review article a reconciliation of packed column
TRANSCRIPT
Review ArticleA Reconciliation of Packed Column Permeability DataColumn Permeability as a Function of Particle Porosity
Hubert M Quinn
TheWrangler Group LLC 40 Nottinghill Road Brighton MA 02135 USA
Correspondence should be addressed to Hubert M Quinn hubertwranglergroupcom
Received 14 November 2013 Accepted 2 January 2014 Published 1 April 2014
Academic Editor Carmen Alvarez-Lorenzo
Copyright copy 2014 Hubert M Quinn This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited
In his textbook teaching of packed bed permeability Georges Guiochon uses mobile phase velocity as the fluid velocity term inhis elaboration of the Darcy permeability equation Although this velocity frame makes a lot of sense from a thermodynamicpoint of view it is valid only with respect to permeability at a single theoretical boundary condition In his more recent writingshowever Guiochon has departed from his long-standing mode of discussing permeability in terms of the Darcy equation andhas embraced the well-known Kozeny-Blake equation In this paper his teaching pertaining to the constant in the Kozeny-Blakeequation is examined and as a result a new correlation coefficient is identified and defined herein based on the velocity frameused in his teaching This coefficient correlates pressure drop and fluid velocity as a function of particle porosity We show thatin their experimental protocols Guiochon et al have not adhered to a strict material balance of permeability which creates amismatch of particle porosity and leads to erroneous conclusions regarding the value of the permeability coefficient in the Kozeny-Blake equation By correcting the experimental data to properly reflect particle porosity we reconcile the experimental results ofGuiochon and Giddings resulting in a permeability reference chart which is presented here for the first time This reference chartdemonstrates that Guiochonrsquos experimental data when properly normalized for particle porosity and other related discrepanciescorroborates the value of 267 for the constant in the Kozeny-Blake equation which was derived by Giddings in 1965
1 Introduction
The value of the constant of proportionality between pressuregradient and fluid flow velocity in a column packed withgranular material and how it relates to column porosity havebeen controversial topics for some time [1] This is especiallytrue in the field of chromatography because columns packedwith porous particles have at least for the last 50 years beenthe main vehicle by which chromatographic separations havebeen carried out and thus this feature has added a newdimension to the controversy The complication arises inpart because of the difficulty of differentiating between thefree space between the particles and the free space withinthe particles in columns packed with porous particles aproblem that does not exist in the case of columns filled withnonporous particles
More recently because current chromatographic ana-lytical and purification applications are being driven by
the need for faster and faster separations the relationshipbetween pressure gradient and high rates of fluid velocityhas taken on a special significance [2] For example it isnow common to see the use of very small particles (circa2 micrometers) in combination with very large operatingcolumn pressures (circa 15000 psi) [3] Because of the obvi-ous impact of high operating pressure on the friability ofchromatographic support particles this development hascreated a new focus on the design criteria for packed columnsespecially with regard to the fractional porosities within thecolumns It will be appreciated that as the internal porosityof chromatographic support particles increases the particlesrsquoability to withstand high pressures decreases This physicalreality would seem to dictate that chromatographic supportparticles suitable for this newly defined operational regimeof high column pressures ought to have a lower particleporosity than those in common use for conventional columnpressures Paradoxically however one finds examples in the
Hindawi Publishing CorporationJournal of MaterialsVolume 2014 Article ID 636507 22 pageshttpdxdoiorg1011552014636507
2 Journal of Materials
recent literature pertaining to fully porous particles wherelow values for the constant of proportionality in the Kozeny-Blake equation are being justified on the basis of particleporosities which purportedly run counter to this principle[4] Clearly something is wrong
The value of the permeability coefficient dictated bythe necessity to balance the pressure gradientfluid flowequation after all independent variables are accounted for(and which presumably has its origin in the fundamentalmechanisms which underlie the generation of pressure in thefirst place) is by no means self-evident From the standpointof first principles it would appear that its value ought to beconstant regardless of the physical parameters comprising thefluid flow apparatus However no fully developed theoreticalexplanation for any constant of permeability has yet beenpublished On the other hand the literature is replete withdiscussions of the value for this parameter which purportedlyare justified on the basis of empirical data The problem isthat the authors of this literature do not agree with eachother and sometimes the same author disagrees with himselfIn the last few years Georges Guiochon in particular haspublished conflicting values for this parameter all of whichare supposedly based uponmeasured values yet he has madeno serious effort to reconcile the conflicts or to provide anyplausible underlying rationale for their differences
As we discuss below there are several inherent flawsin Guiochonrsquos teaching of permeability which are at theroot of his conflicting values for the coefficient We suggestthat these flaws originate in the fact that his teaching doesnot distinguish between the respective impacts on columnpermeability of (a) mass transfer by convection versus masstransfer by diffusion and (b) spherical particles versus par-ticles having an irregular morphology Accordingly withoutsubstantial refinement and adjustment Guiochonrsquos teachingof column permeability at best does little to overcome theshortcomings of Darcyrsquos law and at worst adds fuel to thefire of the aforementioned controversy
2 Guiochonrsquos Textbook Teaching
In his influential textbook Fundamentals of Preparativeand Nonlinear Chromatography [5] Guiochon (togetherwith his coauthors) bases his teaching of the pressureflowrelationship in streamline flow on the Darcy equation whichhe displays therein as his equation (526b) a rearrangementof which we repeat here
Δ119875
119871=
120583119905120578
11989601198892119901119898
(1)
where Δ119875 = pressure drop along column 119871 = length ofcolumn 120583
119905= mobile phase velocity 120578 = absolute viscosity of
fluid 119889119901119898
= average particle diameter and 1198960= Guiochonrsquos
residual permeability coefficientAll terms are self-consistent from a unit-of-measure
perspectiveNote that Guiochon does not discriminate in his expres-
sion of the Darcy equation between spherical and irregularparticles In other words in his statement of the equation
a spherical particle with a measured diameter of 119909 cm isequivalent to an irregularly shaped particle of the samemeasured diameter Peculiarly however he goes on to say thatldquotraditionally a value of 119896
0= 1 times 10minus3 is used for irregular
particles and 12 times 10minus3 for spherical onesrdquo [5 p 153]Similarly without including any porosity dependence
term in the pressurefluid flow relationship or specifyingany particular relationship between columnpressure gradientand column porosity for columns packed with either porousor nonporous particles he states that 119896
0varies ldquobetween 5 times
10minus4 and 2 times 10minus3 depending upon the compactness of thepacked bed that is on the external porosityrdquo [5 p 153]
In essence then Guiochon bundles the variables ofparticle morphology and column porosity into his residualpermeability coefficient 119896
0 Indeed this is why we call it his
residual permeability coefficientIn order to explore these implicit terms which are embed-
ded inGuiochonrsquos teaching for his 1198960 we turn for comparison
purposes to the well-known Kozeny-Blake equation
Δ119875
119871=1198750(1 minus 120576
0)2
120583119904120578
120576301198892119901
(2)
where 1198750= the KozenyBlake permeability coefficient 120576
0=
external porosity of column 120583119904= fluid superficial velocity
and 119889119901= spherical particle diameter equivalent
As we noted above Guiochon acknowledges in his teach-ing that particle irregularity does influence pressure drop Ina modification of the Kozeny-Blake equation proposed byCarman this factor was expressly accounted for by adding aparameter to represent particle sphericity which we denotewith the symbol Ω
119901(Carman gave it the symbol 120601 in his
original publication [6]) It will be appreciated that the valueof Ω
119901is equal to 1 for spherical particles and is less than 1
for particles having an irregular morphology Accordinglythe Kozeny-Blake equation (as modified by Carman)1 isexpressed as
Δ119875
119871=1198750(1 minus 120576
0)2
120583119904120578
12057630(119889
119901119898Ω
119901)2 (3)
where Ω119901= Carmanrsquos particle adjustment factor for devia-
tions from perfect sphericity and 119889119901119898Ω
119901= 119889
119901
As for column porosity note that the Kozeny-Blakeequation includes a combination of porosity terms whichtaken together can be viewed as its porosity dependenceparameter which we notate with the symbol Ψ
119900
Ψ0=(1 minus 120576
0)2
12057630
(4)
where 1205760= the volume of free space between the particles in
the columnthe volume of the empty columnComparing (1) and (3) the reader will observe that
there still is another difference the fluid velocity frames aredifferent Accordingly in order to establish the relationshipbetween the values of Kozeny-Blakersquos 119875
0and the value
of Guiochonrsquos 1198960 one needs to delineate the impact of
Journal of Materials 3
Guiochonrsquos velocity frame upon his permeability coefficientBefore attempting to do this a review of the various velocityframes typically used to characterize flow in porous mediamay be helpful
3 Fluid Velocity Frames
31 Superficial Velocity The fluid velocity term in both theDarcy equation and the Kozeny-Blake equation is the volu-metric flow rate of the fluid divided by the cross-sectional areaof the empty column This velocity is known as the ldquoaveragefluid superficial linear velocityrdquo or simply the ldquosuperficialvelocityrdquo and is defined as follows
120583119904=119876
A (5)
where 119876 = volumetric flow rate 119860 = cross-sectional area ofthe column and 120583
119904= superficial velocity
Thus superficial velocity is not influenced in any way bythe existence or structure of the packed bed Furthermore itdoes not change depending upon whether the particles in thebed are porous or nonporous
32 Interstitial Velocity Theconvective fluid velocity betweenthe particles of a packed bed is called the ldquoaverage fluidinterstitial linear velocityrdquo or simply the ldquointerstitial velocityrdquoInterstitial velocity is the superficial velocity adjusted for thecross-section of the actual fluid flow in a packed columnIt is related to the superficial velocity through the externalporosityof the column 120576
0 as follows
120583119894=120583119904
1205760
(6)
where 120583119894= average fluid interstitial linear velocity
If interstitial velocity is to be substituted for superficialvelocity in the Kozeny-Blake equation one must also includethe external porosity term This term is the compensationfactor that maintains the integrity of the value of the per-meability coefficient 119875
0between pressure gradient and fluid
superficial velocity The equation then becomes
Δ119875
119871=1198750(1 minus 120576
0)2
120583119894120578
120576201198892119901
(7)
33 Mobile Phase Velocity Guiochon does not use either thesuperficial velocity or the interstitial velocity in the equationsin his textbook Instead he uses the ldquomobile phase velocityrdquoMobile phase velocity is uniquely a chromatographic term Itis the velocity calculated as a result of injecting an unretainedfully permeating solute into the column and measuring thetime ofits elution It is defined as follows
120583119905=119871
1199050
(8)
where 1199050= the time it takes an unretained fully permeating
solute to traverse the column and 120583119905=mobile phase velocity
The velocity experienced by the flowing fluid in a columnunder steady state conditions is not altered by the existence ofldquoblindrdquo pores in the particles of that column Blind pores arepores within the particles which have only one fluid openingthat is they are not through-pores Accordingly under steadystate conditions the pores are already filled with identicalstagnant fluid and there exists no driving force to exchangethis stagnant fluid with the fluid flowing in the interstices ofthe particles It will be appreciated that column permeabilitydata is usually taken under steady state conditions in orderto take advantage of a constant value for fluid viscosityAlthough a solute which is dissolved in the flowing fluid willpenetrate the stagnant fluid in the pores (assuming it is notstrictly excluded by either physical dimension or chemicalinteraction) because of molecular diffusion the driving forcefor this solute migration is the concentration gradient of thesolute between the fluid outside the particles and the fluidinside the particles This intraparticle migration of the solutedoes not contribute to friction and thus must not be includedin velocity considerations related to pressure gradientMobilephase velocity therefore is not strictly a fluid velocity terminstead it is a misnomer and it would better be termed theldquounretained fully permeating solute velocityrdquo
Although Guiochon defines the relationship betweenfluid flow and pressure gradient in the context of a residualpermeability coefficient 119896
0 and acknowledges that column
porosity plays a role in that relationship his use of mobilephase velocity fails to recognize that molecular diffusion inblind pores of the particle fraction of a column packed withporous particles does not influence pressure gradient
Mobile phase velocity is related to fluid superficial veloc-ity however through the total porosity of the column ldquoTotalcolumn porosityrdquo 120576
119905 is a term peculiar to the chromato-
graphic literature which relates to the use of columns packedwith porous particles It is the fraction of the column volumetaken up by the free space between the particles plus thefraction of the column volume taken up by the free space inthe internal pores of the particles Thus it is the sum of theexternal and the internal column porosities
120576119905= 120576
0+ 120576
119894 (9)
where 120576119905= total column porosity and 120576
119894= internal column
porosityMobile phase velocity is related to fluid superficial veloc-
ity through the total porosity of the column as follows
120583119905120576119905= 120583
119904 (10)
Thus if mobile phase velocity is used in the Kozeny-Blakeequation one must insert the compensation factor 120576
119905into the
equation in order to maintain the integrity of the value of thepermeability coefficient 119875
0 between pressure gradient and
fluid superficial velocity and the equation then becomes
Δ119875
119871=1198750(1 minus 120576
0)2
120583119905120576119905120578
120576301198892119901
(11)
4 Journal of Materials
And now reconciling (1) and (3) we get
1198750(1 minus 120576
0)2
120583119904
120576301198892119901
=120583119905120576119905
1198960(119889
119901119898Ω
119901)2 (12)
4 The Flow Resistance Parameter
Before proceeding any further we digress to define anotherterm which will become important to our understanding ofGuiochonrsquos work The chromatographic literature containsa term called the ldquoflow resistance parameterrdquo [7] The fluidvelocity underlying this parameter is based upon Giddingsrsquodefinition of ldquomean fluid velocity through a cross-sectionrdquoand is found in his 1965 text book at page 207 [8]2 It is definedas follows
120601 =Δ119875119889
2
119901
120583119905120578119871
(13)
where 120601 = flow resistance parameterThe flow resistance parameter therefore is equivalent to
the reciprocal of Guiochonrsquos term 1198960
120601 =1
1198960
(14)
5 Refining the Terms For Column Porosity
In order to continue with our comparison of Guiochonrsquosteaching to that embodied in the Kozeny-Blake equation andin particular to accommodate columns packed with porousparticles we need to refine the porosity dependence term inthe Kozeny-Blake equation To accomplish this we adopt thedefinition of column porosity as taught by J Calvin Giddings[8] namely
Ψ =(1 minus 120576
119905Φ)
2
(120576119905Φ)
3 (15)
where Ψ = the porosity dependence term in Kozeny-Blake(porous particles) and
Φ =1205760
120576119905
(16)
where Φ = the ratio between the external and total columnporosities
Using thismethodology in themost general case involvingporous particles to express total column porosity (where totalcolumn porosity is not the same as external column porosity)as opposed to in the special case involving nonporous particles(where total column porosity and external column porosityare the same) the Kozeny-Blake equation (using mobilephase velocity as in (10)) may now be reexpressed as follows
Δ119875
119871=1198750(1 minus 120576
119905Φ)
2
120583119905120576119905120578
(120576119905Φ)
3
(119889119901)2
(17)
It will be appreciated that in the special case of columnspacked with nonporous particles (17) is identical to (3)because here 120576
119905= 120576
0and thus 120583
119905= 120583
119894= 120583
119904120576
0and Φ = 1
6 The Relationship between 1198960
and 1198750
We can now connect the teaching of Guiochon to thatembodied in the Kozeny-Blake equation as follows
1198750=
1
1198960Ψ
(18)
or
1198750=120601
Ψ (19)
As illustrated by (19) the term 1198750in the Kozeny-Blake
equation is the flow resistance parameter normalized for thecolumn porosity dependence term
Note that if1198750is a true constant the two parameters upon
which it depends 120601 and Ψ will vary in direct proportion toeach other This is illustrated by the data in Table 1 hereinwhich is a reference chart where particle porosity 120576
119901 is varied
from nonporous at one extreme to highly porous at the otherwhere 120576
119901= volume of the free space within the particles in
a packed columnthe total volume of all the particles in thecolumn
Figure 1(a) is a plot of the data contained in Table 1 as120601 versus Ψ Proceeding from the lower left hand corner tothe upper right hand corner of the plot particle porosityincreases It is obvious from the plot that all the data in Table 1falls on a single line with a slope of 267 which we say is theconstant value of 119875
0 As is illustrated by the plot in the range
of porosity covered by the data the values for 120601 vary fromabout 500 to about 1900 and correspondingly the values ofΨvary from approximately 2 to approximately 7 Note also thatthe upper boundary value of Ψ for packed columns is about55 with larger values related to capillaries
Finally we point out that once the value of 1198750is correctly
identified ameasured value for the flow resistance parametermay then be used to calculate the value of the externalporosity of the column 120576
0 when the latterrsquos value has not or
cannot be measured
7 Guiochonrsquos Modified PermeabilityCoefficient 119875
119905
As reflected in some of his recent publications which wewill explore in more depth later Guiochon inadvertentlyapparently confuses the term 119875
0with another parameter 119875
119905
which we identify here and define for the first time 119875119905is not
the same as 1198750because it is based upon Guiochonrsquos unique
(from a permeability perspective) teaching for fluid velocityIt is defined as follows
119875119905= 119875
0120576119905 (20)
where 119875119905= Guiochonrsquos modified permeability coefficient
It will be appreciated that Guiochonrsquos 119875119905 if used in the
Kozeny-Blake equation in place of 1198750 accommodates his use
of themobile phase velocity in away that is different from butequivalent to a direct conversion of mobile phase velocity tosuperficial velocity
Journal of Materials 5
Table1Colum
nperm
eabilityreferencetable
119870119865
Particlepo
rosity
rang
e120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119896119871
119863Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
120576119905Φ
(120576119905minus1205760)
120576119894
Δ119875119889119901
2Δ119875119889119901
2(1minus120576119905Φ)2120576119905
(1minus1205760)2
1120601
1198750120576119905
119889119901
2120576119905
(119889119901119898Ω119901)
(1minus1205760)
120583119905120578119871
120583119904120578119871
(120576119905Φ)3
(1205760)3
120601Ψ
120601po
iseUnits
Non
eNon
eNon
eNon
eNon
epsi
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
cmNon
ecm
cmgcmminus1 secminus1
cm3 m
inminus1
cmsecminus
1cm
secminus
1cm
secminus
1
Non
porous
particles
0380
1000
0380
000
0000
0160
711
1870
2662
701
141119864minus03
267
101
535119864minus10
15046
100
000100
000100
00100
038
0038
0100
0100
0390
1000
0390
000
0000
0147
653
1675
244
6627
153119864minus03
267
104
597119864minus10
15046
100
000100
000100
00100
039
0039
0100
0100
040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
100
000100
000100
00100
040
004
00100
0100
0410
1000
0410
000
0000
0124
553
1349
2071
505
181119864minus03
267
109
742119864minus10
15046
100
000100
000100
00100
041
004
10100
0100
0420
1000
0420
000
0000
0115
509
1212
1907
454
196119864minus03
267
112
825119864minus10
15046
100
000100
000100
00100
042
004
20100
0100
Lowpo
rosity
particles
0507
0750
0380
0127
0204
213
948
1870
3549
701
106119864minus03
267
135
535119864minus10
15046
100
000100
000100
00100
051
0051
0133
0100
0520
0750
0390
0130
0213
196
871
1675
3261
627
115119864minus03
267
139
597119864minus10
15046
100
000100
000100
00100
052
0052
0133
0100
0533
0750
040
00133
0222
180
801
1502
2999
563
125119864minus03
267
142
666119864minus10
15046
100
000100
000100
00100
053
0053
0133
0100
0546
0750
0410
0136
0231
166
737
1349
2760
505
136119864minus03
267
146
742119864minus10
15046
100
000100
000100
00100
055
0055
0133
0100
0560
0750
0420
0140
0241
153
679
1212
2542
454
147119864minus03
267
149
825119864minus10
15046
100
000100
000100
00100
056
0056
0133
0100
Medium
porosity
particles
0613
0620
0380
0233
0376
258
1146
1870
4293
701
872119864minus04
267
164
535119864minus10
15046
100
000100
000100
00100
061
0061
0161
0100
0629
0620
0390
0239
0392
237
1053
1675
3946
627
949119864minus04
267
168
597119864minus10
15046
100
000100
000100
00100
063
0063
0161
0100
064
50620
040
00245
040
9218
969
1502
3629
563
103119864minus03
267
172
666119864minus10
15046
100
000100
000100
00100
064
0065
0161
0100
0661
0620
0410
0251
0426
201
892
1349
3340
505
112119864minus03
267
177
742119864minus10
15046
100
000100
000100
00100
066
006
60161
0100
0677
0620
0420
0257
044
4185
821
1212
3076
454
122119864minus03
267
181
825119864minus10
15046
100
000100
000100
00100
068
006
80161
0100
Highpo
rosity
particles
0760
0500
0380
0380
0613
320
1422
1870
5325
701
703119864minus04
267
203
535119864minus10
15046
100
000100
000100
00100
076
0076
0200
0100
0780
0500
0390
0390
0639
294
1306
1675
4891
627
766119864minus04
267
208
597119864minus10
15046
100
000100
000100
00100
078
0078
0200
0100
0800
0500
040
0040
0066
7270
1202
1502
4500
563
832119864minus04
267
214
666119864minus10
15046
100
000100
000100
00100
080
0080
0200
0100
0820
0500
0410
0410
0694
249
1105
1349
4140
505
905119864minus04
267
219
742119864minus10
15046
100
000100
000100
00100
082
0082
0200
0100
0840
0500
0420
0420
0724
229
1018
1212
3814
454
982119864minus04
267
224
825119864minus10
15046
100
000100
000100
00100
084
0084
0200
0100
Capillary
1000
0380
0380
0620
1000
421
1870
1870
7005
701
535119864minus04
267
267
535119864minus10
15046
100
000100
000100
00100
100
0100
0263
0100
1000
0390
0390
0610
1000
377
1675
1675
6273
627
597119864minus04
267
267
597119864minus10
15046
100
000100
000100
00100
100
0100
0256
0100
1000
040
0040
0060
010
00338
1502
1502
5625
563
666119864minus04
267
267
666119864minus10
15046
100
000100
000100
00100
100
0100
0250
0100
1000
0410
0410
0590
1000
303
1349
1349
5051
505
742119864minus04
267
267
742119864minus10
15046
100
000100
000100
00100
100
0100
0244
0100
1000
0420
0420
0580
1000
273
1212
1212
4541
454
825119864minus04
267
267
825119864minus10
15046
100
000100
000100
00100
100
0100
0238
0100
6 Journal of Materials
0200400600800
100012001400160018002000
20 30 40 50 60 70
NonporousLow porosityMedium porosity
CapillaryLinear (line)
Nonporous Low porosity particles
Medium porosity particles
High porosity particles
Capillary
Line slope (P0) = 267
y = 267x + 0
Ψ
120601
Line
High porosity
(a)
99111123135147159171183195207219231243255267
035 045 055 065 075 085 095
LineCapillaryLinear (line)
Capillary
Line slope (P0) = 267
y = 267x + 0
Pt
120576t
Φ = 10
Φ = 075
Φ = 062
Φ = 05
Φ = 075
Φ = 062
Φ = 05
Φ = 1
(b)
99111123135147159171183195207219231243255267
000 010 020 030 040 050 060 070 080 090 100
NonporousLow porosityMedium porosityHigh porosity
120576p
y = 160x + 107
Pt
LineCapillaryLinear (line)
(c)
Figure 1 (a) The balance between 120601 and Ψ (b) Column permeability reference chart (c) Particle permeability reference chart
In the 2006 edition of their textbook [9] Guiochon andhis coauthors provide the following worked example of hispressure dropfluid flow teaching3
In Table 2 a cross-reference is established between Guio-chonrsquos teaching for columns packed with both porous andnonporous particles the Kozeny-Blake equation (asmodifiedby Carman) and for validation purposes Giddingsrsquo teachingbased upon actual permeabilitymeasurements fromhis Table53-1 in his 1965 text [8]4
In his worked example Guiochonrsquos fluid velocity 119906is found in his text as his equation (256) [9] at page61 and equates to mobile phase not superficial velocityAccordingly when applied to columns packed with porousparticles his worked example combines the contributionto fluid velocity of mass transfer by molecular diffusionin the free space within the particles with mass transferby convection in the free space between the particles Asdiscussed above if one is attempting to compare Guiochonrsquos
teaching for the value of the permeability coefficient with thatof the Kozeny-Blake equation the conversion factor is thetotal porosity of the column
In addition because Guiochonrsquos particle sizeparameter has an embedded assumption due to particleshaperoughness his worked example also commingles thecontribution to spherical particle diameter equivalent ofparticle shaperoughness with other nonparticle variablesUnfortunately his worked example is not based uponcompletely spherical particles where this would not be anissue We know this because he states that 119896
0in his worked
example equals 1 times 10minus3 which he says is the value of thepermeability coefficient for columns packed with irregularparticles
In order to move forward one must simplify If this wereto be done experimentally it would best be accomplishedby experiments with nonporous particles in which case thevariable of molecular diffusion in the pores of the particles
Journal of Materials 7
Table2Cr
oss-referencefor
term
s119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119896119871
119863Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
120576119905Φ
(120576119905minus1205760)
120576119894
Δ119875119889119901
2Δ119875119889119901
2(1minus120576119905Φ)2120576119905(1minus1205760)2
1120601
1198750120576119905
119889119901
2120576119905
(119889119901119898Ω119901)
(1minus1205760)
120583119905120578119871
120583119904120578119871
(120576119905Φ)3
(1205760)3
120601Ψ
120601po
iseUnits
Non
eNon
eNon
eNon
eNon
epsi
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
cmNon
ecm
cmgcmminus1 secminus1cm
3minminus1cm
secminus
1cm
secminus
1cm
secminus
1
Nonporousp
articles
Gidding
srsquotradition
alno
nporou
scolum
n040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
1000010
00010
0010
0399
004
00100
0100
Guiocho
nrsquostradition
alspheric
alparticle(app
arent)
0362
1000
0362
000
0000
0187
830
2293
3106
858
120119864minus03
267
97436119864minus10
15046
1000010
00010
0010
0361
0036
0100
0100
Gidd
ingrsquos
table5
3-1
5060meshglassb
eads
0422
1000
0422
000
0000
0113
500
1184
1873
444
200119864minus03
267
113
844119864minus10
15046
1000010
00010
0010
0421
004
20100
0100
5060meshglassb
eads
040
910
00040
9000
0000
0126
560
1371
2097
513
179119864minus03
267
109
729119864minus10
15046
1000010
00010
0010
040
70041
0100
0100
Guiochonrsquosteaching
ExA
irregular
particle
(app
arent)
040
010
00040
0000
0000
0225
1000
2500
2250
563
100119864minus03
444
178
400119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
ExA
correctedforp
article
irregularity
040
010
00040
0000
0000
0225
601
1502
2250
563
166119864minus03
267
107
400119864minus10
15046
078
00010
000
080010
0399
004
00100
0100
ExB
spheric
alparticle
(app
arent)
040
010
00040
0000
0000
0187
830
2075
2250
563
120119864minus03
369
148
482119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
ExB
corrected
040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
Porous
particles
Gidding
srsquotradition
alpo
rous
column
060
0066
7040
00200
0333
203
900
1500
3375
563
111119864minus03
267
160
667119864minus10
15046
100
00010
00010
0010
0599
006
00150
0100
Gidd
ingrsquos
table5
3-1
3040meshalum
ina
0803
0498
040
00403
0672
271
1204
1499
4510
562
830119864minus04
267
214
667119864minus10
15046
100
00010
00010
0010
0801
0080
0201
0100
5060meshalum
ina
0837
0499
0417
0420
0721
235
1043
1246
3906
467
959119864minus04
267
224
803119864minus10
15046
100
00010
00010
0010
0835
0084
0201
0100
6080meshchromosorbW
(5DNP)
0766
0498
0382
0384
0621
316
1404
1833
5259
687
712119864minus04
267
204
545119864minus10
15046
100
00010
00010
0010
0764
0077
0201
0100
6080meshchromosorbW
(20
DNP)
0785
0495
0389
0396
064
8300
1333
1698
4991
636
750119864minus04
267
210
589119864minus10
15046
100
00010
00010
0010
0783
0079
0202
0100
8 Journal of Materials
would be eliminated and by using only smooth sphericalparticles such as glass beads in which case the variableof particle shaperoughness would be eliminated If both ofthese things were done one would be able to directly andunambiguously determine the true value of 119875
05
Accordingly since his textbook contains a general teach-ing without any expressed restrictions with respect to particleporosity let us first assume that the column in Guiochonrsquosworked example is packed with nonporous particles Sec-ondly let us assume that the value for the external porosityof the column (which is the same as the total porosity ina column packed with nonporous particles) is the valuetraditionally assumed for a typical well-packed column thatis 04
As illustrated by example A in Table 2 this would leadto an apparent value for 119875
0of 444 which is obviously much
too high6 Using the adjusted particle size of 8 micrometers(rounded) which is calculated based on a value of 078 forΩ
119901
we derive values of 267 for 1198750and 107 for 119875
119905(the relationship
between these two values is of course equal to the totalporosity of the column 120576
119905 as is dictated by (20))
Similarly as illustrated by example B in Table 2 when thecolumn of Guiochonrsquos worked example is adapted to accom-modate spherical nonporous particles Guiochonrsquos teachingof the value for 119896
0of 12 times 10minus3 results in an apparent value
of 369 for 1198750
7 again too high [10 11] On the other hand ifwe correct the value of 119875
0 we get a column external porosity
of 03620which is too low for a typical well-packed column(120576
0= 04) [8] Moreover in the case of both corrections
Guiochonrsquos teaching overstates the pressure gradient (187 psicompared to the 135 psi which Giddingsrsquo experiments wouldgenerate) Accordingly it is only when we correct for thepressure drop that we achieve appropriate values of 04 for1205760 267 for 119875
0 and 107 for 119875
119905
As is evident from Table 2 when the porosity of columnspacked with nonporous particles varies the value of 119875
119905also
varies because it is based upon the measurement of mobilephase velocity which in turn changeswith the porosity of thecolumn Indeed for typical columns packed with nonporousparticles having external porosities close to 04 (038ndash042)[8] Table 2 reveals that the value of 119875
119905varies by as much as 6
points while the value of 1198750remains constant at 267
Having established a range of values for 119875119905based upon
columns packed with nonporous particles one can nowproceed to evaluate the range of values for 119875
119905based upon
columns packed with porous particles which is to say thatwe can now evaluate the impact of particle porosity onpermeability Among other things Table 2 shows that whenit comes to porous particles Guiochonrsquos 119875
119905is even more
variable (and thus even less useful) because the value ofthe coefficient can differ on a column-to-column basis byas much as 90 points depending on the combination of theinternal andexternal column porosities at play
Figures 1(a) and 1(b) are a plot of the data found in Table 1In this plot Guiochonrsquos modified permeability coefficient 119875
119905
is plotted against column total porosity 120576119905 Once again it will
be appreciated that (a) all data falls on a line with a slope of267 which represents the value of 119875
0and (b) the line has no
intercept which means that when 120576119905is zero 119875
119905is also zero As
shown in the plot values of the parameter Φ are highest atlow values of 119875
119905and lowest at high values of 119875
119905 Further as is
evidenced by the respective definitions for the terms119875119905and119875
0
contained herein the only time that the value of Guiochonrsquos119875119905does not differ from the value of 119875
0(267) is when the value
of 120576119905is 1 its value in a theoretical packed column devoid of
particles that is acapillaryFinally Figure 1(c) is another plot of the data in Table 1
but this time the values for 119875119905are plotted against particle
porosity 120576119901 The equation of this line does have an intercept
which represents the value of 119875119905when the particles are
nonporous It is also obvious from the equation of this linethat it is only when particle porosity is unity that is in acapillary that the value of 119875
119905is 267
It will be appreciated that since 120576119905is the sumof the internal
and external porosities of a packed column any given valuefor 119875
119905involving a column packed with porous particles can
represent many different combinations of these parametersVice versa any given value for 120576
119905can generate many different
values for 119875119905depending on the amount of particle porosity
which is present On the other hand as is pointed outearlier column internal porosity andor particle porosityplay no role in determining the permeability of a packedcolumn the only porosity parameter which is relevant forsuch purposes is column external porosity In other words asa permeability coefficient 119875
119905is not a particularly informative
concept because it still fails to provide any meaningful tie tothe underlying physical phenomena that govern the pressuredropfluid flow relationship
In the case of permeability measurements one quan-tifies only total pressure drop Pressure drop however isa commingling of the contributions of the several differ-ent parameters embodied in the pressureflow relationshipTherefore one must accurately represent the contributionof each factor before adding them together to equal thepressure drop Obviously if one mistakes the contribution ofany given element for instance particle size this will skewthe contribution of some other element when summed forinstance column porosity
Our frame of reference has been calibratedwith respect tothe interrelationships between the values of 119875
0 119875
119905 120576
119905 and 120576
119901
based on the data reported in Giddingsrsquo table 53-1 in his 1965textbook Since his data base was derived from experimentsusing nonporous particles the permeability results of whichwere extended to columns packed with fully porous particlesby using his Φ parameter the validity of which in turn wasestablished by independent means there is no uncertaintywith respect to the critical variable of external porosity 120576
0
inherent in our frame of reference It is for this reasontherefore that our frame of reference trumps any reporteddata which uses any other technique including the techniqueof inverse size exclusion chromatography by itself to establishvalues for 120576
0[12] In addition our frame of reference is
internally consistent due to the grounding of the parametersat the boundary condition of a theoretical capillary at whichpoint the values of 120576
119905and 120576
119901are unity and the values of 119875
0and
119875119905are equal In essence these interrelationships exemplify the
conservation laws of nature which are the ultimate standard
Journal of Materials 9
of measure when considering partial fractions of a singleentity
Consequently as Figures 1(b) and 1(c) illustrate if oneknows based upon other independent means the value of acolumnrsquos total porosity or its particlesrsquo porosity respectivelyone can at least verify whether results of permeability exper-iments expressed in terms of 119875
119905are reasonable Or to put
it another way if the 119875119905results reported do not conform to
what these plots tell us one should expect from columns andparticles based upon independently derived porosity valuesfor the particles under study we know that something iswrong Accordingly this is the way in which we use 119875
119905below
to analyze and evaluate Guiochonrsquos experimental work of thelast decade
8 The Origin of Guiochonrsquos Value forColumn Permeability
In 1966 Guiochon wrote a review paper entitled ldquoProblemsRaised by the Operation of Gas Chromatographic Columnsrdquo[13] In that paper he reviewed the development of the per-meability equation from Darcy to the then present Amongother things he reported that ldquoa mean 120576 of 038 is foundfor a number of packings obtained from powdered prod-ucts of various origins consisting of spherical or irregularparticles Almost all the measured values are between035 and 045rdquo (p 14) He then went on to discuss what hecalled the ldquosemiempirical Blake-Kozeny equationrdquo which hereported as 119896 = [119889
2
1199011205762][150(1 minus 120576)
2] [13 p 15 Equation
(11)] What sticks out of course is the fact that Guiochonstates here that 119875
0equals 150 the value for the permeability
coefficient championed by Ergun not the value posited byCarman To be fair we should point out that Guiochongave warning that he was uncomfortable with any particularvalue for the permeability coefficient For even though herecited the Kozeny-Blake equation with a value of 150 for itspermeability coefficient he stated that ldquothe values of 119896 givenby (this) equation are not very accurate and the deviationscan be as large as 50rdquo [13 p 15] Guiochon attributedmost ofthis uncertainty to the inability of practitioners of that era toeffectively measure the value of a columnrsquos external porositySince as he pointed out ldquo(the Kozeny-Blake) equation isvery sensitive to variations in 120576rdquo he concluded that theequation is ldquoof little userdquo and that it was better practice torely upon grosser measurements of permeability [13 p 15]Thus Guiochon at this timeframe made a conscious decisionto staywithin the relatively safe but admittedlymore obscureboundaries of Darcyrsquos teaching on column permeability
In 2006 however Guiochon emerged from under theshadow of Darcy permeability when he published a reviewpaper which purported to be a comprehensive summaryof the current state of the art in chromatography [14]In light of the fact that Guiochon and his coauthors hadvirtually contemporaneously published a new version of theirtextbook one would have expected that Guiochon wouldtake this opportunity to reaffirm his textbook teaching So itshould come as no surprise that he would open his discussionof the pressure dropfluid flow relationshipwith the following
assertion ldquothe permeability of packed beds used in liquidchromatography is in the order of 1 times 10minus3rdquo [14 p 9] Asone would expect the authority which Guiochon cites forthis seemingly familiar proposition is the 2006 edition of hisown textbook For the first time however he also providessome citations to third party sources In particular he cites a1997 chromatography handbook written by Uwe Neue ChiefScientist for Waters Corporation [15 p 9] As one can seeNeuersquos equationmdashlike Guiochonrsquosmdashdoes indeed postulate apermeability coefficient with a value of 1 times 10minus3
At this point however Guiochon abandons his tradi-tional caution about the value of the permeability coeffi-cient and marries his permeability equation to the Kozeny-Blake equation thereby endorsing not only the porositydependence term in that equation but also a specific valuefor 119875
0 Referring to his value of 1 times 10minus3 he states the
following ldquothe specific column permeability is related to thecolumn porosity through the Kozeny-Carman equation 119896
119891=
12057630180 (1 minus 120576
0)2rdquo [14 p 9] Not surprisingly Neue makes
the same connection ldquothe specific permeability depends onthe particle size 119889
119901and the interstitial porosity 120576
0of the
packed bedThe relationship is known as theKozeny-Carmanequation 119861
0= (1185)(1205763
01198892119901(1 minus 120576
0)2)rdquo [15 p 30]
Neuersquos permeability equation and Guiochonrsquos textbookpermeability equation however differ in that Neue uses thesuperficial not the mobile phase velocity Accordingly sinceall other features of their equations are the same one wouldexpect them to get different values for their permeabilitycoefficients Neue himself recognizes this when he says inhis handbook ldquothis value [1000] is also in good agreementwith our experience but values as low as 500 have beenreported in the literature8 However there are several reasonswhy different values reported by different workers do notrepresent true differences in the resistance parameter Firstlower values are obtained for porous packings if the specificpermeability is measured on the basis of the breakthroughtime of an unretained peak instead of using the flow raterdquo see[15 p 32] (when he distinguishes the ldquobreakthrough time ofan unretained peakrdquo from the ldquoflow raterdquo Neue is of coursereferring to the mobile phase velocity and the superficialvelocity resp) Consequently if one takes Neuersquos point andturns it on its head one can see that if he and Guiochonwere to get the same value for their permeability coefficients[1000] but one (Guiochon) is using mobile phase velocityto derive it and the other (Neue) uses superficial velocity toderive it this would indeed represent a ldquotrue differencerdquo intheir equations
So how can Guiochon cite both Neue and himself for theproposition in his review paper that ldquothe permeability of thepacked beds used in liquid chromatography is in the order of1times 10minus3rdquoThe answer is that at least for purposes of his reviewpaper he adopts Kozeny-Blakersquos fluid velocity convention inplace of the velocity frame he uses in his textbook Here iswhat he says ldquothis approximation [1 times 10minus3] is valid only ifthe superficial velocity is usedrdquo [14 p 9] (emphasis supplied)
As a matter of pure mathematics Guiochonrsquos permeabil-ity coefficient of 1 times 10minus3 can only be equal to 1205763
0180(1 minus 120576
0)2
if the value of 1205760is 04 the typical value for the interstitial
10 Journal of Materials
fraction of a well-packed column Guiochon confirms in hisreview paper that this is indeed what he assumes in his state-ment of the Kozeny-Blake equation [14 p 9] Accordinglyhis adoption of the value of 180 for 119875
0solely depends on the
validity of his assertion that the value of the permeabilitycoefficient is 1 times 10minus3
We already know from our discussion of Guiochonrsquostextbook teaching that when this value is associated with apermeability equation based on mobile phase velocity it iswrong As reflected in our Table 2 when Guiochonrsquos workedexample involving this figure is corrected to account for theembedded factor due to the irregularity of the particles whichhis worked example assumes the value of the permeabilitycoefficient generated by his equation is not 1 times 10minus3 it is 166times 10minus3
So where did this value of 1 times 10minus3 come from AlthoughGuiochon in his review paper cites Neuersquos handbook for thisvalue Neue himself offers absolutely no authority for it Onthe other hand Guiochon does provide one other citation inhis review paper for this value The citation is to an article byHalasz et al published in 1975 [16]
If one then goes to that article one sees that Halaszet al do indeed proffer the same equation as Neue (albeitin a somewhat different format) and yes it does containthe value of 1 times 10minus3 See [16 Equation (1)] Neverthelessthe authors fail to identify from whence they obtained thisvalue On the other hand they do make reference to apaper that Endele et al wrote a year earlier in 1974 withKlaus Unger entitled ldquoInfluence of the Particle Size (5ndash35 120583)of Spherical Silica on Column Efficiencies in High-PressureLiquid Chromatographyrdquo [17] It appears that this is wherethe body is buried for it is in this article that Halasz reportshaving actually derived a value of 1 times 10minus3 for the permeabilitycoefficient
In summarizing the results of their experiments whichincidentally are based on measurements of superficial notmobile phase velocity (see their Equation (7)) Halasz et alcalculate an average value for the permeability coefficient(which they notate as119870
119865) of 1198892
1199011080 Accordingly they state
the following ldquo[I]t is proposed to define an average particlesize 119889
119901 for columns packed with spherical silica using the
balanced density packing method by the equation 1198892119901
=
1198701198651000 = 1198651205781198711031199032120587Δ119875rdquo [17 p 382 their Equation (7)]To recapitulate it thus appears that (1) Halasz was the
source of the value of 1 times 10minus3 for the permeability coefficientwhich Guiochon adopts in his review paper (2)Halaszrsquo valuewas based upon measurements which involved sphericalparticles and (3)Halaszrsquo value was properly derived from anequation which utilized the superficial velocity as the fluidvelocity
However we still have a problem if we are going to acceptthe value of 1 times 10minus3 as a basis for calculating the constantin Kozeny-Blake we still need to know the value of Halaszrsquocolumnsrsquo external porosity For example in the absenceof measured values for the external porosity Guiochonrsquosassertion in his review paper that the value of the constantis 180 is without foundation
Unfortunately however Halasz bases his methodologyon ldquoassumedrdquo versus ldquomeasuredrdquo values for external porosityNevertheless he does provide some clues that may helpus peel back this onion First he says that his value forthe permeability coefficient applies to ldquocolumns using thebalanced density packing methodrdquo Secondly after statinghis proposed value for the coefficient we find the followingtelltale sentence ldquothe permeability of columns packed by theconventional ldquodryrdquo method are worse by a factor of about 2than those packed by the balanced density methodrdquo [17 p382] For this proposition Halasz cites yet another of his ownarticles this time a paper written by himself and M Naefein 1972 entitled ldquoInfluence of Column Parameters on PeakBroadening in High-Pressure Liquid Chromatographyrdquo [18]
When one follows this thread by going to Halaszrsquos 1972paper one finds a possible solution to the puzzle For this iswhere Halasz describes his calculations of the permeabilitycoefficient from experiments with a number of conventionaldry-packed columns containing spherical silica particlesNot surprisingly the resulting value for the permeabilitycoefficient for such columns was not 1 times 10minus3 Instead Halaszsays that his experiments with such columns resulted inthe following value for the permeability coefficient ldquo119870 sim
11988921199012000rdquo [18 p 78] In other words consistent with what
Halasz says in his 1974 article the value for the coefficient forthese dry-packed columns was indeed ldquoworse by a factor ofabout 2rdquo compared to its value for columns prepared with thebalanced density method
The obvious implication of this is that when Halasz gota value of 1 times 10minus3 for the permeability coefficient in 1974he must have been evaluating columns that were packed toan interstitial fraction considerably higher than the externalporosity of the columns that were the subject of his 1972experiments Our problem however is still not resolvedbecause Halasz did not accurately determine the externalporosity of the columns under study in either the 1972 or 1974columns
We surmise that in their 1972 paper Halasz and hiscoauthors made use of the teaching of Giddings outlinedin his 1965 textbook which at the time of their 1972 paperwas seven years old Here is what they say ldquoexperiencehas shown that in a regular packed column with the sievefractions usual for chromatography 120576
0is independent of the
particle size if the particles are more or less spherical In anacceptable approximation 120576
0is equal to 04 In those columns
packed with nonporous supports 119906V and 119906 are identicalUsing porous supports (ie silica gel alumina chromosorbPorasil and so forth) 119906 = 4119865(1205871198892
119888120576119879) where 120576
119879is the
total porosity and indicates the pore volume of the supportrdquoIt will be appreciated that 119906V stands for interstitial velocity(our 120583
119894) and 119906 stands for mobile phase velocity (our 120583
119905) The
ldquoexperiencerdquo which Halasz and his coauthors are referringto concerning ldquocolumns packed with nonporous supportsalumina and chromosorbrdquo is that recorded by Giddings inTable 53-1 in his 1965 text Continuing to establish theinterrelationships inherent in the various entities involvedin chromatographic columns with respect to both interstitialandmobile phase velocity the authors conclude ldquoformally 119906V
Journal of Materials 11
can be calculated for a porous support assuming 1205760is equal
to 04rdquo (emphasis supplied) The authors then announce themethodology underlying their frame of reference with regardto column permeability as follows ldquowe determined in all ofour columns the 119906119906V ratio to be 056 with a differentialrefractometer Consequently the total porosity 120576
119905was equal
to 0714rdquo It will be appreciated that their ratio of 119906119906V is oneand the same as GiddingsrsquoΦ parameter as utilized by him inhis 1965 textbook and as we define hereinabove
We can now identify the origin of the discrepancybetween Halasz and Giddings and by extension betweenHalasz and Guiochon (and Neue) For even though Halaszmakes use of the results in Giddingsrsquo Table 53-1 he failsto appreciate the teaching inherent therein regarding theimportance of an accurate value for column external porositywhen using porous particles In establishing his values forΦGiddings was careful to ground his values for the externalporosity of columns packed with porous particles in acomparison of their measured pressure drops with measuredpressure drops in columns packed with nonporous particleswhere the values for (external) porositywere accurate becausethey are equal to the columnsrsquo total porosity
In their 1972 paper Halasz et al used a refractometer todetect the inert solute 119899-pentane in the mobile phase of 119899-heptaneThey injected the 119899-pentane into the flowing mobilephase the flow rate of which they presumably measured witha volumetric cylinder and stop watch Since the exit timeof the 119899-pentane peak was identified by the refractometertheir measurement of holdup volume was accurate andaccordingly so was their value of 0714 for 120576
119905(there is no
uncertainty related to column total porosity when using inertsolutes)This in turn led to an accurate value for their119906 termmobile phase velocity However since they did not measurethe interstitial velocity 119906V but rather used an estimate basedupon the measured flow rate and an ldquoassumedrdquo value of 04for external porosity their calculatedestimated value for 119906Vwas arbitrary This in turn means that the value of 056 fortheirΦ ratio was likewise arbitrary
In Table 3 herein we show the results for column number1 in the 1972 study which we designate as Column A
1 Note
that the authorsrsquo raw values correspond to a value of 500 for1198750 a valuewhich is obviously far too high As is plainly evident
in our Figures 2 and 3 this column data set does not conformin any reasonable fashion to the norms in our referencecharts Accordingly the data is badly flawed In our correcteddata for Column A
1in Table 3 herein we adjust the column
external porosity to the lowest value permitted in our frame ofreference (038) This correction results in a revised value forGiddingsrsquoΦparameter of 0532 In additionwe are compelledto adjust the particle downward (to 11 micrometers approx)as is also dictated by our frame of reference We justify ourlow corrected value for external porosity and our downwardadjustment for particle size on the aggressive ldquodry-packingrdquoprocess described by the authors ldquo[T] the best results wereachieved with the following method the column (id =2mm) 25 cm long was packed with 25 equal portions of theldquobrushrdquo After adding each portion the column was vibratedand afterward tapped on the floor and also vigorously tappedwith a rod from the siderdquo Based upon a column packing
experience of the principal author herein spanningmore than30 years we believe that these packing conditions generateda packed column with a low external porosity and also one inwhich the particles experienced significant breakage
The main differences between the columns in the 1972and 1974 papers concerned the particle sizes and themethodsused to pack the columns The particle diameters in the 1972study fell in a range of about 15ndash200 micrometers while thosein the 1974 study were much smaller being in the range of 5ndash40 micrometers approximately Moreover while the columnsin the 1972 study were dry-packed the columns in the 1974study were slurry-packed (what Halasz calls ldquothe balanceddensity packing methodrdquo) The authors of the 1974 paperrecite that their permeability results shown in their Table 4correspond to values for 120601
0of 1080 and values for 120601 of 910
In Table 3 herein we show the results for the 1974 study as anaverage value in our column designated as A
2 Note that the
raw data reported by the authors which again has the built-inassumption that the column external porosity was 04 placesthe value of 119875
0at 192 In our corrected values for Column A
2
we show that a value of 267 for 1198750places the column external
porosity at a value of 043 The ratio of the value for 120601 of 910when compared to the value for 120601 in the 1972 study of 2007is about 22 (2007910 = 22) which is in good agreementwith the authorsrsquo statement that the permeability of the 1972columns was worse by a factor of about 2 Again based onthe experience of the principal author herein in the packingof hundreds of thousands of commercially sold slurry packedcolumns we conclude that the balanced density techniquedescribed by the authors in the 1974 paper is fully capable ofproducing columns with this high and external porosity
In summary we conclude that for purposes of theKozeny-Blake equation (a) the external porosity of theHalasz 1972 columns was 038 (b) the external porosity of thecolumns which he tested in 1974 was 043 and (c) Halaszrsquos1972 and 1974 studies when properly analyzed corroborateGiddingsrsquo value for 119875
0of 267 not Carmanrsquos value of 180
Although the fault for all of the misdirection about the valueof the permeability coefficient being 1 times 10minus3 therefore restsprimarily on the shoulders of Halasz one has to question whyGuiochon (andNeue) so readily adopted itWhatevermay bethe reason Guiochonrsquos support for this proposition has beenand continues to be a cause for much of the conventionalacceptance of Carmanrsquos (erroneous) figure of 180 as thecharacteristic value of the coefficient in the Kozeny-Blakeequation
9 Guiochonrsquos Experimental Data
In addition to the data from Halasz 1972 and 1974 studiesTable 3 herein contains data from a representative sampleof Guiochonrsquos personal experimental permeability investi-gations over approximately the last decade as reported inthe numerous trade journals particularly the Journal ofChromatography A In each case the table has a single rowdedicated to Guiochonrsquos reported raw data and a single rowdedicated to corrected values for each column based on thereference chart in Table 1 herein
12 Journal of MaterialsTa
ble3Summaryexperim
entald
ata
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Naefe
andHalasz
columns
1198601(19
72)
Colum
nA1rawdata
07140
0560
040
000314
0523
782007
2811
4016
563
498119864minus04
500
357
908119864minus10
648119864minus10
2500
020
100
000135
0001350
00038
100
053
132
074
Colum
nA1corrected
07140
0532
03800
0334
0539
7813
3518
705002
701
749119864minus04
267
191
908119864minus10
648119864minus10
2500
020
100
000110
000110
100038
100
053
139
074
Endelean
dHalasz
columns
1198602(19
74)
Colum
nA2rawdata
08426
0475
040
00044
30738
11910
1080
4740
563
110119864minus03
192
162
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
033
016
Colum
nA2corrected
08426
0511
04309
0412
0723
11910
1080
3411
405
110119864minus03
267
225
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
031
016
Guiochon
columns
1198612ndash1198614(19
98)
Colum
nB 2
rawdata
05639
0722
040
730157
0264
22771
1367
2931
520
130119864minus03
263
148
467119864minus10
263119864minus10
429
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 3
rawdata
05582
0704
03932
0165
0272
44764
1369
3382
606
131119864minus03
226
126
471119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 4
rawdata
05509
0710
03909
0160
0263
72784
1422
3422
621
128119864minus03
229
126
459119864minus10
253119864minus10
1328
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 3
corrected
05653
0723
040
860157
0265
44774
1369
2899
513
129119864minus03
267
151
465119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 4
corrected
05651
0723
040
830157
0265
72776
1374
2907
514
129119864minus03
267
151
464119864minus10
262119864minus10
1375
500
100
000
060
000
0600
00165
98008
020
015
Guiochon
columnC
(1999)
Colum
nCrawdata
05930
0673
03990
0194
0323
183
865
1459
3372
569
116119864minus03
257
152
109119864minus09
645119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Colum
nCcorrected
05930
0673
03990
0194
0323
190
900
1518
3372
569
111119864minus03
267
158
105119864minus09
620119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Guiochon
columns
1198631ndash1198636(2006)
Colum
nD
1rawdata
07830
0456
03570
0426
0663
86694
886
7115
909
144119864minus03
975
76334119864minus10
261119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
013
Colum
nD
2rawdata
07230
0491
03550
0368
0571
88656
907
6726
930
152119864minus03
975
70353119864minus10
255119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
014
Colum
nD
3rawdata
06850
0506
03466
0338
0518
97685
1000
7025
1026
146119864minus03
975
67338119864minus10
231119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD
4rawdata
06560
0521
03415
0315
0478
103
696
1062
7143
1089
144119864minus03
975
64332119864minus10
218119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD5rawdata
06190
0545
03371
0282
0425
109
692
1118
7100
1147
144119864minus03
975
60334119864minus10
207119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
016
Colum
nD
6rawdata
05760
0587
03383
0238
0359
107
635
1103
6516
1131
157119864minus03
975
56364119864minus10
210119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
017
Colum
nD
1corrected
07830
0542
04243
0359
0623
86907
1158
3397
434
110119864minus03
267
209
334119864minus10
261119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
013
Colum
nD
2corrected
07230
0584
04221
0301
0521
88857
1186
3212
444
117119864minus03
267
193
353119864minus10
255119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
014
Colum
nD
3corrected
06850
0603
04129
0272
0463
97896
1307
3354
490
112119864minus03
267
183
338119864minus10
231119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
015
Colum
nD
4corrected
06560
0621
040
730249
0420
103
911
1388
3411
520
110119864minus03
267
175
332119864minus10
218119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
015
Colum
nD
5corrected
06190
0650
04025
0217
0362
109
905
1462
3390
548
110119864minus03
267
165
334119864minus10
207119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
016
Colum
nD
6corrected
05760
0701
04038
0172
0289
107
831
1442
3111
540
120119864minus03
267
154
364119864minus10
210119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
017
Guiochon
columns
1198641-1198642
(2007)
Colum
nE 1
rawdata
064
500594
03830
0262
0425
356
1119
1735
4371
678
894119864minus04
256
165
804119864minus11
519119864minus11
1500
046
100
000
030
000
0300
000
41300
030
079
047
Colum
nE 2
rawdata
05400
0800
04320
0108
0190
470
1000
1853
2161
400
100119864minus03
463
250
729119864minus11
393119864minus11
1500
046
100
000
027
000
0270
000
41300
030
070
056
Colum
nE 1
corrected
064
500594
040
000245
040
8356
969
1502
3628
563
103119864minus03
267
172
804119864minus11
519119864minus11
1500
046
100
000
028
000
0279
000
41300
030
075
047
Colum
nE 2
corrected
05400
0800
04320
0108
0190
470
577
1068
2161
400
173119864minus03
267
144
729119864minus11
393119864minus11
1500
046
088
000
023
000
0205
000
41300
030
070
056
Guiochon
columns
1198651ndash1198654
(2008)
Colum
nF 1
rawdata
06350
0586
03720
0263
0419
197
725
1142
4865
766
138119864minus03
149
95399119864minus11
253119864minus11
300
021
100
000
017
000
0170
00035
100
048
129
076
Colum
nF 2
rawdata
064
200581
03730
0269
0429
361
807
1258
4863
758
124119864minus03
166
107
358119864minus11
230119864minus11
500
021
100
000
017
000
0170
00035
100
048
129
075
Colum
nF 3
rawdata
064
100588
03770
0264
0424
670
748
1166
4643
724
134119864minus03
161
103
387119864minus11
248119864minus11
1000
021
100
000
017
000
0170
00035
100
048
128
075
Colum
nF 4
rawdata
06390
0595
03800
0259
0418
1014
752
1177
4476
701
133119864minus03
168
107
384119864minus11
246119864minus11
1500
021
100
000
017
000
0170
00035
100
048
127
075
Colum
nF 1
corrected
06350
0630
040
000235
0392
197
954
1502
3572
563
105119864minus03
267
170
399119864minus11
253119864minus11
300
021
100
000
019
000
0195
00035
100
048
120
076
Colum
nF 2
corrected
064
200623
040
000242
0403
361
964
1502
3611
563
104119864minus03
267
171
358119864minus11
230119864minus11
500
021
100
000
019
000
0186
00035
100
048
120
075
Colum
nF 3
corrected
064
100624
040
000241
0402
670
963
1502
360
6563
104119864minus03
267
171
386119864minus11
248119864minus11
1000
021
100
000
019
000
0193
00035
100
048
120
075
Colum
nF 4
corrected
06390
0626
040
000239
0398
1014
960
1502
3594
563
104119864minus03
267
171
384119864minus11
246119864minus11
1500
021
100
000
019
000
0192
00035
100
048
120
075
Journal of Materials 13
Table3Con
tinued
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Guiochon
columns
1198671-1198672serie
s(2010)
Colum
nH
1rawdata
05320
0744
03960
0136
0225
120
563
1057
3125
587
178119864minus03
180
96122119864minus10
652119864minus11
1500
046
100
000
026
000
0262
00104
050
005
013
009
Colum
nH
2rawdata
05420
0686
03720
0170
0271
61747
1379
4152
766
134119864minus03
180
98823119864minus11
446119864minus11
1000
046
100
000
025
000
0248
00054
050
005
013
009
Colum
nH
1corrected
05320
0752
040
000132
0220
120
799
1502
2993
563
125119864minus03
267
142
122119864minus10
652119864minus11
1500
046
100
000
031
000
0313
00104
050
005
013
009
Colum
nH
2corrected
05420
0738
040
000142
0237
61814
1502
3049
563
123119864minus03
267
145
823119864minus11
446119864minus11
1000
046
100
000
026
000
0259
00054
050
005
013
009
Guiochon
columns
1198673-1198674serie
s(2010)
Colum
nH
3rawdata
06270
060
903820
0245
0396
463
700
1117
4296
685
143119864minus03
163
102
447119864minus11
280119864minus11
1500
021
100
000
018
000
0177
00036
050
024
063
038
Colum
nH
4rawdata
05170
0791
040
900108
0183
433
616
1191
2639
511
162119864minus03
233
121
580119864minus11
300119864minus11
1500
021
100
000
019
000
0189
00036
050
024
059
047
Colum
nH
3corrected
06270
0638
040
000227
0378
463
942
1502
3527
563
106119864minus03
267
167
447119864minus11
280119864minus11
1500
021
100
000
021
000
0205
00036
050
024
060
038
Colum
nH
4corrected
05170
0791
040
900108
0183
433
699
1353
2639
511
143119864minus03
265
137
580119864minus11
300119864minus11
1500
021
100
000
020
000
0201
00036
050
024
059
047
Guiochon
columns
1198681-1198686
(2010)
Colum
nI 1rawdata
06580
0562
03700
0288
0457
488
741
1127
5156
784
135119864minus03
144
95390119864minus11
256119864minus11
500
021
100
000
017
000
0170
00104
050
024
065
037
Colum
nI 2rawdata
06550
0576
03770
0278
044
6873
661
1009
4745
724
151119864minus03
139
91437119864minus11
286119864minus11
1000
021
100
000
017
000
0170
00104
050
024
064
037
Colum
nI 3rawdata
06550
0588
03850
0270
0439
1209
610
931
4341
663
164119864minus03
141
92474119864minus11
310119864minus11
1500
021
100
000
017
000
0170
00104
050
024
062
037
Colum
nI 4rawdata
06430
0572
03680
0275
0435
260
787
1225
5153
801
127119864minus03
153
98367119864minus11
236119864minus11
500
030
100
000
017
000
0170
00104
050
012
032
018
Colum
nI 5rawdata
06540
0583
03810
0273
044
144
3683
1044
4531
693
146119864minus03
151
99423119864minus11
277119864minus11
1000
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 6rawdata
06550
0585
03830
0272
044
164
6665
1016
4438
678
150119864minus03
150
98434119864minus11
285119864minus11
1500
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 1corrected
06580
060
8040
000258
0430
488
988
1502
3701
563
101119864minus03
267
176
390119864minus11
256119864minus11
500
021
100
000
020
000
0196
00104
050
024
060
037
Colum
nI 2corrected
06550
0611
040
000255
0425
873
984
1502
3684
563
102119864minus03
267
175
437119864minus11
286119864minus11
1000
021
100
000
021
000
0207
00104
050
024
060
037
Colum
nI 3corrected
06550
0611
040
000255
0425
1209
984
1502
3684
563
102119864minus03
267
175
474119864minus11
310119864minus11
1500
021
100
000
022
000
0216
00104
050
024
060
037
Colum
nI 4corrected
06430
0622
040
000243
040
5260
966
1502
3617
563
104119864minus03
267
172
367119864minus11
236119864minus11
500
030
100
000
019
000
0188
00104
050
012
029
018
Colum
nI 5corrected
06540
0612
040
000254
0423
443
982
1502
3679
563
102119864minus03
267
175
423119864minus11
277119864minus11
1000
030
100
000
020
000
0204
00104
050
012
029
018
Colum
nI 6corrected
06550
0611
040
000255
0425
646
984
1502
3684
563
102119864minus03
267
175
434119864minus11
285119864minus11
1500
030
100
000
021
000
0207
00104
050
012
029
018
Guiochon
columns
1198691-1198692
(2011)
Colum
nJ 1rawdata90
∘A(120578
=00054
P)04980
0803
040
000098
0163
65654
1313
2801
563
153119864minus03
234
116126119864minus10
627119864minus11
1500
046
100
000
029
000
0287
00054
050
005
013
010
Colum
nJ 2rawdata90
∘A(120578
=00104
P)04980
0803
040
000098
0163
124
651
1307
2801
563
154119864minus03
232
116127119864minus10
630119864minus11
1500
046
100
000
029
000
0287
00104
050
005
013
010
Colum
nJ 1rawdata160
∘A(120578
=00054
P)05630
0714
04020
0161
0269
71896
1592
3099
550
112119864minus03
289
163
102119864minus10
573119864minus11
1500
046
100
000
030
000
0302
00054
050
005
012
009
Colum
nJ 2rawdata160
∘A(120578
=00104
P)05630
0714
04020
0161
0269
129
847
1504
3099
550
118119864minus03
273
154
108119864minus10
606119864minus11
1500
046
100
000
030
000
0302
00104
050
005
012
009
Colum
nJ 1corrected
04980
0803
040
000098
0163
65748
1502
2801
563
134119864minus03
267
133
126119864minus10
627119864minus11
1500
046
100
000
031
000
0307
00054
050
005
013
010
Colum
nJ 2corrected
04980
0803
040
000098
0163
124
748
1502
2801
563
134119864minus03
267
133
127119864minus10
630119864minus11
1500
046
100
000
031
000
0308
00104
050
005
013
010
Colum
nJ 1corrected
05630
0714
04020
0161
0269
71827
1470
3099
550
121119864minus03
267
150
102119864minus10
573119864minus11
1500
046
100
000
029
000
0290
00054
050
005
012
009
Colum
nJ 2corrected
05630
0714
04020
0161
0269
129
827
1470
3099
550
121119864minus03
267
150
108119864minus10
606119864minus11
1500
046
100
000
030
000
0299
00104
050
005
012
009
14 Journal of Materials
050
100150200250300350400
045 050 055 060 065 070 075 080 085
LineAll corrected data
Linear (all corrected data)Squares raw data Triangles corrected data
Pt
120576t
D seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
J1 J2 H4
H1H2
D6
B2 B3 B4
J3 J4
E2
C
H3
D5 D4
E1
D2
A1
D1
A2
y = 267xLine slope = 267
Figure 2 Summary of experimental results
0
500
1000
1500
2000
2500
20 25 30 35 40 45 50 55 60 65 70 75 80
Linear (line)
Ψ
LineAll corrected dataD seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
Line slope (P0) = 267
J1 J2
J3 J4E2
E1
A1
A2
H4 H1
C
120601
H2H3I1 I2 I3 I4 I5
F1 F2 F3 F4
D1 D2 D3 D4 D5 D6
Squares raw data Triangles corrected data
Figure 3 Summary of experimental results
Figures 2 and 3 herein are graphical representations ofTable 31015840s raw and corrected data for Guiochonrsquos columnsplotted in the same reference frames as in Figures 1(a) and1(b) respectively
As shown in the plots of the thirty-one total columnsevaluated in this data base only six columns had reportedraw data values for 119875
0close to the normThese were columns
B2 C E
1 H
4 J
1and J
2 Interestingly of the six conforming
columns 3 of the columns were packed with fully porousparticles and 3 columns with partially porous particles Itcan be seen however that when properly interpreted andadjusted all of Guiochonrsquos experimental data falls fairlywithin the norms outlined in the reference chart in Table 1In order to explore the reasons why the raw data does notconform more closely to the norms outlined in Table 1 eachof Guiochonrsquos studies included in Table 3 is evaluated in turn
Finally Table 4 herein is a list of symbols and parameterdefinitions used throughout this paper as well as their unitsof measure
91 Guiochonrsquos 1998 Study In a 1998 publication by Kohand Guiochon entitled ldquoEffect of the Column Length on theCharacteristics of the Packed Bed and the Column Efficiencyin a Dynamic Axial Compression Columnrdquo [19] Guiochonand his coauthor report the results of a study they hadcarried out on column packing using a rather novel techniquewhich was a combination of slurry packing and mechanicalcompression
The study involved the packing of a single cylinder ofdiameter 5 cm to various column lengths ranging between 1and 25 cm approximately The particles were all silica basedand coated with a reverse stationary phase C
18 However two
different categories of particles were involved One categorycontained highly irregular particles which were about 130micrometers in average diameter In addition these particleshad a high specific pore volume (11mLg) and exhibitedsignificant particle breakage under the packing conditionsinvestigated in the study Conversely the other particlecategory involved very small particles (6 micrometers) whichwere spherical in shape had a low specific pore volume(03mLg) and exhibited a greater ability to withstandparticle breakage under the studyrsquos packing conditions
The permeability data for the irregular particles isneglected herein because according to the authors theparticles exhibited significant breakage during packing andwithout an accurate value for the characteristic dimensionof the particle the data cannot be used in the Kozeny-Blake model to determine the value of 119875
0 The spherical
particles on the other hand behaved successfully under thepacking conditions chosen and yielded values calculated bythe authors for 119875
0of 619 268 226 and 229 for the four
different lengths of columns studied The data for thesecolumns is reported in our Table 3 in the rows for ColumnsB1through B
4 The higher value of 619 (the shortest column)
was rejected by the authors because they felt that either theparticles had been somewhat broken during packing or thecolumn frits had been blocked with fines Accordingly thepermeability data for this column is likewise neglected herein
The remaining three columns however were not con-sistent with respect to the authorsrsquo calculated values for 119875
0
Since the particles were identical the range of values for 1198750
of 226 to 268 must be due to experimental error The rawdata for column B
2 which produced a 119875
0value of 268 is
all self-consistent For example it generates a value for 119875119905
of 148 which is a reasonable value based upon the knownlow particle porosity for these Nova Pak particles and acorresponding value for 120576
0of 04073 which is a reasonable
value for a well-packed column However the other twocolumns columns B
3and B
4 had the much lower value of
126 for 119875119905 Accordingly in Table 3 herein corrected values
for column external porosity for columns B3and B
4are
displayed These corrected values are only slightly differentthan the reported values are within the experimental errorof the measurement and therefore in combination with the
Journal of Materials 15
Table 4 Glossary of terms
Symbol Unit dimensions (cgs) Definition119860 cm2 Column cross-sectional area119863 cm Column diameter119889119901
cm Particle spherical diameter equivalentΔ119875 gcmminus1secminus2 Pressure drop along column119889119901119898
cm Average particle diameter (measured)1198960
None Guiochonrsquos residual reciprocal permeability coefficient119875119905
None Guiochonrsquos modified permeability coefficient in Kozeny-Blake equation119871 cm Column length1198750
None Permeability coefficient in Kozeny-Blake equation119876 cm3minminus1 (exception) Volumetric fluid flow rate1199050
sec Time for elution of totally permeating unretained solute119896 cm2 General permeability coefficient in Darcy equation119870
119865cm2 Halaszrsquos permeability coefficient in Darcy equation (1974 paper)
119870 cm2 Halaszrsquos permeability coefficient in Darcy equation (1972 paper)119870
119888None Guiochonrsquos permeability coefficient in the Kozeny-Blake equation (as modified by Carmen)
Greek1205760
None External column porosity120576119894
None Internal column porosity120576119905
None Total column porosity120576119901
None Particle porosityΦ None Giddingsrsquo ratio of external and total column porosity120601 None Flow resistance parameter based on mobile phase velocity1206010
None Flow resistance parameter based on superficial velocityΩ
119901None Particle sphericity factor
120578 gcmminus1secminus1 Fluid absolute viscosity120583119894
cmsecminus1 Average fluid interstitial linear velocity120583119904
cmsecminus1 Average fluid superficial linear velocity120583119905
cmsecminus1 Average mobile phase velocityΨ None Porosity dependence term in the Kozeny-Blake equation based on mobile phase velocityΨ
0None Porosity dependence term in the Kozeny-Blake equation based on superficial velocity
raw data for column B2 are supportive of the value of 267 for
1198750
92 Guiochonrsquos 1999 Study In a 1999 publication by Farkaset al entitled ldquoValidity of Darcyrsquos Law at Low Flow Rates inLiquid Chromatographyrdquo [20] Guiochon and his coauthorsreport the results of some very exactingmeasurements whichthey made to validate Darcyrsquos law at extremely low fluidflow rates Using a column containing particles packed toa measured external column porosity of 0399 and havinga measured total column porosity of 0593 (Figure 2 in thepaper) Guiochon et al measured the column pressure dropand fluid flow rate at flow rates ranging between 0015 and05mLminwith ethylene glycol as the fluidThe authors statethat the particles in the column are spherical and they appearto have gone to extraordinary lengths to obtain an accuratemeasure of the particle diameter and of course as a result ofmodern techniques of particle size classification the particlesize distribution is very narrow
In Figure 1 of the paper the authors show a plot ofcolumn measured pressure drop in bar versus ldquofluid linearvelocityrdquo in cmsec This plot represents their measurementstaken in the empty column The only velocity that exists in
an empty column is the superficial velocity and thus theabscissa values in the plot in Figure 1 designated as ldquofluidlinear velocityrdquo represent superficial linear velocity
In Figure 2 in their paper however the authors showanother plot of measured pressure drop in bar also versusldquofluid linear velocityrdquo in cmsec The velocity on the abscissain this plot although also designated as ldquofluid linear velocityrdquodoes not equate (for permeability related purposes) to theldquofluid linear velocityrdquo in Figure 1 This is because in Figure 2the columnwas filled with porous particles and themeasuredvelocity was the mobile phase velocity Indeed the onlyvelocity used throughout the entire paper is the mobile phasevelocity (mobile phase velocity and superficial velocity areone and the same in an empty column)
In Table 1 of their paper the authors report that for thecolumn represented in Figure 2 the slope of a line achievedwhen the mobile phase velocityin units of cmsec is plottedversus pressure drop in units of bar is 1829 They also reportthat the flow resistance parameter 120601 for this column is 865In Table 3 herein it can be seen that the raw data for thiscolumn (Column C) generates an apparent value of 257 for1198750 This value is very close to Giddingsrsquo value of 267 Indeed
by referring to Figures 2 and 3 herein one can see that thereported raw data in this study without any modification
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
2 Journal of Materials
recent literature pertaining to fully porous particles wherelow values for the constant of proportionality in the Kozeny-Blake equation are being justified on the basis of particleporosities which purportedly run counter to this principle[4] Clearly something is wrong
The value of the permeability coefficient dictated bythe necessity to balance the pressure gradientfluid flowequation after all independent variables are accounted for(and which presumably has its origin in the fundamentalmechanisms which underlie the generation of pressure in thefirst place) is by no means self-evident From the standpointof first principles it would appear that its value ought to beconstant regardless of the physical parameters comprising thefluid flow apparatus However no fully developed theoreticalexplanation for any constant of permeability has yet beenpublished On the other hand the literature is replete withdiscussions of the value for this parameter which purportedlyare justified on the basis of empirical data The problem isthat the authors of this literature do not agree with eachother and sometimes the same author disagrees with himselfIn the last few years Georges Guiochon in particular haspublished conflicting values for this parameter all of whichare supposedly based uponmeasured values yet he has madeno serious effort to reconcile the conflicts or to provide anyplausible underlying rationale for their differences
As we discuss below there are several inherent flawsin Guiochonrsquos teaching of permeability which are at theroot of his conflicting values for the coefficient We suggestthat these flaws originate in the fact that his teaching doesnot distinguish between the respective impacts on columnpermeability of (a) mass transfer by convection versus masstransfer by diffusion and (b) spherical particles versus par-ticles having an irregular morphology Accordingly withoutsubstantial refinement and adjustment Guiochonrsquos teachingof column permeability at best does little to overcome theshortcomings of Darcyrsquos law and at worst adds fuel to thefire of the aforementioned controversy
2 Guiochonrsquos Textbook Teaching
In his influential textbook Fundamentals of Preparativeand Nonlinear Chromatography [5] Guiochon (togetherwith his coauthors) bases his teaching of the pressureflowrelationship in streamline flow on the Darcy equation whichhe displays therein as his equation (526b) a rearrangementof which we repeat here
Δ119875
119871=
120583119905120578
11989601198892119901119898
(1)
where Δ119875 = pressure drop along column 119871 = length ofcolumn 120583
119905= mobile phase velocity 120578 = absolute viscosity of
fluid 119889119901119898
= average particle diameter and 1198960= Guiochonrsquos
residual permeability coefficientAll terms are self-consistent from a unit-of-measure
perspectiveNote that Guiochon does not discriminate in his expres-
sion of the Darcy equation between spherical and irregularparticles In other words in his statement of the equation
a spherical particle with a measured diameter of 119909 cm isequivalent to an irregularly shaped particle of the samemeasured diameter Peculiarly however he goes on to say thatldquotraditionally a value of 119896
0= 1 times 10minus3 is used for irregular
particles and 12 times 10minus3 for spherical onesrdquo [5 p 153]Similarly without including any porosity dependence
term in the pressurefluid flow relationship or specifyingany particular relationship between columnpressure gradientand column porosity for columns packed with either porousor nonporous particles he states that 119896
0varies ldquobetween 5 times
10minus4 and 2 times 10minus3 depending upon the compactness of thepacked bed that is on the external porosityrdquo [5 p 153]
In essence then Guiochon bundles the variables ofparticle morphology and column porosity into his residualpermeability coefficient 119896
0 Indeed this is why we call it his
residual permeability coefficientIn order to explore these implicit terms which are embed-
ded inGuiochonrsquos teaching for his 1198960 we turn for comparison
purposes to the well-known Kozeny-Blake equation
Δ119875
119871=1198750(1 minus 120576
0)2
120583119904120578
120576301198892119901
(2)
where 1198750= the KozenyBlake permeability coefficient 120576
0=
external porosity of column 120583119904= fluid superficial velocity
and 119889119901= spherical particle diameter equivalent
As we noted above Guiochon acknowledges in his teach-ing that particle irregularity does influence pressure drop Ina modification of the Kozeny-Blake equation proposed byCarman this factor was expressly accounted for by adding aparameter to represent particle sphericity which we denotewith the symbol Ω
119901(Carman gave it the symbol 120601 in his
original publication [6]) It will be appreciated that the valueof Ω
119901is equal to 1 for spherical particles and is less than 1
for particles having an irregular morphology Accordinglythe Kozeny-Blake equation (as modified by Carman)1 isexpressed as
Δ119875
119871=1198750(1 minus 120576
0)2
120583119904120578
12057630(119889
119901119898Ω
119901)2 (3)
where Ω119901= Carmanrsquos particle adjustment factor for devia-
tions from perfect sphericity and 119889119901119898Ω
119901= 119889
119901
As for column porosity note that the Kozeny-Blakeequation includes a combination of porosity terms whichtaken together can be viewed as its porosity dependenceparameter which we notate with the symbol Ψ
119900
Ψ0=(1 minus 120576
0)2
12057630
(4)
where 1205760= the volume of free space between the particles in
the columnthe volume of the empty columnComparing (1) and (3) the reader will observe that
there still is another difference the fluid velocity frames aredifferent Accordingly in order to establish the relationshipbetween the values of Kozeny-Blakersquos 119875
0and the value
of Guiochonrsquos 1198960 one needs to delineate the impact of
Journal of Materials 3
Guiochonrsquos velocity frame upon his permeability coefficientBefore attempting to do this a review of the various velocityframes typically used to characterize flow in porous mediamay be helpful
3 Fluid Velocity Frames
31 Superficial Velocity The fluid velocity term in both theDarcy equation and the Kozeny-Blake equation is the volu-metric flow rate of the fluid divided by the cross-sectional areaof the empty column This velocity is known as the ldquoaveragefluid superficial linear velocityrdquo or simply the ldquosuperficialvelocityrdquo and is defined as follows
120583119904=119876
A (5)
where 119876 = volumetric flow rate 119860 = cross-sectional area ofthe column and 120583
119904= superficial velocity
Thus superficial velocity is not influenced in any way bythe existence or structure of the packed bed Furthermore itdoes not change depending upon whether the particles in thebed are porous or nonporous
32 Interstitial Velocity Theconvective fluid velocity betweenthe particles of a packed bed is called the ldquoaverage fluidinterstitial linear velocityrdquo or simply the ldquointerstitial velocityrdquoInterstitial velocity is the superficial velocity adjusted for thecross-section of the actual fluid flow in a packed columnIt is related to the superficial velocity through the externalporosityof the column 120576
0 as follows
120583119894=120583119904
1205760
(6)
where 120583119894= average fluid interstitial linear velocity
If interstitial velocity is to be substituted for superficialvelocity in the Kozeny-Blake equation one must also includethe external porosity term This term is the compensationfactor that maintains the integrity of the value of the per-meability coefficient 119875
0between pressure gradient and fluid
superficial velocity The equation then becomes
Δ119875
119871=1198750(1 minus 120576
0)2
120583119894120578
120576201198892119901
(7)
33 Mobile Phase Velocity Guiochon does not use either thesuperficial velocity or the interstitial velocity in the equationsin his textbook Instead he uses the ldquomobile phase velocityrdquoMobile phase velocity is uniquely a chromatographic term Itis the velocity calculated as a result of injecting an unretainedfully permeating solute into the column and measuring thetime ofits elution It is defined as follows
120583119905=119871
1199050
(8)
where 1199050= the time it takes an unretained fully permeating
solute to traverse the column and 120583119905=mobile phase velocity
The velocity experienced by the flowing fluid in a columnunder steady state conditions is not altered by the existence ofldquoblindrdquo pores in the particles of that column Blind pores arepores within the particles which have only one fluid openingthat is they are not through-pores Accordingly under steadystate conditions the pores are already filled with identicalstagnant fluid and there exists no driving force to exchangethis stagnant fluid with the fluid flowing in the interstices ofthe particles It will be appreciated that column permeabilitydata is usually taken under steady state conditions in orderto take advantage of a constant value for fluid viscosityAlthough a solute which is dissolved in the flowing fluid willpenetrate the stagnant fluid in the pores (assuming it is notstrictly excluded by either physical dimension or chemicalinteraction) because of molecular diffusion the driving forcefor this solute migration is the concentration gradient of thesolute between the fluid outside the particles and the fluidinside the particles This intraparticle migration of the solutedoes not contribute to friction and thus must not be includedin velocity considerations related to pressure gradientMobilephase velocity therefore is not strictly a fluid velocity terminstead it is a misnomer and it would better be termed theldquounretained fully permeating solute velocityrdquo
Although Guiochon defines the relationship betweenfluid flow and pressure gradient in the context of a residualpermeability coefficient 119896
0 and acknowledges that column
porosity plays a role in that relationship his use of mobilephase velocity fails to recognize that molecular diffusion inblind pores of the particle fraction of a column packed withporous particles does not influence pressure gradient
Mobile phase velocity is related to fluid superficial veloc-ity however through the total porosity of the column ldquoTotalcolumn porosityrdquo 120576
119905 is a term peculiar to the chromato-
graphic literature which relates to the use of columns packedwith porous particles It is the fraction of the column volumetaken up by the free space between the particles plus thefraction of the column volume taken up by the free space inthe internal pores of the particles Thus it is the sum of theexternal and the internal column porosities
120576119905= 120576
0+ 120576
119894 (9)
where 120576119905= total column porosity and 120576
119894= internal column
porosityMobile phase velocity is related to fluid superficial veloc-
ity through the total porosity of the column as follows
120583119905120576119905= 120583
119904 (10)
Thus if mobile phase velocity is used in the Kozeny-Blakeequation one must insert the compensation factor 120576
119905into the
equation in order to maintain the integrity of the value of thepermeability coefficient 119875
0 between pressure gradient and
fluid superficial velocity and the equation then becomes
Δ119875
119871=1198750(1 minus 120576
0)2
120583119905120576119905120578
120576301198892119901
(11)
4 Journal of Materials
And now reconciling (1) and (3) we get
1198750(1 minus 120576
0)2
120583119904
120576301198892119901
=120583119905120576119905
1198960(119889
119901119898Ω
119901)2 (12)
4 The Flow Resistance Parameter
Before proceeding any further we digress to define anotherterm which will become important to our understanding ofGuiochonrsquos work The chromatographic literature containsa term called the ldquoflow resistance parameterrdquo [7] The fluidvelocity underlying this parameter is based upon Giddingsrsquodefinition of ldquomean fluid velocity through a cross-sectionrdquoand is found in his 1965 text book at page 207 [8]2 It is definedas follows
120601 =Δ119875119889
2
119901
120583119905120578119871
(13)
where 120601 = flow resistance parameterThe flow resistance parameter therefore is equivalent to
the reciprocal of Guiochonrsquos term 1198960
120601 =1
1198960
(14)
5 Refining the Terms For Column Porosity
In order to continue with our comparison of Guiochonrsquosteaching to that embodied in the Kozeny-Blake equation andin particular to accommodate columns packed with porousparticles we need to refine the porosity dependence term inthe Kozeny-Blake equation To accomplish this we adopt thedefinition of column porosity as taught by J Calvin Giddings[8] namely
Ψ =(1 minus 120576
119905Φ)
2
(120576119905Φ)
3 (15)
where Ψ = the porosity dependence term in Kozeny-Blake(porous particles) and
Φ =1205760
120576119905
(16)
where Φ = the ratio between the external and total columnporosities
Using thismethodology in themost general case involvingporous particles to express total column porosity (where totalcolumn porosity is not the same as external column porosity)as opposed to in the special case involving nonporous particles(where total column porosity and external column porosityare the same) the Kozeny-Blake equation (using mobilephase velocity as in (10)) may now be reexpressed as follows
Δ119875
119871=1198750(1 minus 120576
119905Φ)
2
120583119905120576119905120578
(120576119905Φ)
3
(119889119901)2
(17)
It will be appreciated that in the special case of columnspacked with nonporous particles (17) is identical to (3)because here 120576
119905= 120576
0and thus 120583
119905= 120583
119894= 120583
119904120576
0and Φ = 1
6 The Relationship between 1198960
and 1198750
We can now connect the teaching of Guiochon to thatembodied in the Kozeny-Blake equation as follows
1198750=
1
1198960Ψ
(18)
or
1198750=120601
Ψ (19)
As illustrated by (19) the term 1198750in the Kozeny-Blake
equation is the flow resistance parameter normalized for thecolumn porosity dependence term
Note that if1198750is a true constant the two parameters upon
which it depends 120601 and Ψ will vary in direct proportion toeach other This is illustrated by the data in Table 1 hereinwhich is a reference chart where particle porosity 120576
119901 is varied
from nonporous at one extreme to highly porous at the otherwhere 120576
119901= volume of the free space within the particles in
a packed columnthe total volume of all the particles in thecolumn
Figure 1(a) is a plot of the data contained in Table 1 as120601 versus Ψ Proceeding from the lower left hand corner tothe upper right hand corner of the plot particle porosityincreases It is obvious from the plot that all the data in Table 1falls on a single line with a slope of 267 which we say is theconstant value of 119875
0 As is illustrated by the plot in the range
of porosity covered by the data the values for 120601 vary fromabout 500 to about 1900 and correspondingly the values ofΨvary from approximately 2 to approximately 7 Note also thatthe upper boundary value of Ψ for packed columns is about55 with larger values related to capillaries
Finally we point out that once the value of 1198750is correctly
identified ameasured value for the flow resistance parametermay then be used to calculate the value of the externalporosity of the column 120576
0 when the latterrsquos value has not or
cannot be measured
7 Guiochonrsquos Modified PermeabilityCoefficient 119875
119905
As reflected in some of his recent publications which wewill explore in more depth later Guiochon inadvertentlyapparently confuses the term 119875
0with another parameter 119875
119905
which we identify here and define for the first time 119875119905is not
the same as 1198750because it is based upon Guiochonrsquos unique
(from a permeability perspective) teaching for fluid velocityIt is defined as follows
119875119905= 119875
0120576119905 (20)
where 119875119905= Guiochonrsquos modified permeability coefficient
It will be appreciated that Guiochonrsquos 119875119905 if used in the
Kozeny-Blake equation in place of 1198750 accommodates his use
of themobile phase velocity in away that is different from butequivalent to a direct conversion of mobile phase velocity tosuperficial velocity
Journal of Materials 5
Table1Colum
nperm
eabilityreferencetable
119870119865
Particlepo
rosity
rang
e120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119896119871
119863Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
120576119905Φ
(120576119905minus1205760)
120576119894
Δ119875119889119901
2Δ119875119889119901
2(1minus120576119905Φ)2120576119905
(1minus1205760)2
1120601
1198750120576119905
119889119901
2120576119905
(119889119901119898Ω119901)
(1minus1205760)
120583119905120578119871
120583119904120578119871
(120576119905Φ)3
(1205760)3
120601Ψ
120601po
iseUnits
Non
eNon
eNon
eNon
eNon
epsi
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
cmNon
ecm
cmgcmminus1 secminus1
cm3 m
inminus1
cmsecminus
1cm
secminus
1cm
secminus
1
Non
porous
particles
0380
1000
0380
000
0000
0160
711
1870
2662
701
141119864minus03
267
101
535119864minus10
15046
100
000100
000100
00100
038
0038
0100
0100
0390
1000
0390
000
0000
0147
653
1675
244
6627
153119864minus03
267
104
597119864minus10
15046
100
000100
000100
00100
039
0039
0100
0100
040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
100
000100
000100
00100
040
004
00100
0100
0410
1000
0410
000
0000
0124
553
1349
2071
505
181119864minus03
267
109
742119864minus10
15046
100
000100
000100
00100
041
004
10100
0100
0420
1000
0420
000
0000
0115
509
1212
1907
454
196119864minus03
267
112
825119864minus10
15046
100
000100
000100
00100
042
004
20100
0100
Lowpo
rosity
particles
0507
0750
0380
0127
0204
213
948
1870
3549
701
106119864minus03
267
135
535119864minus10
15046
100
000100
000100
00100
051
0051
0133
0100
0520
0750
0390
0130
0213
196
871
1675
3261
627
115119864minus03
267
139
597119864minus10
15046
100
000100
000100
00100
052
0052
0133
0100
0533
0750
040
00133
0222
180
801
1502
2999
563
125119864minus03
267
142
666119864minus10
15046
100
000100
000100
00100
053
0053
0133
0100
0546
0750
0410
0136
0231
166
737
1349
2760
505
136119864minus03
267
146
742119864minus10
15046
100
000100
000100
00100
055
0055
0133
0100
0560
0750
0420
0140
0241
153
679
1212
2542
454
147119864minus03
267
149
825119864minus10
15046
100
000100
000100
00100
056
0056
0133
0100
Medium
porosity
particles
0613
0620
0380
0233
0376
258
1146
1870
4293
701
872119864minus04
267
164
535119864minus10
15046
100
000100
000100
00100
061
0061
0161
0100
0629
0620
0390
0239
0392
237
1053
1675
3946
627
949119864minus04
267
168
597119864minus10
15046
100
000100
000100
00100
063
0063
0161
0100
064
50620
040
00245
040
9218
969
1502
3629
563
103119864minus03
267
172
666119864minus10
15046
100
000100
000100
00100
064
0065
0161
0100
0661
0620
0410
0251
0426
201
892
1349
3340
505
112119864minus03
267
177
742119864minus10
15046
100
000100
000100
00100
066
006
60161
0100
0677
0620
0420
0257
044
4185
821
1212
3076
454
122119864minus03
267
181
825119864minus10
15046
100
000100
000100
00100
068
006
80161
0100
Highpo
rosity
particles
0760
0500
0380
0380
0613
320
1422
1870
5325
701
703119864minus04
267
203
535119864minus10
15046
100
000100
000100
00100
076
0076
0200
0100
0780
0500
0390
0390
0639
294
1306
1675
4891
627
766119864minus04
267
208
597119864minus10
15046
100
000100
000100
00100
078
0078
0200
0100
0800
0500
040
0040
0066
7270
1202
1502
4500
563
832119864minus04
267
214
666119864minus10
15046
100
000100
000100
00100
080
0080
0200
0100
0820
0500
0410
0410
0694
249
1105
1349
4140
505
905119864minus04
267
219
742119864minus10
15046
100
000100
000100
00100
082
0082
0200
0100
0840
0500
0420
0420
0724
229
1018
1212
3814
454
982119864minus04
267
224
825119864minus10
15046
100
000100
000100
00100
084
0084
0200
0100
Capillary
1000
0380
0380
0620
1000
421
1870
1870
7005
701
535119864minus04
267
267
535119864minus10
15046
100
000100
000100
00100
100
0100
0263
0100
1000
0390
0390
0610
1000
377
1675
1675
6273
627
597119864minus04
267
267
597119864minus10
15046
100
000100
000100
00100
100
0100
0256
0100
1000
040
0040
0060
010
00338
1502
1502
5625
563
666119864minus04
267
267
666119864minus10
15046
100
000100
000100
00100
100
0100
0250
0100
1000
0410
0410
0590
1000
303
1349
1349
5051
505
742119864minus04
267
267
742119864minus10
15046
100
000100
000100
00100
100
0100
0244
0100
1000
0420
0420
0580
1000
273
1212
1212
4541
454
825119864minus04
267
267
825119864minus10
15046
100
000100
000100
00100
100
0100
0238
0100
6 Journal of Materials
0200400600800
100012001400160018002000
20 30 40 50 60 70
NonporousLow porosityMedium porosity
CapillaryLinear (line)
Nonporous Low porosity particles
Medium porosity particles
High porosity particles
Capillary
Line slope (P0) = 267
y = 267x + 0
Ψ
120601
Line
High porosity
(a)
99111123135147159171183195207219231243255267
035 045 055 065 075 085 095
LineCapillaryLinear (line)
Capillary
Line slope (P0) = 267
y = 267x + 0
Pt
120576t
Φ = 10
Φ = 075
Φ = 062
Φ = 05
Φ = 075
Φ = 062
Φ = 05
Φ = 1
(b)
99111123135147159171183195207219231243255267
000 010 020 030 040 050 060 070 080 090 100
NonporousLow porosityMedium porosityHigh porosity
120576p
y = 160x + 107
Pt
LineCapillaryLinear (line)
(c)
Figure 1 (a) The balance between 120601 and Ψ (b) Column permeability reference chart (c) Particle permeability reference chart
In the 2006 edition of their textbook [9] Guiochon andhis coauthors provide the following worked example of hispressure dropfluid flow teaching3
In Table 2 a cross-reference is established between Guio-chonrsquos teaching for columns packed with both porous andnonporous particles the Kozeny-Blake equation (asmodifiedby Carman) and for validation purposes Giddingsrsquo teachingbased upon actual permeabilitymeasurements fromhis Table53-1 in his 1965 text [8]4
In his worked example Guiochonrsquos fluid velocity 119906is found in his text as his equation (256) [9] at page61 and equates to mobile phase not superficial velocityAccordingly when applied to columns packed with porousparticles his worked example combines the contributionto fluid velocity of mass transfer by molecular diffusionin the free space within the particles with mass transferby convection in the free space between the particles Asdiscussed above if one is attempting to compare Guiochonrsquos
teaching for the value of the permeability coefficient with thatof the Kozeny-Blake equation the conversion factor is thetotal porosity of the column
In addition because Guiochonrsquos particle sizeparameter has an embedded assumption due to particleshaperoughness his worked example also commingles thecontribution to spherical particle diameter equivalent ofparticle shaperoughness with other nonparticle variablesUnfortunately his worked example is not based uponcompletely spherical particles where this would not be anissue We know this because he states that 119896
0in his worked
example equals 1 times 10minus3 which he says is the value of thepermeability coefficient for columns packed with irregularparticles
In order to move forward one must simplify If this wereto be done experimentally it would best be accomplishedby experiments with nonporous particles in which case thevariable of molecular diffusion in the pores of the particles
Journal of Materials 7
Table2Cr
oss-referencefor
term
s119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119896119871
119863Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
120576119905Φ
(120576119905minus1205760)
120576119894
Δ119875119889119901
2Δ119875119889119901
2(1minus120576119905Φ)2120576119905(1minus1205760)2
1120601
1198750120576119905
119889119901
2120576119905
(119889119901119898Ω119901)
(1minus1205760)
120583119905120578119871
120583119904120578119871
(120576119905Φ)3
(1205760)3
120601Ψ
120601po
iseUnits
Non
eNon
eNon
eNon
eNon
epsi
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
cmNon
ecm
cmgcmminus1 secminus1cm
3minminus1cm
secminus
1cm
secminus
1cm
secminus
1
Nonporousp
articles
Gidding
srsquotradition
alno
nporou
scolum
n040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
1000010
00010
0010
0399
004
00100
0100
Guiocho
nrsquostradition
alspheric
alparticle(app
arent)
0362
1000
0362
000
0000
0187
830
2293
3106
858
120119864minus03
267
97436119864minus10
15046
1000010
00010
0010
0361
0036
0100
0100
Gidd
ingrsquos
table5
3-1
5060meshglassb
eads
0422
1000
0422
000
0000
0113
500
1184
1873
444
200119864minus03
267
113
844119864minus10
15046
1000010
00010
0010
0421
004
20100
0100
5060meshglassb
eads
040
910
00040
9000
0000
0126
560
1371
2097
513
179119864minus03
267
109
729119864minus10
15046
1000010
00010
0010
040
70041
0100
0100
Guiochonrsquosteaching
ExA
irregular
particle
(app
arent)
040
010
00040
0000
0000
0225
1000
2500
2250
563
100119864minus03
444
178
400119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
ExA
correctedforp
article
irregularity
040
010
00040
0000
0000
0225
601
1502
2250
563
166119864minus03
267
107
400119864minus10
15046
078
00010
000
080010
0399
004
00100
0100
ExB
spheric
alparticle
(app
arent)
040
010
00040
0000
0000
0187
830
2075
2250
563
120119864minus03
369
148
482119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
ExB
corrected
040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
Porous
particles
Gidding
srsquotradition
alpo
rous
column
060
0066
7040
00200
0333
203
900
1500
3375
563
111119864minus03
267
160
667119864minus10
15046
100
00010
00010
0010
0599
006
00150
0100
Gidd
ingrsquos
table5
3-1
3040meshalum
ina
0803
0498
040
00403
0672
271
1204
1499
4510
562
830119864minus04
267
214
667119864minus10
15046
100
00010
00010
0010
0801
0080
0201
0100
5060meshalum
ina
0837
0499
0417
0420
0721
235
1043
1246
3906
467
959119864minus04
267
224
803119864minus10
15046
100
00010
00010
0010
0835
0084
0201
0100
6080meshchromosorbW
(5DNP)
0766
0498
0382
0384
0621
316
1404
1833
5259
687
712119864minus04
267
204
545119864minus10
15046
100
00010
00010
0010
0764
0077
0201
0100
6080meshchromosorbW
(20
DNP)
0785
0495
0389
0396
064
8300
1333
1698
4991
636
750119864minus04
267
210
589119864minus10
15046
100
00010
00010
0010
0783
0079
0202
0100
8 Journal of Materials
would be eliminated and by using only smooth sphericalparticles such as glass beads in which case the variableof particle shaperoughness would be eliminated If both ofthese things were done one would be able to directly andunambiguously determine the true value of 119875
05
Accordingly since his textbook contains a general teach-ing without any expressed restrictions with respect to particleporosity let us first assume that the column in Guiochonrsquosworked example is packed with nonporous particles Sec-ondly let us assume that the value for the external porosityof the column (which is the same as the total porosity ina column packed with nonporous particles) is the valuetraditionally assumed for a typical well-packed column thatis 04
As illustrated by example A in Table 2 this would leadto an apparent value for 119875
0of 444 which is obviously much
too high6 Using the adjusted particle size of 8 micrometers(rounded) which is calculated based on a value of 078 forΩ
119901
we derive values of 267 for 1198750and 107 for 119875
119905(the relationship
between these two values is of course equal to the totalporosity of the column 120576
119905 as is dictated by (20))
Similarly as illustrated by example B in Table 2 when thecolumn of Guiochonrsquos worked example is adapted to accom-modate spherical nonporous particles Guiochonrsquos teachingof the value for 119896
0of 12 times 10minus3 results in an apparent value
of 369 for 1198750
7 again too high [10 11] On the other hand ifwe correct the value of 119875
0 we get a column external porosity
of 03620which is too low for a typical well-packed column(120576
0= 04) [8] Moreover in the case of both corrections
Guiochonrsquos teaching overstates the pressure gradient (187 psicompared to the 135 psi which Giddingsrsquo experiments wouldgenerate) Accordingly it is only when we correct for thepressure drop that we achieve appropriate values of 04 for1205760 267 for 119875
0 and 107 for 119875
119905
As is evident from Table 2 when the porosity of columnspacked with nonporous particles varies the value of 119875
119905also
varies because it is based upon the measurement of mobilephase velocity which in turn changeswith the porosity of thecolumn Indeed for typical columns packed with nonporousparticles having external porosities close to 04 (038ndash042)[8] Table 2 reveals that the value of 119875
119905varies by as much as 6
points while the value of 1198750remains constant at 267
Having established a range of values for 119875119905based upon
columns packed with nonporous particles one can nowproceed to evaluate the range of values for 119875
119905based upon
columns packed with porous particles which is to say thatwe can now evaluate the impact of particle porosity onpermeability Among other things Table 2 shows that whenit comes to porous particles Guiochonrsquos 119875
119905is even more
variable (and thus even less useful) because the value ofthe coefficient can differ on a column-to-column basis byas much as 90 points depending on the combination of theinternal andexternal column porosities at play
Figures 1(a) and 1(b) are a plot of the data found in Table 1In this plot Guiochonrsquos modified permeability coefficient 119875
119905
is plotted against column total porosity 120576119905 Once again it will
be appreciated that (a) all data falls on a line with a slope of267 which represents the value of 119875
0and (b) the line has no
intercept which means that when 120576119905is zero 119875
119905is also zero As
shown in the plot values of the parameter Φ are highest atlow values of 119875
119905and lowest at high values of 119875
119905 Further as is
evidenced by the respective definitions for the terms119875119905and119875
0
contained herein the only time that the value of Guiochonrsquos119875119905does not differ from the value of 119875
0(267) is when the value
of 120576119905is 1 its value in a theoretical packed column devoid of
particles that is acapillaryFinally Figure 1(c) is another plot of the data in Table 1
but this time the values for 119875119905are plotted against particle
porosity 120576119901 The equation of this line does have an intercept
which represents the value of 119875119905when the particles are
nonporous It is also obvious from the equation of this linethat it is only when particle porosity is unity that is in acapillary that the value of 119875
119905is 267
It will be appreciated that since 120576119905is the sumof the internal
and external porosities of a packed column any given valuefor 119875
119905involving a column packed with porous particles can
represent many different combinations of these parametersVice versa any given value for 120576
119905can generate many different
values for 119875119905depending on the amount of particle porosity
which is present On the other hand as is pointed outearlier column internal porosity andor particle porosityplay no role in determining the permeability of a packedcolumn the only porosity parameter which is relevant forsuch purposes is column external porosity In other words asa permeability coefficient 119875
119905is not a particularly informative
concept because it still fails to provide any meaningful tie tothe underlying physical phenomena that govern the pressuredropfluid flow relationship
In the case of permeability measurements one quan-tifies only total pressure drop Pressure drop however isa commingling of the contributions of the several differ-ent parameters embodied in the pressureflow relationshipTherefore one must accurately represent the contributionof each factor before adding them together to equal thepressure drop Obviously if one mistakes the contribution ofany given element for instance particle size this will skewthe contribution of some other element when summed forinstance column porosity
Our frame of reference has been calibratedwith respect tothe interrelationships between the values of 119875
0 119875
119905 120576
119905 and 120576
119901
based on the data reported in Giddingsrsquo table 53-1 in his 1965textbook Since his data base was derived from experimentsusing nonporous particles the permeability results of whichwere extended to columns packed with fully porous particlesby using his Φ parameter the validity of which in turn wasestablished by independent means there is no uncertaintywith respect to the critical variable of external porosity 120576
0
inherent in our frame of reference It is for this reasontherefore that our frame of reference trumps any reporteddata which uses any other technique including the techniqueof inverse size exclusion chromatography by itself to establishvalues for 120576
0[12] In addition our frame of reference is
internally consistent due to the grounding of the parametersat the boundary condition of a theoretical capillary at whichpoint the values of 120576
119905and 120576
119901are unity and the values of 119875
0and
119875119905are equal In essence these interrelationships exemplify the
conservation laws of nature which are the ultimate standard
Journal of Materials 9
of measure when considering partial fractions of a singleentity
Consequently as Figures 1(b) and 1(c) illustrate if oneknows based upon other independent means the value of acolumnrsquos total porosity or its particlesrsquo porosity respectivelyone can at least verify whether results of permeability exper-iments expressed in terms of 119875
119905are reasonable Or to put
it another way if the 119875119905results reported do not conform to
what these plots tell us one should expect from columns andparticles based upon independently derived porosity valuesfor the particles under study we know that something iswrong Accordingly this is the way in which we use 119875
119905below
to analyze and evaluate Guiochonrsquos experimental work of thelast decade
8 The Origin of Guiochonrsquos Value forColumn Permeability
In 1966 Guiochon wrote a review paper entitled ldquoProblemsRaised by the Operation of Gas Chromatographic Columnsrdquo[13] In that paper he reviewed the development of the per-meability equation from Darcy to the then present Amongother things he reported that ldquoa mean 120576 of 038 is foundfor a number of packings obtained from powdered prod-ucts of various origins consisting of spherical or irregularparticles Almost all the measured values are between035 and 045rdquo (p 14) He then went on to discuss what hecalled the ldquosemiempirical Blake-Kozeny equationrdquo which hereported as 119896 = [119889
2
1199011205762][150(1 minus 120576)
2] [13 p 15 Equation
(11)] What sticks out of course is the fact that Guiochonstates here that 119875
0equals 150 the value for the permeability
coefficient championed by Ergun not the value posited byCarman To be fair we should point out that Guiochongave warning that he was uncomfortable with any particularvalue for the permeability coefficient For even though herecited the Kozeny-Blake equation with a value of 150 for itspermeability coefficient he stated that ldquothe values of 119896 givenby (this) equation are not very accurate and the deviationscan be as large as 50rdquo [13 p 15] Guiochon attributedmost ofthis uncertainty to the inability of practitioners of that era toeffectively measure the value of a columnrsquos external porositySince as he pointed out ldquo(the Kozeny-Blake) equation isvery sensitive to variations in 120576rdquo he concluded that theequation is ldquoof little userdquo and that it was better practice torely upon grosser measurements of permeability [13 p 15]Thus Guiochon at this timeframe made a conscious decisionto staywithin the relatively safe but admittedlymore obscureboundaries of Darcyrsquos teaching on column permeability
In 2006 however Guiochon emerged from under theshadow of Darcy permeability when he published a reviewpaper which purported to be a comprehensive summaryof the current state of the art in chromatography [14]In light of the fact that Guiochon and his coauthors hadvirtually contemporaneously published a new version of theirtextbook one would have expected that Guiochon wouldtake this opportunity to reaffirm his textbook teaching So itshould come as no surprise that he would open his discussionof the pressure dropfluid flow relationshipwith the following
assertion ldquothe permeability of packed beds used in liquidchromatography is in the order of 1 times 10minus3rdquo [14 p 9] Asone would expect the authority which Guiochon cites forthis seemingly familiar proposition is the 2006 edition of hisown textbook For the first time however he also providessome citations to third party sources In particular he cites a1997 chromatography handbook written by Uwe Neue ChiefScientist for Waters Corporation [15 p 9] As one can seeNeuersquos equationmdashlike Guiochonrsquosmdashdoes indeed postulate apermeability coefficient with a value of 1 times 10minus3
At this point however Guiochon abandons his tradi-tional caution about the value of the permeability coeffi-cient and marries his permeability equation to the Kozeny-Blake equation thereby endorsing not only the porositydependence term in that equation but also a specific valuefor 119875
0 Referring to his value of 1 times 10minus3 he states the
following ldquothe specific column permeability is related to thecolumn porosity through the Kozeny-Carman equation 119896
119891=
12057630180 (1 minus 120576
0)2rdquo [14 p 9] Not surprisingly Neue makes
the same connection ldquothe specific permeability depends onthe particle size 119889
119901and the interstitial porosity 120576
0of the
packed bedThe relationship is known as theKozeny-Carmanequation 119861
0= (1185)(1205763
01198892119901(1 minus 120576
0)2)rdquo [15 p 30]
Neuersquos permeability equation and Guiochonrsquos textbookpermeability equation however differ in that Neue uses thesuperficial not the mobile phase velocity Accordingly sinceall other features of their equations are the same one wouldexpect them to get different values for their permeabilitycoefficients Neue himself recognizes this when he says inhis handbook ldquothis value [1000] is also in good agreementwith our experience but values as low as 500 have beenreported in the literature8 However there are several reasonswhy different values reported by different workers do notrepresent true differences in the resistance parameter Firstlower values are obtained for porous packings if the specificpermeability is measured on the basis of the breakthroughtime of an unretained peak instead of using the flow raterdquo see[15 p 32] (when he distinguishes the ldquobreakthrough time ofan unretained peakrdquo from the ldquoflow raterdquo Neue is of coursereferring to the mobile phase velocity and the superficialvelocity resp) Consequently if one takes Neuersquos point andturns it on its head one can see that if he and Guiochonwere to get the same value for their permeability coefficients[1000] but one (Guiochon) is using mobile phase velocityto derive it and the other (Neue) uses superficial velocity toderive it this would indeed represent a ldquotrue differencerdquo intheir equations
So how can Guiochon cite both Neue and himself for theproposition in his review paper that ldquothe permeability of thepacked beds used in liquid chromatography is in the order of1times 10minus3rdquoThe answer is that at least for purposes of his reviewpaper he adopts Kozeny-Blakersquos fluid velocity convention inplace of the velocity frame he uses in his textbook Here iswhat he says ldquothis approximation [1 times 10minus3] is valid only ifthe superficial velocity is usedrdquo [14 p 9] (emphasis supplied)
As a matter of pure mathematics Guiochonrsquos permeabil-ity coefficient of 1 times 10minus3 can only be equal to 1205763
0180(1 minus 120576
0)2
if the value of 1205760is 04 the typical value for the interstitial
10 Journal of Materials
fraction of a well-packed column Guiochon confirms in hisreview paper that this is indeed what he assumes in his state-ment of the Kozeny-Blake equation [14 p 9] Accordinglyhis adoption of the value of 180 for 119875
0solely depends on the
validity of his assertion that the value of the permeabilitycoefficient is 1 times 10minus3
We already know from our discussion of Guiochonrsquostextbook teaching that when this value is associated with apermeability equation based on mobile phase velocity it iswrong As reflected in our Table 2 when Guiochonrsquos workedexample involving this figure is corrected to account for theembedded factor due to the irregularity of the particles whichhis worked example assumes the value of the permeabilitycoefficient generated by his equation is not 1 times 10minus3 it is 166times 10minus3
So where did this value of 1 times 10minus3 come from AlthoughGuiochon in his review paper cites Neuersquos handbook for thisvalue Neue himself offers absolutely no authority for it Onthe other hand Guiochon does provide one other citation inhis review paper for this value The citation is to an article byHalasz et al published in 1975 [16]
If one then goes to that article one sees that Halaszet al do indeed proffer the same equation as Neue (albeitin a somewhat different format) and yes it does containthe value of 1 times 10minus3 See [16 Equation (1)] Neverthelessthe authors fail to identify from whence they obtained thisvalue On the other hand they do make reference to apaper that Endele et al wrote a year earlier in 1974 withKlaus Unger entitled ldquoInfluence of the Particle Size (5ndash35 120583)of Spherical Silica on Column Efficiencies in High-PressureLiquid Chromatographyrdquo [17] It appears that this is wherethe body is buried for it is in this article that Halasz reportshaving actually derived a value of 1 times 10minus3 for the permeabilitycoefficient
In summarizing the results of their experiments whichincidentally are based on measurements of superficial notmobile phase velocity (see their Equation (7)) Halasz et alcalculate an average value for the permeability coefficient(which they notate as119870
119865) of 1198892
1199011080 Accordingly they state
the following ldquo[I]t is proposed to define an average particlesize 119889
119901 for columns packed with spherical silica using the
balanced density packing method by the equation 1198892119901
=
1198701198651000 = 1198651205781198711031199032120587Δ119875rdquo [17 p 382 their Equation (7)]To recapitulate it thus appears that (1) Halasz was the
source of the value of 1 times 10minus3 for the permeability coefficientwhich Guiochon adopts in his review paper (2)Halaszrsquo valuewas based upon measurements which involved sphericalparticles and (3)Halaszrsquo value was properly derived from anequation which utilized the superficial velocity as the fluidvelocity
However we still have a problem if we are going to acceptthe value of 1 times 10minus3 as a basis for calculating the constantin Kozeny-Blake we still need to know the value of Halaszrsquocolumnsrsquo external porosity For example in the absenceof measured values for the external porosity Guiochonrsquosassertion in his review paper that the value of the constantis 180 is without foundation
Unfortunately however Halasz bases his methodologyon ldquoassumedrdquo versus ldquomeasuredrdquo values for external porosityNevertheless he does provide some clues that may helpus peel back this onion First he says that his value forthe permeability coefficient applies to ldquocolumns using thebalanced density packing methodrdquo Secondly after statinghis proposed value for the coefficient we find the followingtelltale sentence ldquothe permeability of columns packed by theconventional ldquodryrdquo method are worse by a factor of about 2than those packed by the balanced density methodrdquo [17 p382] For this proposition Halasz cites yet another of his ownarticles this time a paper written by himself and M Naefein 1972 entitled ldquoInfluence of Column Parameters on PeakBroadening in High-Pressure Liquid Chromatographyrdquo [18]
When one follows this thread by going to Halaszrsquos 1972paper one finds a possible solution to the puzzle For this iswhere Halasz describes his calculations of the permeabilitycoefficient from experiments with a number of conventionaldry-packed columns containing spherical silica particlesNot surprisingly the resulting value for the permeabilitycoefficient for such columns was not 1 times 10minus3 Instead Halaszsays that his experiments with such columns resulted inthe following value for the permeability coefficient ldquo119870 sim
11988921199012000rdquo [18 p 78] In other words consistent with what
Halasz says in his 1974 article the value for the coefficient forthese dry-packed columns was indeed ldquoworse by a factor ofabout 2rdquo compared to its value for columns prepared with thebalanced density method
The obvious implication of this is that when Halasz gota value of 1 times 10minus3 for the permeability coefficient in 1974he must have been evaluating columns that were packed toan interstitial fraction considerably higher than the externalporosity of the columns that were the subject of his 1972experiments Our problem however is still not resolvedbecause Halasz did not accurately determine the externalporosity of the columns under study in either the 1972 or 1974columns
We surmise that in their 1972 paper Halasz and hiscoauthors made use of the teaching of Giddings outlinedin his 1965 textbook which at the time of their 1972 paperwas seven years old Here is what they say ldquoexperiencehas shown that in a regular packed column with the sievefractions usual for chromatography 120576
0is independent of the
particle size if the particles are more or less spherical In anacceptable approximation 120576
0is equal to 04 In those columns
packed with nonporous supports 119906V and 119906 are identicalUsing porous supports (ie silica gel alumina chromosorbPorasil and so forth) 119906 = 4119865(1205871198892
119888120576119879) where 120576
119879is the
total porosity and indicates the pore volume of the supportrdquoIt will be appreciated that 119906V stands for interstitial velocity(our 120583
119894) and 119906 stands for mobile phase velocity (our 120583
119905) The
ldquoexperiencerdquo which Halasz and his coauthors are referringto concerning ldquocolumns packed with nonporous supportsalumina and chromosorbrdquo is that recorded by Giddings inTable 53-1 in his 1965 text Continuing to establish theinterrelationships inherent in the various entities involvedin chromatographic columns with respect to both interstitialandmobile phase velocity the authors conclude ldquoformally 119906V
Journal of Materials 11
can be calculated for a porous support assuming 1205760is equal
to 04rdquo (emphasis supplied) The authors then announce themethodology underlying their frame of reference with regardto column permeability as follows ldquowe determined in all ofour columns the 119906119906V ratio to be 056 with a differentialrefractometer Consequently the total porosity 120576
119905was equal
to 0714rdquo It will be appreciated that their ratio of 119906119906V is oneand the same as GiddingsrsquoΦ parameter as utilized by him inhis 1965 textbook and as we define hereinabove
We can now identify the origin of the discrepancybetween Halasz and Giddings and by extension betweenHalasz and Guiochon (and Neue) For even though Halaszmakes use of the results in Giddingsrsquo Table 53-1 he failsto appreciate the teaching inherent therein regarding theimportance of an accurate value for column external porositywhen using porous particles In establishing his values forΦGiddings was careful to ground his values for the externalporosity of columns packed with porous particles in acomparison of their measured pressure drops with measuredpressure drops in columns packed with nonporous particleswhere the values for (external) porositywere accurate becausethey are equal to the columnsrsquo total porosity
In their 1972 paper Halasz et al used a refractometer todetect the inert solute 119899-pentane in the mobile phase of 119899-heptaneThey injected the 119899-pentane into the flowing mobilephase the flow rate of which they presumably measured witha volumetric cylinder and stop watch Since the exit timeof the 119899-pentane peak was identified by the refractometertheir measurement of holdup volume was accurate andaccordingly so was their value of 0714 for 120576
119905(there is no
uncertainty related to column total porosity when using inertsolutes)This in turn led to an accurate value for their119906 termmobile phase velocity However since they did not measurethe interstitial velocity 119906V but rather used an estimate basedupon the measured flow rate and an ldquoassumedrdquo value of 04for external porosity their calculatedestimated value for 119906Vwas arbitrary This in turn means that the value of 056 fortheirΦ ratio was likewise arbitrary
In Table 3 herein we show the results for column number1 in the 1972 study which we designate as Column A
1 Note
that the authorsrsquo raw values correspond to a value of 500 for1198750 a valuewhich is obviously far too high As is plainly evident
in our Figures 2 and 3 this column data set does not conformin any reasonable fashion to the norms in our referencecharts Accordingly the data is badly flawed In our correcteddata for Column A
1in Table 3 herein we adjust the column
external porosity to the lowest value permitted in our frame ofreference (038) This correction results in a revised value forGiddingsrsquoΦparameter of 0532 In additionwe are compelledto adjust the particle downward (to 11 micrometers approx)as is also dictated by our frame of reference We justify ourlow corrected value for external porosity and our downwardadjustment for particle size on the aggressive ldquodry-packingrdquoprocess described by the authors ldquo[T] the best results wereachieved with the following method the column (id =2mm) 25 cm long was packed with 25 equal portions of theldquobrushrdquo After adding each portion the column was vibratedand afterward tapped on the floor and also vigorously tappedwith a rod from the siderdquo Based upon a column packing
experience of the principal author herein spanningmore than30 years we believe that these packing conditions generateda packed column with a low external porosity and also one inwhich the particles experienced significant breakage
The main differences between the columns in the 1972and 1974 papers concerned the particle sizes and themethodsused to pack the columns The particle diameters in the 1972study fell in a range of about 15ndash200 micrometers while thosein the 1974 study were much smaller being in the range of 5ndash40 micrometers approximately Moreover while the columnsin the 1972 study were dry-packed the columns in the 1974study were slurry-packed (what Halasz calls ldquothe balanceddensity packing methodrdquo) The authors of the 1974 paperrecite that their permeability results shown in their Table 4correspond to values for 120601
0of 1080 and values for 120601 of 910
In Table 3 herein we show the results for the 1974 study as anaverage value in our column designated as A
2 Note that the
raw data reported by the authors which again has the built-inassumption that the column external porosity was 04 placesthe value of 119875
0at 192 In our corrected values for Column A
2
we show that a value of 267 for 1198750places the column external
porosity at a value of 043 The ratio of the value for 120601 of 910when compared to the value for 120601 in the 1972 study of 2007is about 22 (2007910 = 22) which is in good agreementwith the authorsrsquo statement that the permeability of the 1972columns was worse by a factor of about 2 Again based onthe experience of the principal author herein in the packingof hundreds of thousands of commercially sold slurry packedcolumns we conclude that the balanced density techniquedescribed by the authors in the 1974 paper is fully capable ofproducing columns with this high and external porosity
In summary we conclude that for purposes of theKozeny-Blake equation (a) the external porosity of theHalasz 1972 columns was 038 (b) the external porosity of thecolumns which he tested in 1974 was 043 and (c) Halaszrsquos1972 and 1974 studies when properly analyzed corroborateGiddingsrsquo value for 119875
0of 267 not Carmanrsquos value of 180
Although the fault for all of the misdirection about the valueof the permeability coefficient being 1 times 10minus3 therefore restsprimarily on the shoulders of Halasz one has to question whyGuiochon (andNeue) so readily adopted itWhatevermay bethe reason Guiochonrsquos support for this proposition has beenand continues to be a cause for much of the conventionalacceptance of Carmanrsquos (erroneous) figure of 180 as thecharacteristic value of the coefficient in the Kozeny-Blakeequation
9 Guiochonrsquos Experimental Data
In addition to the data from Halasz 1972 and 1974 studiesTable 3 herein contains data from a representative sampleof Guiochonrsquos personal experimental permeability investi-gations over approximately the last decade as reported inthe numerous trade journals particularly the Journal ofChromatography A In each case the table has a single rowdedicated to Guiochonrsquos reported raw data and a single rowdedicated to corrected values for each column based on thereference chart in Table 1 herein
12 Journal of MaterialsTa
ble3Summaryexperim
entald
ata
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Naefe
andHalasz
columns
1198601(19
72)
Colum
nA1rawdata
07140
0560
040
000314
0523
782007
2811
4016
563
498119864minus04
500
357
908119864minus10
648119864minus10
2500
020
100
000135
0001350
00038
100
053
132
074
Colum
nA1corrected
07140
0532
03800
0334
0539
7813
3518
705002
701
749119864minus04
267
191
908119864minus10
648119864minus10
2500
020
100
000110
000110
100038
100
053
139
074
Endelean
dHalasz
columns
1198602(19
74)
Colum
nA2rawdata
08426
0475
040
00044
30738
11910
1080
4740
563
110119864minus03
192
162
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
033
016
Colum
nA2corrected
08426
0511
04309
0412
0723
11910
1080
3411
405
110119864minus03
267
225
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
031
016
Guiochon
columns
1198612ndash1198614(19
98)
Colum
nB 2
rawdata
05639
0722
040
730157
0264
22771
1367
2931
520
130119864minus03
263
148
467119864minus10
263119864minus10
429
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 3
rawdata
05582
0704
03932
0165
0272
44764
1369
3382
606
131119864minus03
226
126
471119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 4
rawdata
05509
0710
03909
0160
0263
72784
1422
3422
621
128119864minus03
229
126
459119864minus10
253119864minus10
1328
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 3
corrected
05653
0723
040
860157
0265
44774
1369
2899
513
129119864minus03
267
151
465119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 4
corrected
05651
0723
040
830157
0265
72776
1374
2907
514
129119864minus03
267
151
464119864minus10
262119864minus10
1375
500
100
000
060
000
0600
00165
98008
020
015
Guiochon
columnC
(1999)
Colum
nCrawdata
05930
0673
03990
0194
0323
183
865
1459
3372
569
116119864minus03
257
152
109119864minus09
645119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Colum
nCcorrected
05930
0673
03990
0194
0323
190
900
1518
3372
569
111119864minus03
267
158
105119864minus09
620119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Guiochon
columns
1198631ndash1198636(2006)
Colum
nD
1rawdata
07830
0456
03570
0426
0663
86694
886
7115
909
144119864minus03
975
76334119864minus10
261119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
013
Colum
nD
2rawdata
07230
0491
03550
0368
0571
88656
907
6726
930
152119864minus03
975
70353119864minus10
255119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
014
Colum
nD
3rawdata
06850
0506
03466
0338
0518
97685
1000
7025
1026
146119864minus03
975
67338119864minus10
231119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD
4rawdata
06560
0521
03415
0315
0478
103
696
1062
7143
1089
144119864minus03
975
64332119864minus10
218119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD5rawdata
06190
0545
03371
0282
0425
109
692
1118
7100
1147
144119864minus03
975
60334119864minus10
207119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
016
Colum
nD
6rawdata
05760
0587
03383
0238
0359
107
635
1103
6516
1131
157119864minus03
975
56364119864minus10
210119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
017
Colum
nD
1corrected
07830
0542
04243
0359
0623
86907
1158
3397
434
110119864minus03
267
209
334119864minus10
261119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
013
Colum
nD
2corrected
07230
0584
04221
0301
0521
88857
1186
3212
444
117119864minus03
267
193
353119864minus10
255119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
014
Colum
nD
3corrected
06850
0603
04129
0272
0463
97896
1307
3354
490
112119864minus03
267
183
338119864minus10
231119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
015
Colum
nD
4corrected
06560
0621
040
730249
0420
103
911
1388
3411
520
110119864minus03
267
175
332119864minus10
218119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
015
Colum
nD
5corrected
06190
0650
04025
0217
0362
109
905
1462
3390
548
110119864minus03
267
165
334119864minus10
207119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
016
Colum
nD
6corrected
05760
0701
04038
0172
0289
107
831
1442
3111
540
120119864minus03
267
154
364119864minus10
210119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
017
Guiochon
columns
1198641-1198642
(2007)
Colum
nE 1
rawdata
064
500594
03830
0262
0425
356
1119
1735
4371
678
894119864minus04
256
165
804119864minus11
519119864minus11
1500
046
100
000
030
000
0300
000
41300
030
079
047
Colum
nE 2
rawdata
05400
0800
04320
0108
0190
470
1000
1853
2161
400
100119864minus03
463
250
729119864minus11
393119864minus11
1500
046
100
000
027
000
0270
000
41300
030
070
056
Colum
nE 1
corrected
064
500594
040
000245
040
8356
969
1502
3628
563
103119864minus03
267
172
804119864minus11
519119864minus11
1500
046
100
000
028
000
0279
000
41300
030
075
047
Colum
nE 2
corrected
05400
0800
04320
0108
0190
470
577
1068
2161
400
173119864minus03
267
144
729119864minus11
393119864minus11
1500
046
088
000
023
000
0205
000
41300
030
070
056
Guiochon
columns
1198651ndash1198654
(2008)
Colum
nF 1
rawdata
06350
0586
03720
0263
0419
197
725
1142
4865
766
138119864minus03
149
95399119864minus11
253119864minus11
300
021
100
000
017
000
0170
00035
100
048
129
076
Colum
nF 2
rawdata
064
200581
03730
0269
0429
361
807
1258
4863
758
124119864minus03
166
107
358119864minus11
230119864minus11
500
021
100
000
017
000
0170
00035
100
048
129
075
Colum
nF 3
rawdata
064
100588
03770
0264
0424
670
748
1166
4643
724
134119864minus03
161
103
387119864minus11
248119864minus11
1000
021
100
000
017
000
0170
00035
100
048
128
075
Colum
nF 4
rawdata
06390
0595
03800
0259
0418
1014
752
1177
4476
701
133119864minus03
168
107
384119864minus11
246119864minus11
1500
021
100
000
017
000
0170
00035
100
048
127
075
Colum
nF 1
corrected
06350
0630
040
000235
0392
197
954
1502
3572
563
105119864minus03
267
170
399119864minus11
253119864minus11
300
021
100
000
019
000
0195
00035
100
048
120
076
Colum
nF 2
corrected
064
200623
040
000242
0403
361
964
1502
3611
563
104119864minus03
267
171
358119864minus11
230119864minus11
500
021
100
000
019
000
0186
00035
100
048
120
075
Colum
nF 3
corrected
064
100624
040
000241
0402
670
963
1502
360
6563
104119864minus03
267
171
386119864minus11
248119864minus11
1000
021
100
000
019
000
0193
00035
100
048
120
075
Colum
nF 4
corrected
06390
0626
040
000239
0398
1014
960
1502
3594
563
104119864minus03
267
171
384119864minus11
246119864minus11
1500
021
100
000
019
000
0192
00035
100
048
120
075
Journal of Materials 13
Table3Con
tinued
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Guiochon
columns
1198671-1198672serie
s(2010)
Colum
nH
1rawdata
05320
0744
03960
0136
0225
120
563
1057
3125
587
178119864minus03
180
96122119864minus10
652119864minus11
1500
046
100
000
026
000
0262
00104
050
005
013
009
Colum
nH
2rawdata
05420
0686
03720
0170
0271
61747
1379
4152
766
134119864minus03
180
98823119864minus11
446119864minus11
1000
046
100
000
025
000
0248
00054
050
005
013
009
Colum
nH
1corrected
05320
0752
040
000132
0220
120
799
1502
2993
563
125119864minus03
267
142
122119864minus10
652119864minus11
1500
046
100
000
031
000
0313
00104
050
005
013
009
Colum
nH
2corrected
05420
0738
040
000142
0237
61814
1502
3049
563
123119864minus03
267
145
823119864minus11
446119864minus11
1000
046
100
000
026
000
0259
00054
050
005
013
009
Guiochon
columns
1198673-1198674serie
s(2010)
Colum
nH
3rawdata
06270
060
903820
0245
0396
463
700
1117
4296
685
143119864minus03
163
102
447119864minus11
280119864minus11
1500
021
100
000
018
000
0177
00036
050
024
063
038
Colum
nH
4rawdata
05170
0791
040
900108
0183
433
616
1191
2639
511
162119864minus03
233
121
580119864minus11
300119864minus11
1500
021
100
000
019
000
0189
00036
050
024
059
047
Colum
nH
3corrected
06270
0638
040
000227
0378
463
942
1502
3527
563
106119864minus03
267
167
447119864minus11
280119864minus11
1500
021
100
000
021
000
0205
00036
050
024
060
038
Colum
nH
4corrected
05170
0791
040
900108
0183
433
699
1353
2639
511
143119864minus03
265
137
580119864minus11
300119864minus11
1500
021
100
000
020
000
0201
00036
050
024
059
047
Guiochon
columns
1198681-1198686
(2010)
Colum
nI 1rawdata
06580
0562
03700
0288
0457
488
741
1127
5156
784
135119864minus03
144
95390119864minus11
256119864minus11
500
021
100
000
017
000
0170
00104
050
024
065
037
Colum
nI 2rawdata
06550
0576
03770
0278
044
6873
661
1009
4745
724
151119864minus03
139
91437119864minus11
286119864minus11
1000
021
100
000
017
000
0170
00104
050
024
064
037
Colum
nI 3rawdata
06550
0588
03850
0270
0439
1209
610
931
4341
663
164119864minus03
141
92474119864minus11
310119864minus11
1500
021
100
000
017
000
0170
00104
050
024
062
037
Colum
nI 4rawdata
06430
0572
03680
0275
0435
260
787
1225
5153
801
127119864minus03
153
98367119864minus11
236119864minus11
500
030
100
000
017
000
0170
00104
050
012
032
018
Colum
nI 5rawdata
06540
0583
03810
0273
044
144
3683
1044
4531
693
146119864minus03
151
99423119864minus11
277119864minus11
1000
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 6rawdata
06550
0585
03830
0272
044
164
6665
1016
4438
678
150119864minus03
150
98434119864minus11
285119864minus11
1500
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 1corrected
06580
060
8040
000258
0430
488
988
1502
3701
563
101119864minus03
267
176
390119864minus11
256119864minus11
500
021
100
000
020
000
0196
00104
050
024
060
037
Colum
nI 2corrected
06550
0611
040
000255
0425
873
984
1502
3684
563
102119864minus03
267
175
437119864minus11
286119864minus11
1000
021
100
000
021
000
0207
00104
050
024
060
037
Colum
nI 3corrected
06550
0611
040
000255
0425
1209
984
1502
3684
563
102119864minus03
267
175
474119864minus11
310119864minus11
1500
021
100
000
022
000
0216
00104
050
024
060
037
Colum
nI 4corrected
06430
0622
040
000243
040
5260
966
1502
3617
563
104119864minus03
267
172
367119864minus11
236119864minus11
500
030
100
000
019
000
0188
00104
050
012
029
018
Colum
nI 5corrected
06540
0612
040
000254
0423
443
982
1502
3679
563
102119864minus03
267
175
423119864minus11
277119864minus11
1000
030
100
000
020
000
0204
00104
050
012
029
018
Colum
nI 6corrected
06550
0611
040
000255
0425
646
984
1502
3684
563
102119864minus03
267
175
434119864minus11
285119864minus11
1500
030
100
000
021
000
0207
00104
050
012
029
018
Guiochon
columns
1198691-1198692
(2011)
Colum
nJ 1rawdata90
∘A(120578
=00054
P)04980
0803
040
000098
0163
65654
1313
2801
563
153119864minus03
234
116126119864minus10
627119864minus11
1500
046
100
000
029
000
0287
00054
050
005
013
010
Colum
nJ 2rawdata90
∘A(120578
=00104
P)04980
0803
040
000098
0163
124
651
1307
2801
563
154119864minus03
232
116127119864minus10
630119864minus11
1500
046
100
000
029
000
0287
00104
050
005
013
010
Colum
nJ 1rawdata160
∘A(120578
=00054
P)05630
0714
04020
0161
0269
71896
1592
3099
550
112119864minus03
289
163
102119864minus10
573119864minus11
1500
046
100
000
030
000
0302
00054
050
005
012
009
Colum
nJ 2rawdata160
∘A(120578
=00104
P)05630
0714
04020
0161
0269
129
847
1504
3099
550
118119864minus03
273
154
108119864minus10
606119864minus11
1500
046
100
000
030
000
0302
00104
050
005
012
009
Colum
nJ 1corrected
04980
0803
040
000098
0163
65748
1502
2801
563
134119864minus03
267
133
126119864minus10
627119864minus11
1500
046
100
000
031
000
0307
00054
050
005
013
010
Colum
nJ 2corrected
04980
0803
040
000098
0163
124
748
1502
2801
563
134119864minus03
267
133
127119864minus10
630119864minus11
1500
046
100
000
031
000
0308
00104
050
005
013
010
Colum
nJ 1corrected
05630
0714
04020
0161
0269
71827
1470
3099
550
121119864minus03
267
150
102119864minus10
573119864minus11
1500
046
100
000
029
000
0290
00054
050
005
012
009
Colum
nJ 2corrected
05630
0714
04020
0161
0269
129
827
1470
3099
550
121119864minus03
267
150
108119864minus10
606119864minus11
1500
046
100
000
030
000
0299
00104
050
005
012
009
14 Journal of Materials
050
100150200250300350400
045 050 055 060 065 070 075 080 085
LineAll corrected data
Linear (all corrected data)Squares raw data Triangles corrected data
Pt
120576t
D seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
J1 J2 H4
H1H2
D6
B2 B3 B4
J3 J4
E2
C
H3
D5 D4
E1
D2
A1
D1
A2
y = 267xLine slope = 267
Figure 2 Summary of experimental results
0
500
1000
1500
2000
2500
20 25 30 35 40 45 50 55 60 65 70 75 80
Linear (line)
Ψ
LineAll corrected dataD seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
Line slope (P0) = 267
J1 J2
J3 J4E2
E1
A1
A2
H4 H1
C
120601
H2H3I1 I2 I3 I4 I5
F1 F2 F3 F4
D1 D2 D3 D4 D5 D6
Squares raw data Triangles corrected data
Figure 3 Summary of experimental results
Figures 2 and 3 herein are graphical representations ofTable 31015840s raw and corrected data for Guiochonrsquos columnsplotted in the same reference frames as in Figures 1(a) and1(b) respectively
As shown in the plots of the thirty-one total columnsevaluated in this data base only six columns had reportedraw data values for 119875
0close to the normThese were columns
B2 C E
1 H
4 J
1and J
2 Interestingly of the six conforming
columns 3 of the columns were packed with fully porousparticles and 3 columns with partially porous particles Itcan be seen however that when properly interpreted andadjusted all of Guiochonrsquos experimental data falls fairlywithin the norms outlined in the reference chart in Table 1In order to explore the reasons why the raw data does notconform more closely to the norms outlined in Table 1 eachof Guiochonrsquos studies included in Table 3 is evaluated in turn
Finally Table 4 herein is a list of symbols and parameterdefinitions used throughout this paper as well as their unitsof measure
91 Guiochonrsquos 1998 Study In a 1998 publication by Kohand Guiochon entitled ldquoEffect of the Column Length on theCharacteristics of the Packed Bed and the Column Efficiencyin a Dynamic Axial Compression Columnrdquo [19] Guiochonand his coauthor report the results of a study they hadcarried out on column packing using a rather novel techniquewhich was a combination of slurry packing and mechanicalcompression
The study involved the packing of a single cylinder ofdiameter 5 cm to various column lengths ranging between 1and 25 cm approximately The particles were all silica basedand coated with a reverse stationary phase C
18 However two
different categories of particles were involved One categorycontained highly irregular particles which were about 130micrometers in average diameter In addition these particleshad a high specific pore volume (11mLg) and exhibitedsignificant particle breakage under the packing conditionsinvestigated in the study Conversely the other particlecategory involved very small particles (6 micrometers) whichwere spherical in shape had a low specific pore volume(03mLg) and exhibited a greater ability to withstandparticle breakage under the studyrsquos packing conditions
The permeability data for the irregular particles isneglected herein because according to the authors theparticles exhibited significant breakage during packing andwithout an accurate value for the characteristic dimensionof the particle the data cannot be used in the Kozeny-Blake model to determine the value of 119875
0 The spherical
particles on the other hand behaved successfully under thepacking conditions chosen and yielded values calculated bythe authors for 119875
0of 619 268 226 and 229 for the four
different lengths of columns studied The data for thesecolumns is reported in our Table 3 in the rows for ColumnsB1through B
4 The higher value of 619 (the shortest column)
was rejected by the authors because they felt that either theparticles had been somewhat broken during packing or thecolumn frits had been blocked with fines Accordingly thepermeability data for this column is likewise neglected herein
The remaining three columns however were not con-sistent with respect to the authorsrsquo calculated values for 119875
0
Since the particles were identical the range of values for 1198750
of 226 to 268 must be due to experimental error The rawdata for column B
2 which produced a 119875
0value of 268 is
all self-consistent For example it generates a value for 119875119905
of 148 which is a reasonable value based upon the knownlow particle porosity for these Nova Pak particles and acorresponding value for 120576
0of 04073 which is a reasonable
value for a well-packed column However the other twocolumns columns B
3and B
4 had the much lower value of
126 for 119875119905 Accordingly in Table 3 herein corrected values
for column external porosity for columns B3and B
4are
displayed These corrected values are only slightly differentthan the reported values are within the experimental errorof the measurement and therefore in combination with the
Journal of Materials 15
Table 4 Glossary of terms
Symbol Unit dimensions (cgs) Definition119860 cm2 Column cross-sectional area119863 cm Column diameter119889119901
cm Particle spherical diameter equivalentΔ119875 gcmminus1secminus2 Pressure drop along column119889119901119898
cm Average particle diameter (measured)1198960
None Guiochonrsquos residual reciprocal permeability coefficient119875119905
None Guiochonrsquos modified permeability coefficient in Kozeny-Blake equation119871 cm Column length1198750
None Permeability coefficient in Kozeny-Blake equation119876 cm3minminus1 (exception) Volumetric fluid flow rate1199050
sec Time for elution of totally permeating unretained solute119896 cm2 General permeability coefficient in Darcy equation119870
119865cm2 Halaszrsquos permeability coefficient in Darcy equation (1974 paper)
119870 cm2 Halaszrsquos permeability coefficient in Darcy equation (1972 paper)119870
119888None Guiochonrsquos permeability coefficient in the Kozeny-Blake equation (as modified by Carmen)
Greek1205760
None External column porosity120576119894
None Internal column porosity120576119905
None Total column porosity120576119901
None Particle porosityΦ None Giddingsrsquo ratio of external and total column porosity120601 None Flow resistance parameter based on mobile phase velocity1206010
None Flow resistance parameter based on superficial velocityΩ
119901None Particle sphericity factor
120578 gcmminus1secminus1 Fluid absolute viscosity120583119894
cmsecminus1 Average fluid interstitial linear velocity120583119904
cmsecminus1 Average fluid superficial linear velocity120583119905
cmsecminus1 Average mobile phase velocityΨ None Porosity dependence term in the Kozeny-Blake equation based on mobile phase velocityΨ
0None Porosity dependence term in the Kozeny-Blake equation based on superficial velocity
raw data for column B2 are supportive of the value of 267 for
1198750
92 Guiochonrsquos 1999 Study In a 1999 publication by Farkaset al entitled ldquoValidity of Darcyrsquos Law at Low Flow Rates inLiquid Chromatographyrdquo [20] Guiochon and his coauthorsreport the results of some very exactingmeasurements whichthey made to validate Darcyrsquos law at extremely low fluidflow rates Using a column containing particles packed toa measured external column porosity of 0399 and havinga measured total column porosity of 0593 (Figure 2 in thepaper) Guiochon et al measured the column pressure dropand fluid flow rate at flow rates ranging between 0015 and05mLminwith ethylene glycol as the fluidThe authors statethat the particles in the column are spherical and they appearto have gone to extraordinary lengths to obtain an accuratemeasure of the particle diameter and of course as a result ofmodern techniques of particle size classification the particlesize distribution is very narrow
In Figure 1 of the paper the authors show a plot ofcolumn measured pressure drop in bar versus ldquofluid linearvelocityrdquo in cmsec This plot represents their measurementstaken in the empty column The only velocity that exists in
an empty column is the superficial velocity and thus theabscissa values in the plot in Figure 1 designated as ldquofluidlinear velocityrdquo represent superficial linear velocity
In Figure 2 in their paper however the authors showanother plot of measured pressure drop in bar also versusldquofluid linear velocityrdquo in cmsec The velocity on the abscissain this plot although also designated as ldquofluid linear velocityrdquodoes not equate (for permeability related purposes) to theldquofluid linear velocityrdquo in Figure 1 This is because in Figure 2the columnwas filled with porous particles and themeasuredvelocity was the mobile phase velocity Indeed the onlyvelocity used throughout the entire paper is the mobile phasevelocity (mobile phase velocity and superficial velocity areone and the same in an empty column)
In Table 1 of their paper the authors report that for thecolumn represented in Figure 2 the slope of a line achievedwhen the mobile phase velocityin units of cmsec is plottedversus pressure drop in units of bar is 1829 They also reportthat the flow resistance parameter 120601 for this column is 865In Table 3 herein it can be seen that the raw data for thiscolumn (Column C) generates an apparent value of 257 for1198750 This value is very close to Giddingsrsquo value of 267 Indeed
by referring to Figures 2 and 3 herein one can see that thereported raw data in this study without any modification
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
Journal of Materials 3
Guiochonrsquos velocity frame upon his permeability coefficientBefore attempting to do this a review of the various velocityframes typically used to characterize flow in porous mediamay be helpful
3 Fluid Velocity Frames
31 Superficial Velocity The fluid velocity term in both theDarcy equation and the Kozeny-Blake equation is the volu-metric flow rate of the fluid divided by the cross-sectional areaof the empty column This velocity is known as the ldquoaveragefluid superficial linear velocityrdquo or simply the ldquosuperficialvelocityrdquo and is defined as follows
120583119904=119876
A (5)
where 119876 = volumetric flow rate 119860 = cross-sectional area ofthe column and 120583
119904= superficial velocity
Thus superficial velocity is not influenced in any way bythe existence or structure of the packed bed Furthermore itdoes not change depending upon whether the particles in thebed are porous or nonporous
32 Interstitial Velocity Theconvective fluid velocity betweenthe particles of a packed bed is called the ldquoaverage fluidinterstitial linear velocityrdquo or simply the ldquointerstitial velocityrdquoInterstitial velocity is the superficial velocity adjusted for thecross-section of the actual fluid flow in a packed columnIt is related to the superficial velocity through the externalporosityof the column 120576
0 as follows
120583119894=120583119904
1205760
(6)
where 120583119894= average fluid interstitial linear velocity
If interstitial velocity is to be substituted for superficialvelocity in the Kozeny-Blake equation one must also includethe external porosity term This term is the compensationfactor that maintains the integrity of the value of the per-meability coefficient 119875
0between pressure gradient and fluid
superficial velocity The equation then becomes
Δ119875
119871=1198750(1 minus 120576
0)2
120583119894120578
120576201198892119901
(7)
33 Mobile Phase Velocity Guiochon does not use either thesuperficial velocity or the interstitial velocity in the equationsin his textbook Instead he uses the ldquomobile phase velocityrdquoMobile phase velocity is uniquely a chromatographic term Itis the velocity calculated as a result of injecting an unretainedfully permeating solute into the column and measuring thetime ofits elution It is defined as follows
120583119905=119871
1199050
(8)
where 1199050= the time it takes an unretained fully permeating
solute to traverse the column and 120583119905=mobile phase velocity
The velocity experienced by the flowing fluid in a columnunder steady state conditions is not altered by the existence ofldquoblindrdquo pores in the particles of that column Blind pores arepores within the particles which have only one fluid openingthat is they are not through-pores Accordingly under steadystate conditions the pores are already filled with identicalstagnant fluid and there exists no driving force to exchangethis stagnant fluid with the fluid flowing in the interstices ofthe particles It will be appreciated that column permeabilitydata is usually taken under steady state conditions in orderto take advantage of a constant value for fluid viscosityAlthough a solute which is dissolved in the flowing fluid willpenetrate the stagnant fluid in the pores (assuming it is notstrictly excluded by either physical dimension or chemicalinteraction) because of molecular diffusion the driving forcefor this solute migration is the concentration gradient of thesolute between the fluid outside the particles and the fluidinside the particles This intraparticle migration of the solutedoes not contribute to friction and thus must not be includedin velocity considerations related to pressure gradientMobilephase velocity therefore is not strictly a fluid velocity terminstead it is a misnomer and it would better be termed theldquounretained fully permeating solute velocityrdquo
Although Guiochon defines the relationship betweenfluid flow and pressure gradient in the context of a residualpermeability coefficient 119896
0 and acknowledges that column
porosity plays a role in that relationship his use of mobilephase velocity fails to recognize that molecular diffusion inblind pores of the particle fraction of a column packed withporous particles does not influence pressure gradient
Mobile phase velocity is related to fluid superficial veloc-ity however through the total porosity of the column ldquoTotalcolumn porosityrdquo 120576
119905 is a term peculiar to the chromato-
graphic literature which relates to the use of columns packedwith porous particles It is the fraction of the column volumetaken up by the free space between the particles plus thefraction of the column volume taken up by the free space inthe internal pores of the particles Thus it is the sum of theexternal and the internal column porosities
120576119905= 120576
0+ 120576
119894 (9)
where 120576119905= total column porosity and 120576
119894= internal column
porosityMobile phase velocity is related to fluid superficial veloc-
ity through the total porosity of the column as follows
120583119905120576119905= 120583
119904 (10)
Thus if mobile phase velocity is used in the Kozeny-Blakeequation one must insert the compensation factor 120576
119905into the
equation in order to maintain the integrity of the value of thepermeability coefficient 119875
0 between pressure gradient and
fluid superficial velocity and the equation then becomes
Δ119875
119871=1198750(1 minus 120576
0)2
120583119905120576119905120578
120576301198892119901
(11)
4 Journal of Materials
And now reconciling (1) and (3) we get
1198750(1 minus 120576
0)2
120583119904
120576301198892119901
=120583119905120576119905
1198960(119889
119901119898Ω
119901)2 (12)
4 The Flow Resistance Parameter
Before proceeding any further we digress to define anotherterm which will become important to our understanding ofGuiochonrsquos work The chromatographic literature containsa term called the ldquoflow resistance parameterrdquo [7] The fluidvelocity underlying this parameter is based upon Giddingsrsquodefinition of ldquomean fluid velocity through a cross-sectionrdquoand is found in his 1965 text book at page 207 [8]2 It is definedas follows
120601 =Δ119875119889
2
119901
120583119905120578119871
(13)
where 120601 = flow resistance parameterThe flow resistance parameter therefore is equivalent to
the reciprocal of Guiochonrsquos term 1198960
120601 =1
1198960
(14)
5 Refining the Terms For Column Porosity
In order to continue with our comparison of Guiochonrsquosteaching to that embodied in the Kozeny-Blake equation andin particular to accommodate columns packed with porousparticles we need to refine the porosity dependence term inthe Kozeny-Blake equation To accomplish this we adopt thedefinition of column porosity as taught by J Calvin Giddings[8] namely
Ψ =(1 minus 120576
119905Φ)
2
(120576119905Φ)
3 (15)
where Ψ = the porosity dependence term in Kozeny-Blake(porous particles) and
Φ =1205760
120576119905
(16)
where Φ = the ratio between the external and total columnporosities
Using thismethodology in themost general case involvingporous particles to express total column porosity (where totalcolumn porosity is not the same as external column porosity)as opposed to in the special case involving nonporous particles(where total column porosity and external column porosityare the same) the Kozeny-Blake equation (using mobilephase velocity as in (10)) may now be reexpressed as follows
Δ119875
119871=1198750(1 minus 120576
119905Φ)
2
120583119905120576119905120578
(120576119905Φ)
3
(119889119901)2
(17)
It will be appreciated that in the special case of columnspacked with nonporous particles (17) is identical to (3)because here 120576
119905= 120576
0and thus 120583
119905= 120583
119894= 120583
119904120576
0and Φ = 1
6 The Relationship between 1198960
and 1198750
We can now connect the teaching of Guiochon to thatembodied in the Kozeny-Blake equation as follows
1198750=
1
1198960Ψ
(18)
or
1198750=120601
Ψ (19)
As illustrated by (19) the term 1198750in the Kozeny-Blake
equation is the flow resistance parameter normalized for thecolumn porosity dependence term
Note that if1198750is a true constant the two parameters upon
which it depends 120601 and Ψ will vary in direct proportion toeach other This is illustrated by the data in Table 1 hereinwhich is a reference chart where particle porosity 120576
119901 is varied
from nonporous at one extreme to highly porous at the otherwhere 120576
119901= volume of the free space within the particles in
a packed columnthe total volume of all the particles in thecolumn
Figure 1(a) is a plot of the data contained in Table 1 as120601 versus Ψ Proceeding from the lower left hand corner tothe upper right hand corner of the plot particle porosityincreases It is obvious from the plot that all the data in Table 1falls on a single line with a slope of 267 which we say is theconstant value of 119875
0 As is illustrated by the plot in the range
of porosity covered by the data the values for 120601 vary fromabout 500 to about 1900 and correspondingly the values ofΨvary from approximately 2 to approximately 7 Note also thatthe upper boundary value of Ψ for packed columns is about55 with larger values related to capillaries
Finally we point out that once the value of 1198750is correctly
identified ameasured value for the flow resistance parametermay then be used to calculate the value of the externalporosity of the column 120576
0 when the latterrsquos value has not or
cannot be measured
7 Guiochonrsquos Modified PermeabilityCoefficient 119875
119905
As reflected in some of his recent publications which wewill explore in more depth later Guiochon inadvertentlyapparently confuses the term 119875
0with another parameter 119875
119905
which we identify here and define for the first time 119875119905is not
the same as 1198750because it is based upon Guiochonrsquos unique
(from a permeability perspective) teaching for fluid velocityIt is defined as follows
119875119905= 119875
0120576119905 (20)
where 119875119905= Guiochonrsquos modified permeability coefficient
It will be appreciated that Guiochonrsquos 119875119905 if used in the
Kozeny-Blake equation in place of 1198750 accommodates his use
of themobile phase velocity in away that is different from butequivalent to a direct conversion of mobile phase velocity tosuperficial velocity
Journal of Materials 5
Table1Colum
nperm
eabilityreferencetable
119870119865
Particlepo
rosity
rang
e120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119896119871
119863Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
120576119905Φ
(120576119905minus1205760)
120576119894
Δ119875119889119901
2Δ119875119889119901
2(1minus120576119905Φ)2120576119905
(1minus1205760)2
1120601
1198750120576119905
119889119901
2120576119905
(119889119901119898Ω119901)
(1minus1205760)
120583119905120578119871
120583119904120578119871
(120576119905Φ)3
(1205760)3
120601Ψ
120601po
iseUnits
Non
eNon
eNon
eNon
eNon
epsi
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
cmNon
ecm
cmgcmminus1 secminus1
cm3 m
inminus1
cmsecminus
1cm
secminus
1cm
secminus
1
Non
porous
particles
0380
1000
0380
000
0000
0160
711
1870
2662
701
141119864minus03
267
101
535119864minus10
15046
100
000100
000100
00100
038
0038
0100
0100
0390
1000
0390
000
0000
0147
653
1675
244
6627
153119864minus03
267
104
597119864minus10
15046
100
000100
000100
00100
039
0039
0100
0100
040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
100
000100
000100
00100
040
004
00100
0100
0410
1000
0410
000
0000
0124
553
1349
2071
505
181119864minus03
267
109
742119864minus10
15046
100
000100
000100
00100
041
004
10100
0100
0420
1000
0420
000
0000
0115
509
1212
1907
454
196119864minus03
267
112
825119864minus10
15046
100
000100
000100
00100
042
004
20100
0100
Lowpo
rosity
particles
0507
0750
0380
0127
0204
213
948
1870
3549
701
106119864minus03
267
135
535119864minus10
15046
100
000100
000100
00100
051
0051
0133
0100
0520
0750
0390
0130
0213
196
871
1675
3261
627
115119864minus03
267
139
597119864minus10
15046
100
000100
000100
00100
052
0052
0133
0100
0533
0750
040
00133
0222
180
801
1502
2999
563
125119864minus03
267
142
666119864minus10
15046
100
000100
000100
00100
053
0053
0133
0100
0546
0750
0410
0136
0231
166
737
1349
2760
505
136119864minus03
267
146
742119864minus10
15046
100
000100
000100
00100
055
0055
0133
0100
0560
0750
0420
0140
0241
153
679
1212
2542
454
147119864minus03
267
149
825119864minus10
15046
100
000100
000100
00100
056
0056
0133
0100
Medium
porosity
particles
0613
0620
0380
0233
0376
258
1146
1870
4293
701
872119864minus04
267
164
535119864minus10
15046
100
000100
000100
00100
061
0061
0161
0100
0629
0620
0390
0239
0392
237
1053
1675
3946
627
949119864minus04
267
168
597119864minus10
15046
100
000100
000100
00100
063
0063
0161
0100
064
50620
040
00245
040
9218
969
1502
3629
563
103119864minus03
267
172
666119864minus10
15046
100
000100
000100
00100
064
0065
0161
0100
0661
0620
0410
0251
0426
201
892
1349
3340
505
112119864minus03
267
177
742119864minus10
15046
100
000100
000100
00100
066
006
60161
0100
0677
0620
0420
0257
044
4185
821
1212
3076
454
122119864minus03
267
181
825119864minus10
15046
100
000100
000100
00100
068
006
80161
0100
Highpo
rosity
particles
0760
0500
0380
0380
0613
320
1422
1870
5325
701
703119864minus04
267
203
535119864minus10
15046
100
000100
000100
00100
076
0076
0200
0100
0780
0500
0390
0390
0639
294
1306
1675
4891
627
766119864minus04
267
208
597119864minus10
15046
100
000100
000100
00100
078
0078
0200
0100
0800
0500
040
0040
0066
7270
1202
1502
4500
563
832119864minus04
267
214
666119864minus10
15046
100
000100
000100
00100
080
0080
0200
0100
0820
0500
0410
0410
0694
249
1105
1349
4140
505
905119864minus04
267
219
742119864minus10
15046
100
000100
000100
00100
082
0082
0200
0100
0840
0500
0420
0420
0724
229
1018
1212
3814
454
982119864minus04
267
224
825119864minus10
15046
100
000100
000100
00100
084
0084
0200
0100
Capillary
1000
0380
0380
0620
1000
421
1870
1870
7005
701
535119864minus04
267
267
535119864minus10
15046
100
000100
000100
00100
100
0100
0263
0100
1000
0390
0390
0610
1000
377
1675
1675
6273
627
597119864minus04
267
267
597119864minus10
15046
100
000100
000100
00100
100
0100
0256
0100
1000
040
0040
0060
010
00338
1502
1502
5625
563
666119864minus04
267
267
666119864minus10
15046
100
000100
000100
00100
100
0100
0250
0100
1000
0410
0410
0590
1000
303
1349
1349
5051
505
742119864minus04
267
267
742119864minus10
15046
100
000100
000100
00100
100
0100
0244
0100
1000
0420
0420
0580
1000
273
1212
1212
4541
454
825119864minus04
267
267
825119864minus10
15046
100
000100
000100
00100
100
0100
0238
0100
6 Journal of Materials
0200400600800
100012001400160018002000
20 30 40 50 60 70
NonporousLow porosityMedium porosity
CapillaryLinear (line)
Nonporous Low porosity particles
Medium porosity particles
High porosity particles
Capillary
Line slope (P0) = 267
y = 267x + 0
Ψ
120601
Line
High porosity
(a)
99111123135147159171183195207219231243255267
035 045 055 065 075 085 095
LineCapillaryLinear (line)
Capillary
Line slope (P0) = 267
y = 267x + 0
Pt
120576t
Φ = 10
Φ = 075
Φ = 062
Φ = 05
Φ = 075
Φ = 062
Φ = 05
Φ = 1
(b)
99111123135147159171183195207219231243255267
000 010 020 030 040 050 060 070 080 090 100
NonporousLow porosityMedium porosityHigh porosity
120576p
y = 160x + 107
Pt
LineCapillaryLinear (line)
(c)
Figure 1 (a) The balance between 120601 and Ψ (b) Column permeability reference chart (c) Particle permeability reference chart
In the 2006 edition of their textbook [9] Guiochon andhis coauthors provide the following worked example of hispressure dropfluid flow teaching3
In Table 2 a cross-reference is established between Guio-chonrsquos teaching for columns packed with both porous andnonporous particles the Kozeny-Blake equation (asmodifiedby Carman) and for validation purposes Giddingsrsquo teachingbased upon actual permeabilitymeasurements fromhis Table53-1 in his 1965 text [8]4
In his worked example Guiochonrsquos fluid velocity 119906is found in his text as his equation (256) [9] at page61 and equates to mobile phase not superficial velocityAccordingly when applied to columns packed with porousparticles his worked example combines the contributionto fluid velocity of mass transfer by molecular diffusionin the free space within the particles with mass transferby convection in the free space between the particles Asdiscussed above if one is attempting to compare Guiochonrsquos
teaching for the value of the permeability coefficient with thatof the Kozeny-Blake equation the conversion factor is thetotal porosity of the column
In addition because Guiochonrsquos particle sizeparameter has an embedded assumption due to particleshaperoughness his worked example also commingles thecontribution to spherical particle diameter equivalent ofparticle shaperoughness with other nonparticle variablesUnfortunately his worked example is not based uponcompletely spherical particles where this would not be anissue We know this because he states that 119896
0in his worked
example equals 1 times 10minus3 which he says is the value of thepermeability coefficient for columns packed with irregularparticles
In order to move forward one must simplify If this wereto be done experimentally it would best be accomplishedby experiments with nonporous particles in which case thevariable of molecular diffusion in the pores of the particles
Journal of Materials 7
Table2Cr
oss-referencefor
term
s119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119896119871
119863Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
120576119905Φ
(120576119905minus1205760)
120576119894
Δ119875119889119901
2Δ119875119889119901
2(1minus120576119905Φ)2120576119905(1minus1205760)2
1120601
1198750120576119905
119889119901
2120576119905
(119889119901119898Ω119901)
(1minus1205760)
120583119905120578119871
120583119904120578119871
(120576119905Φ)3
(1205760)3
120601Ψ
120601po
iseUnits
Non
eNon
eNon
eNon
eNon
epsi
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
cmNon
ecm
cmgcmminus1 secminus1cm
3minminus1cm
secminus
1cm
secminus
1cm
secminus
1
Nonporousp
articles
Gidding
srsquotradition
alno
nporou
scolum
n040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
1000010
00010
0010
0399
004
00100
0100
Guiocho
nrsquostradition
alspheric
alparticle(app
arent)
0362
1000
0362
000
0000
0187
830
2293
3106
858
120119864minus03
267
97436119864minus10
15046
1000010
00010
0010
0361
0036
0100
0100
Gidd
ingrsquos
table5
3-1
5060meshglassb
eads
0422
1000
0422
000
0000
0113
500
1184
1873
444
200119864minus03
267
113
844119864minus10
15046
1000010
00010
0010
0421
004
20100
0100
5060meshglassb
eads
040
910
00040
9000
0000
0126
560
1371
2097
513
179119864minus03
267
109
729119864minus10
15046
1000010
00010
0010
040
70041
0100
0100
Guiochonrsquosteaching
ExA
irregular
particle
(app
arent)
040
010
00040
0000
0000
0225
1000
2500
2250
563
100119864minus03
444
178
400119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
ExA
correctedforp
article
irregularity
040
010
00040
0000
0000
0225
601
1502
2250
563
166119864minus03
267
107
400119864minus10
15046
078
00010
000
080010
0399
004
00100
0100
ExB
spheric
alparticle
(app
arent)
040
010
00040
0000
0000
0187
830
2075
2250
563
120119864minus03
369
148
482119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
ExB
corrected
040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
Porous
particles
Gidding
srsquotradition
alpo
rous
column
060
0066
7040
00200
0333
203
900
1500
3375
563
111119864minus03
267
160
667119864minus10
15046
100
00010
00010
0010
0599
006
00150
0100
Gidd
ingrsquos
table5
3-1
3040meshalum
ina
0803
0498
040
00403
0672
271
1204
1499
4510
562
830119864minus04
267
214
667119864minus10
15046
100
00010
00010
0010
0801
0080
0201
0100
5060meshalum
ina
0837
0499
0417
0420
0721
235
1043
1246
3906
467
959119864minus04
267
224
803119864minus10
15046
100
00010
00010
0010
0835
0084
0201
0100
6080meshchromosorbW
(5DNP)
0766
0498
0382
0384
0621
316
1404
1833
5259
687
712119864minus04
267
204
545119864minus10
15046
100
00010
00010
0010
0764
0077
0201
0100
6080meshchromosorbW
(20
DNP)
0785
0495
0389
0396
064
8300
1333
1698
4991
636
750119864minus04
267
210
589119864minus10
15046
100
00010
00010
0010
0783
0079
0202
0100
8 Journal of Materials
would be eliminated and by using only smooth sphericalparticles such as glass beads in which case the variableof particle shaperoughness would be eliminated If both ofthese things were done one would be able to directly andunambiguously determine the true value of 119875
05
Accordingly since his textbook contains a general teach-ing without any expressed restrictions with respect to particleporosity let us first assume that the column in Guiochonrsquosworked example is packed with nonporous particles Sec-ondly let us assume that the value for the external porosityof the column (which is the same as the total porosity ina column packed with nonporous particles) is the valuetraditionally assumed for a typical well-packed column thatis 04
As illustrated by example A in Table 2 this would leadto an apparent value for 119875
0of 444 which is obviously much
too high6 Using the adjusted particle size of 8 micrometers(rounded) which is calculated based on a value of 078 forΩ
119901
we derive values of 267 for 1198750and 107 for 119875
119905(the relationship
between these two values is of course equal to the totalporosity of the column 120576
119905 as is dictated by (20))
Similarly as illustrated by example B in Table 2 when thecolumn of Guiochonrsquos worked example is adapted to accom-modate spherical nonporous particles Guiochonrsquos teachingof the value for 119896
0of 12 times 10minus3 results in an apparent value
of 369 for 1198750
7 again too high [10 11] On the other hand ifwe correct the value of 119875
0 we get a column external porosity
of 03620which is too low for a typical well-packed column(120576
0= 04) [8] Moreover in the case of both corrections
Guiochonrsquos teaching overstates the pressure gradient (187 psicompared to the 135 psi which Giddingsrsquo experiments wouldgenerate) Accordingly it is only when we correct for thepressure drop that we achieve appropriate values of 04 for1205760 267 for 119875
0 and 107 for 119875
119905
As is evident from Table 2 when the porosity of columnspacked with nonporous particles varies the value of 119875
119905also
varies because it is based upon the measurement of mobilephase velocity which in turn changeswith the porosity of thecolumn Indeed for typical columns packed with nonporousparticles having external porosities close to 04 (038ndash042)[8] Table 2 reveals that the value of 119875
119905varies by as much as 6
points while the value of 1198750remains constant at 267
Having established a range of values for 119875119905based upon
columns packed with nonporous particles one can nowproceed to evaluate the range of values for 119875
119905based upon
columns packed with porous particles which is to say thatwe can now evaluate the impact of particle porosity onpermeability Among other things Table 2 shows that whenit comes to porous particles Guiochonrsquos 119875
119905is even more
variable (and thus even less useful) because the value ofthe coefficient can differ on a column-to-column basis byas much as 90 points depending on the combination of theinternal andexternal column porosities at play
Figures 1(a) and 1(b) are a plot of the data found in Table 1In this plot Guiochonrsquos modified permeability coefficient 119875
119905
is plotted against column total porosity 120576119905 Once again it will
be appreciated that (a) all data falls on a line with a slope of267 which represents the value of 119875
0and (b) the line has no
intercept which means that when 120576119905is zero 119875
119905is also zero As
shown in the plot values of the parameter Φ are highest atlow values of 119875
119905and lowest at high values of 119875
119905 Further as is
evidenced by the respective definitions for the terms119875119905and119875
0
contained herein the only time that the value of Guiochonrsquos119875119905does not differ from the value of 119875
0(267) is when the value
of 120576119905is 1 its value in a theoretical packed column devoid of
particles that is acapillaryFinally Figure 1(c) is another plot of the data in Table 1
but this time the values for 119875119905are plotted against particle
porosity 120576119901 The equation of this line does have an intercept
which represents the value of 119875119905when the particles are
nonporous It is also obvious from the equation of this linethat it is only when particle porosity is unity that is in acapillary that the value of 119875
119905is 267
It will be appreciated that since 120576119905is the sumof the internal
and external porosities of a packed column any given valuefor 119875
119905involving a column packed with porous particles can
represent many different combinations of these parametersVice versa any given value for 120576
119905can generate many different
values for 119875119905depending on the amount of particle porosity
which is present On the other hand as is pointed outearlier column internal porosity andor particle porosityplay no role in determining the permeability of a packedcolumn the only porosity parameter which is relevant forsuch purposes is column external porosity In other words asa permeability coefficient 119875
119905is not a particularly informative
concept because it still fails to provide any meaningful tie tothe underlying physical phenomena that govern the pressuredropfluid flow relationship
In the case of permeability measurements one quan-tifies only total pressure drop Pressure drop however isa commingling of the contributions of the several differ-ent parameters embodied in the pressureflow relationshipTherefore one must accurately represent the contributionof each factor before adding them together to equal thepressure drop Obviously if one mistakes the contribution ofany given element for instance particle size this will skewthe contribution of some other element when summed forinstance column porosity
Our frame of reference has been calibratedwith respect tothe interrelationships between the values of 119875
0 119875
119905 120576
119905 and 120576
119901
based on the data reported in Giddingsrsquo table 53-1 in his 1965textbook Since his data base was derived from experimentsusing nonporous particles the permeability results of whichwere extended to columns packed with fully porous particlesby using his Φ parameter the validity of which in turn wasestablished by independent means there is no uncertaintywith respect to the critical variable of external porosity 120576
0
inherent in our frame of reference It is for this reasontherefore that our frame of reference trumps any reporteddata which uses any other technique including the techniqueof inverse size exclusion chromatography by itself to establishvalues for 120576
0[12] In addition our frame of reference is
internally consistent due to the grounding of the parametersat the boundary condition of a theoretical capillary at whichpoint the values of 120576
119905and 120576
119901are unity and the values of 119875
0and
119875119905are equal In essence these interrelationships exemplify the
conservation laws of nature which are the ultimate standard
Journal of Materials 9
of measure when considering partial fractions of a singleentity
Consequently as Figures 1(b) and 1(c) illustrate if oneknows based upon other independent means the value of acolumnrsquos total porosity or its particlesrsquo porosity respectivelyone can at least verify whether results of permeability exper-iments expressed in terms of 119875
119905are reasonable Or to put
it another way if the 119875119905results reported do not conform to
what these plots tell us one should expect from columns andparticles based upon independently derived porosity valuesfor the particles under study we know that something iswrong Accordingly this is the way in which we use 119875
119905below
to analyze and evaluate Guiochonrsquos experimental work of thelast decade
8 The Origin of Guiochonrsquos Value forColumn Permeability
In 1966 Guiochon wrote a review paper entitled ldquoProblemsRaised by the Operation of Gas Chromatographic Columnsrdquo[13] In that paper he reviewed the development of the per-meability equation from Darcy to the then present Amongother things he reported that ldquoa mean 120576 of 038 is foundfor a number of packings obtained from powdered prod-ucts of various origins consisting of spherical or irregularparticles Almost all the measured values are between035 and 045rdquo (p 14) He then went on to discuss what hecalled the ldquosemiempirical Blake-Kozeny equationrdquo which hereported as 119896 = [119889
2
1199011205762][150(1 minus 120576)
2] [13 p 15 Equation
(11)] What sticks out of course is the fact that Guiochonstates here that 119875
0equals 150 the value for the permeability
coefficient championed by Ergun not the value posited byCarman To be fair we should point out that Guiochongave warning that he was uncomfortable with any particularvalue for the permeability coefficient For even though herecited the Kozeny-Blake equation with a value of 150 for itspermeability coefficient he stated that ldquothe values of 119896 givenby (this) equation are not very accurate and the deviationscan be as large as 50rdquo [13 p 15] Guiochon attributedmost ofthis uncertainty to the inability of practitioners of that era toeffectively measure the value of a columnrsquos external porositySince as he pointed out ldquo(the Kozeny-Blake) equation isvery sensitive to variations in 120576rdquo he concluded that theequation is ldquoof little userdquo and that it was better practice torely upon grosser measurements of permeability [13 p 15]Thus Guiochon at this timeframe made a conscious decisionto staywithin the relatively safe but admittedlymore obscureboundaries of Darcyrsquos teaching on column permeability
In 2006 however Guiochon emerged from under theshadow of Darcy permeability when he published a reviewpaper which purported to be a comprehensive summaryof the current state of the art in chromatography [14]In light of the fact that Guiochon and his coauthors hadvirtually contemporaneously published a new version of theirtextbook one would have expected that Guiochon wouldtake this opportunity to reaffirm his textbook teaching So itshould come as no surprise that he would open his discussionof the pressure dropfluid flow relationshipwith the following
assertion ldquothe permeability of packed beds used in liquidchromatography is in the order of 1 times 10minus3rdquo [14 p 9] Asone would expect the authority which Guiochon cites forthis seemingly familiar proposition is the 2006 edition of hisown textbook For the first time however he also providessome citations to third party sources In particular he cites a1997 chromatography handbook written by Uwe Neue ChiefScientist for Waters Corporation [15 p 9] As one can seeNeuersquos equationmdashlike Guiochonrsquosmdashdoes indeed postulate apermeability coefficient with a value of 1 times 10minus3
At this point however Guiochon abandons his tradi-tional caution about the value of the permeability coeffi-cient and marries his permeability equation to the Kozeny-Blake equation thereby endorsing not only the porositydependence term in that equation but also a specific valuefor 119875
0 Referring to his value of 1 times 10minus3 he states the
following ldquothe specific column permeability is related to thecolumn porosity through the Kozeny-Carman equation 119896
119891=
12057630180 (1 minus 120576
0)2rdquo [14 p 9] Not surprisingly Neue makes
the same connection ldquothe specific permeability depends onthe particle size 119889
119901and the interstitial porosity 120576
0of the
packed bedThe relationship is known as theKozeny-Carmanequation 119861
0= (1185)(1205763
01198892119901(1 minus 120576
0)2)rdquo [15 p 30]
Neuersquos permeability equation and Guiochonrsquos textbookpermeability equation however differ in that Neue uses thesuperficial not the mobile phase velocity Accordingly sinceall other features of their equations are the same one wouldexpect them to get different values for their permeabilitycoefficients Neue himself recognizes this when he says inhis handbook ldquothis value [1000] is also in good agreementwith our experience but values as low as 500 have beenreported in the literature8 However there are several reasonswhy different values reported by different workers do notrepresent true differences in the resistance parameter Firstlower values are obtained for porous packings if the specificpermeability is measured on the basis of the breakthroughtime of an unretained peak instead of using the flow raterdquo see[15 p 32] (when he distinguishes the ldquobreakthrough time ofan unretained peakrdquo from the ldquoflow raterdquo Neue is of coursereferring to the mobile phase velocity and the superficialvelocity resp) Consequently if one takes Neuersquos point andturns it on its head one can see that if he and Guiochonwere to get the same value for their permeability coefficients[1000] but one (Guiochon) is using mobile phase velocityto derive it and the other (Neue) uses superficial velocity toderive it this would indeed represent a ldquotrue differencerdquo intheir equations
So how can Guiochon cite both Neue and himself for theproposition in his review paper that ldquothe permeability of thepacked beds used in liquid chromatography is in the order of1times 10minus3rdquoThe answer is that at least for purposes of his reviewpaper he adopts Kozeny-Blakersquos fluid velocity convention inplace of the velocity frame he uses in his textbook Here iswhat he says ldquothis approximation [1 times 10minus3] is valid only ifthe superficial velocity is usedrdquo [14 p 9] (emphasis supplied)
As a matter of pure mathematics Guiochonrsquos permeabil-ity coefficient of 1 times 10minus3 can only be equal to 1205763
0180(1 minus 120576
0)2
if the value of 1205760is 04 the typical value for the interstitial
10 Journal of Materials
fraction of a well-packed column Guiochon confirms in hisreview paper that this is indeed what he assumes in his state-ment of the Kozeny-Blake equation [14 p 9] Accordinglyhis adoption of the value of 180 for 119875
0solely depends on the
validity of his assertion that the value of the permeabilitycoefficient is 1 times 10minus3
We already know from our discussion of Guiochonrsquostextbook teaching that when this value is associated with apermeability equation based on mobile phase velocity it iswrong As reflected in our Table 2 when Guiochonrsquos workedexample involving this figure is corrected to account for theembedded factor due to the irregularity of the particles whichhis worked example assumes the value of the permeabilitycoefficient generated by his equation is not 1 times 10minus3 it is 166times 10minus3
So where did this value of 1 times 10minus3 come from AlthoughGuiochon in his review paper cites Neuersquos handbook for thisvalue Neue himself offers absolutely no authority for it Onthe other hand Guiochon does provide one other citation inhis review paper for this value The citation is to an article byHalasz et al published in 1975 [16]
If one then goes to that article one sees that Halaszet al do indeed proffer the same equation as Neue (albeitin a somewhat different format) and yes it does containthe value of 1 times 10minus3 See [16 Equation (1)] Neverthelessthe authors fail to identify from whence they obtained thisvalue On the other hand they do make reference to apaper that Endele et al wrote a year earlier in 1974 withKlaus Unger entitled ldquoInfluence of the Particle Size (5ndash35 120583)of Spherical Silica on Column Efficiencies in High-PressureLiquid Chromatographyrdquo [17] It appears that this is wherethe body is buried for it is in this article that Halasz reportshaving actually derived a value of 1 times 10minus3 for the permeabilitycoefficient
In summarizing the results of their experiments whichincidentally are based on measurements of superficial notmobile phase velocity (see their Equation (7)) Halasz et alcalculate an average value for the permeability coefficient(which they notate as119870
119865) of 1198892
1199011080 Accordingly they state
the following ldquo[I]t is proposed to define an average particlesize 119889
119901 for columns packed with spherical silica using the
balanced density packing method by the equation 1198892119901
=
1198701198651000 = 1198651205781198711031199032120587Δ119875rdquo [17 p 382 their Equation (7)]To recapitulate it thus appears that (1) Halasz was the
source of the value of 1 times 10minus3 for the permeability coefficientwhich Guiochon adopts in his review paper (2)Halaszrsquo valuewas based upon measurements which involved sphericalparticles and (3)Halaszrsquo value was properly derived from anequation which utilized the superficial velocity as the fluidvelocity
However we still have a problem if we are going to acceptthe value of 1 times 10minus3 as a basis for calculating the constantin Kozeny-Blake we still need to know the value of Halaszrsquocolumnsrsquo external porosity For example in the absenceof measured values for the external porosity Guiochonrsquosassertion in his review paper that the value of the constantis 180 is without foundation
Unfortunately however Halasz bases his methodologyon ldquoassumedrdquo versus ldquomeasuredrdquo values for external porosityNevertheless he does provide some clues that may helpus peel back this onion First he says that his value forthe permeability coefficient applies to ldquocolumns using thebalanced density packing methodrdquo Secondly after statinghis proposed value for the coefficient we find the followingtelltale sentence ldquothe permeability of columns packed by theconventional ldquodryrdquo method are worse by a factor of about 2than those packed by the balanced density methodrdquo [17 p382] For this proposition Halasz cites yet another of his ownarticles this time a paper written by himself and M Naefein 1972 entitled ldquoInfluence of Column Parameters on PeakBroadening in High-Pressure Liquid Chromatographyrdquo [18]
When one follows this thread by going to Halaszrsquos 1972paper one finds a possible solution to the puzzle For this iswhere Halasz describes his calculations of the permeabilitycoefficient from experiments with a number of conventionaldry-packed columns containing spherical silica particlesNot surprisingly the resulting value for the permeabilitycoefficient for such columns was not 1 times 10minus3 Instead Halaszsays that his experiments with such columns resulted inthe following value for the permeability coefficient ldquo119870 sim
11988921199012000rdquo [18 p 78] In other words consistent with what
Halasz says in his 1974 article the value for the coefficient forthese dry-packed columns was indeed ldquoworse by a factor ofabout 2rdquo compared to its value for columns prepared with thebalanced density method
The obvious implication of this is that when Halasz gota value of 1 times 10minus3 for the permeability coefficient in 1974he must have been evaluating columns that were packed toan interstitial fraction considerably higher than the externalporosity of the columns that were the subject of his 1972experiments Our problem however is still not resolvedbecause Halasz did not accurately determine the externalporosity of the columns under study in either the 1972 or 1974columns
We surmise that in their 1972 paper Halasz and hiscoauthors made use of the teaching of Giddings outlinedin his 1965 textbook which at the time of their 1972 paperwas seven years old Here is what they say ldquoexperiencehas shown that in a regular packed column with the sievefractions usual for chromatography 120576
0is independent of the
particle size if the particles are more or less spherical In anacceptable approximation 120576
0is equal to 04 In those columns
packed with nonporous supports 119906V and 119906 are identicalUsing porous supports (ie silica gel alumina chromosorbPorasil and so forth) 119906 = 4119865(1205871198892
119888120576119879) where 120576
119879is the
total porosity and indicates the pore volume of the supportrdquoIt will be appreciated that 119906V stands for interstitial velocity(our 120583
119894) and 119906 stands for mobile phase velocity (our 120583
119905) The
ldquoexperiencerdquo which Halasz and his coauthors are referringto concerning ldquocolumns packed with nonporous supportsalumina and chromosorbrdquo is that recorded by Giddings inTable 53-1 in his 1965 text Continuing to establish theinterrelationships inherent in the various entities involvedin chromatographic columns with respect to both interstitialandmobile phase velocity the authors conclude ldquoformally 119906V
Journal of Materials 11
can be calculated for a porous support assuming 1205760is equal
to 04rdquo (emphasis supplied) The authors then announce themethodology underlying their frame of reference with regardto column permeability as follows ldquowe determined in all ofour columns the 119906119906V ratio to be 056 with a differentialrefractometer Consequently the total porosity 120576
119905was equal
to 0714rdquo It will be appreciated that their ratio of 119906119906V is oneand the same as GiddingsrsquoΦ parameter as utilized by him inhis 1965 textbook and as we define hereinabove
We can now identify the origin of the discrepancybetween Halasz and Giddings and by extension betweenHalasz and Guiochon (and Neue) For even though Halaszmakes use of the results in Giddingsrsquo Table 53-1 he failsto appreciate the teaching inherent therein regarding theimportance of an accurate value for column external porositywhen using porous particles In establishing his values forΦGiddings was careful to ground his values for the externalporosity of columns packed with porous particles in acomparison of their measured pressure drops with measuredpressure drops in columns packed with nonporous particleswhere the values for (external) porositywere accurate becausethey are equal to the columnsrsquo total porosity
In their 1972 paper Halasz et al used a refractometer todetect the inert solute 119899-pentane in the mobile phase of 119899-heptaneThey injected the 119899-pentane into the flowing mobilephase the flow rate of which they presumably measured witha volumetric cylinder and stop watch Since the exit timeof the 119899-pentane peak was identified by the refractometertheir measurement of holdup volume was accurate andaccordingly so was their value of 0714 for 120576
119905(there is no
uncertainty related to column total porosity when using inertsolutes)This in turn led to an accurate value for their119906 termmobile phase velocity However since they did not measurethe interstitial velocity 119906V but rather used an estimate basedupon the measured flow rate and an ldquoassumedrdquo value of 04for external porosity their calculatedestimated value for 119906Vwas arbitrary This in turn means that the value of 056 fortheirΦ ratio was likewise arbitrary
In Table 3 herein we show the results for column number1 in the 1972 study which we designate as Column A
1 Note
that the authorsrsquo raw values correspond to a value of 500 for1198750 a valuewhich is obviously far too high As is plainly evident
in our Figures 2 and 3 this column data set does not conformin any reasonable fashion to the norms in our referencecharts Accordingly the data is badly flawed In our correcteddata for Column A
1in Table 3 herein we adjust the column
external porosity to the lowest value permitted in our frame ofreference (038) This correction results in a revised value forGiddingsrsquoΦparameter of 0532 In additionwe are compelledto adjust the particle downward (to 11 micrometers approx)as is also dictated by our frame of reference We justify ourlow corrected value for external porosity and our downwardadjustment for particle size on the aggressive ldquodry-packingrdquoprocess described by the authors ldquo[T] the best results wereachieved with the following method the column (id =2mm) 25 cm long was packed with 25 equal portions of theldquobrushrdquo After adding each portion the column was vibratedand afterward tapped on the floor and also vigorously tappedwith a rod from the siderdquo Based upon a column packing
experience of the principal author herein spanningmore than30 years we believe that these packing conditions generateda packed column with a low external porosity and also one inwhich the particles experienced significant breakage
The main differences between the columns in the 1972and 1974 papers concerned the particle sizes and themethodsused to pack the columns The particle diameters in the 1972study fell in a range of about 15ndash200 micrometers while thosein the 1974 study were much smaller being in the range of 5ndash40 micrometers approximately Moreover while the columnsin the 1972 study were dry-packed the columns in the 1974study were slurry-packed (what Halasz calls ldquothe balanceddensity packing methodrdquo) The authors of the 1974 paperrecite that their permeability results shown in their Table 4correspond to values for 120601
0of 1080 and values for 120601 of 910
In Table 3 herein we show the results for the 1974 study as anaverage value in our column designated as A
2 Note that the
raw data reported by the authors which again has the built-inassumption that the column external porosity was 04 placesthe value of 119875
0at 192 In our corrected values for Column A
2
we show that a value of 267 for 1198750places the column external
porosity at a value of 043 The ratio of the value for 120601 of 910when compared to the value for 120601 in the 1972 study of 2007is about 22 (2007910 = 22) which is in good agreementwith the authorsrsquo statement that the permeability of the 1972columns was worse by a factor of about 2 Again based onthe experience of the principal author herein in the packingof hundreds of thousands of commercially sold slurry packedcolumns we conclude that the balanced density techniquedescribed by the authors in the 1974 paper is fully capable ofproducing columns with this high and external porosity
In summary we conclude that for purposes of theKozeny-Blake equation (a) the external porosity of theHalasz 1972 columns was 038 (b) the external porosity of thecolumns which he tested in 1974 was 043 and (c) Halaszrsquos1972 and 1974 studies when properly analyzed corroborateGiddingsrsquo value for 119875
0of 267 not Carmanrsquos value of 180
Although the fault for all of the misdirection about the valueof the permeability coefficient being 1 times 10minus3 therefore restsprimarily on the shoulders of Halasz one has to question whyGuiochon (andNeue) so readily adopted itWhatevermay bethe reason Guiochonrsquos support for this proposition has beenand continues to be a cause for much of the conventionalacceptance of Carmanrsquos (erroneous) figure of 180 as thecharacteristic value of the coefficient in the Kozeny-Blakeequation
9 Guiochonrsquos Experimental Data
In addition to the data from Halasz 1972 and 1974 studiesTable 3 herein contains data from a representative sampleof Guiochonrsquos personal experimental permeability investi-gations over approximately the last decade as reported inthe numerous trade journals particularly the Journal ofChromatography A In each case the table has a single rowdedicated to Guiochonrsquos reported raw data and a single rowdedicated to corrected values for each column based on thereference chart in Table 1 herein
12 Journal of MaterialsTa
ble3Summaryexperim
entald
ata
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Naefe
andHalasz
columns
1198601(19
72)
Colum
nA1rawdata
07140
0560
040
000314
0523
782007
2811
4016
563
498119864minus04
500
357
908119864minus10
648119864minus10
2500
020
100
000135
0001350
00038
100
053
132
074
Colum
nA1corrected
07140
0532
03800
0334
0539
7813
3518
705002
701
749119864minus04
267
191
908119864minus10
648119864minus10
2500
020
100
000110
000110
100038
100
053
139
074
Endelean
dHalasz
columns
1198602(19
74)
Colum
nA2rawdata
08426
0475
040
00044
30738
11910
1080
4740
563
110119864minus03
192
162
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
033
016
Colum
nA2corrected
08426
0511
04309
0412
0723
11910
1080
3411
405
110119864minus03
267
225
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
031
016
Guiochon
columns
1198612ndash1198614(19
98)
Colum
nB 2
rawdata
05639
0722
040
730157
0264
22771
1367
2931
520
130119864minus03
263
148
467119864minus10
263119864minus10
429
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 3
rawdata
05582
0704
03932
0165
0272
44764
1369
3382
606
131119864minus03
226
126
471119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 4
rawdata
05509
0710
03909
0160
0263
72784
1422
3422
621
128119864minus03
229
126
459119864minus10
253119864minus10
1328
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 3
corrected
05653
0723
040
860157
0265
44774
1369
2899
513
129119864minus03
267
151
465119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 4
corrected
05651
0723
040
830157
0265
72776
1374
2907
514
129119864minus03
267
151
464119864minus10
262119864minus10
1375
500
100
000
060
000
0600
00165
98008
020
015
Guiochon
columnC
(1999)
Colum
nCrawdata
05930
0673
03990
0194
0323
183
865
1459
3372
569
116119864minus03
257
152
109119864minus09
645119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Colum
nCcorrected
05930
0673
03990
0194
0323
190
900
1518
3372
569
111119864minus03
267
158
105119864minus09
620119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Guiochon
columns
1198631ndash1198636(2006)
Colum
nD
1rawdata
07830
0456
03570
0426
0663
86694
886
7115
909
144119864minus03
975
76334119864minus10
261119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
013
Colum
nD
2rawdata
07230
0491
03550
0368
0571
88656
907
6726
930
152119864minus03
975
70353119864minus10
255119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
014
Colum
nD
3rawdata
06850
0506
03466
0338
0518
97685
1000
7025
1026
146119864minus03
975
67338119864minus10
231119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD
4rawdata
06560
0521
03415
0315
0478
103
696
1062
7143
1089
144119864minus03
975
64332119864minus10
218119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD5rawdata
06190
0545
03371
0282
0425
109
692
1118
7100
1147
144119864minus03
975
60334119864minus10
207119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
016
Colum
nD
6rawdata
05760
0587
03383
0238
0359
107
635
1103
6516
1131
157119864minus03
975
56364119864minus10
210119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
017
Colum
nD
1corrected
07830
0542
04243
0359
0623
86907
1158
3397
434
110119864minus03
267
209
334119864minus10
261119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
013
Colum
nD
2corrected
07230
0584
04221
0301
0521
88857
1186
3212
444
117119864minus03
267
193
353119864minus10
255119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
014
Colum
nD
3corrected
06850
0603
04129
0272
0463
97896
1307
3354
490
112119864minus03
267
183
338119864minus10
231119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
015
Colum
nD
4corrected
06560
0621
040
730249
0420
103
911
1388
3411
520
110119864minus03
267
175
332119864minus10
218119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
015
Colum
nD
5corrected
06190
0650
04025
0217
0362
109
905
1462
3390
548
110119864minus03
267
165
334119864minus10
207119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
016
Colum
nD
6corrected
05760
0701
04038
0172
0289
107
831
1442
3111
540
120119864minus03
267
154
364119864minus10
210119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
017
Guiochon
columns
1198641-1198642
(2007)
Colum
nE 1
rawdata
064
500594
03830
0262
0425
356
1119
1735
4371
678
894119864minus04
256
165
804119864minus11
519119864minus11
1500
046
100
000
030
000
0300
000
41300
030
079
047
Colum
nE 2
rawdata
05400
0800
04320
0108
0190
470
1000
1853
2161
400
100119864minus03
463
250
729119864minus11
393119864minus11
1500
046
100
000
027
000
0270
000
41300
030
070
056
Colum
nE 1
corrected
064
500594
040
000245
040
8356
969
1502
3628
563
103119864minus03
267
172
804119864minus11
519119864minus11
1500
046
100
000
028
000
0279
000
41300
030
075
047
Colum
nE 2
corrected
05400
0800
04320
0108
0190
470
577
1068
2161
400
173119864minus03
267
144
729119864minus11
393119864minus11
1500
046
088
000
023
000
0205
000
41300
030
070
056
Guiochon
columns
1198651ndash1198654
(2008)
Colum
nF 1
rawdata
06350
0586
03720
0263
0419
197
725
1142
4865
766
138119864minus03
149
95399119864minus11
253119864minus11
300
021
100
000
017
000
0170
00035
100
048
129
076
Colum
nF 2
rawdata
064
200581
03730
0269
0429
361
807
1258
4863
758
124119864minus03
166
107
358119864minus11
230119864minus11
500
021
100
000
017
000
0170
00035
100
048
129
075
Colum
nF 3
rawdata
064
100588
03770
0264
0424
670
748
1166
4643
724
134119864minus03
161
103
387119864minus11
248119864minus11
1000
021
100
000
017
000
0170
00035
100
048
128
075
Colum
nF 4
rawdata
06390
0595
03800
0259
0418
1014
752
1177
4476
701
133119864minus03
168
107
384119864minus11
246119864minus11
1500
021
100
000
017
000
0170
00035
100
048
127
075
Colum
nF 1
corrected
06350
0630
040
000235
0392
197
954
1502
3572
563
105119864minus03
267
170
399119864minus11
253119864minus11
300
021
100
000
019
000
0195
00035
100
048
120
076
Colum
nF 2
corrected
064
200623
040
000242
0403
361
964
1502
3611
563
104119864minus03
267
171
358119864minus11
230119864minus11
500
021
100
000
019
000
0186
00035
100
048
120
075
Colum
nF 3
corrected
064
100624
040
000241
0402
670
963
1502
360
6563
104119864minus03
267
171
386119864minus11
248119864minus11
1000
021
100
000
019
000
0193
00035
100
048
120
075
Colum
nF 4
corrected
06390
0626
040
000239
0398
1014
960
1502
3594
563
104119864minus03
267
171
384119864minus11
246119864minus11
1500
021
100
000
019
000
0192
00035
100
048
120
075
Journal of Materials 13
Table3Con
tinued
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Guiochon
columns
1198671-1198672serie
s(2010)
Colum
nH
1rawdata
05320
0744
03960
0136
0225
120
563
1057
3125
587
178119864minus03
180
96122119864minus10
652119864minus11
1500
046
100
000
026
000
0262
00104
050
005
013
009
Colum
nH
2rawdata
05420
0686
03720
0170
0271
61747
1379
4152
766
134119864minus03
180
98823119864minus11
446119864minus11
1000
046
100
000
025
000
0248
00054
050
005
013
009
Colum
nH
1corrected
05320
0752
040
000132
0220
120
799
1502
2993
563
125119864minus03
267
142
122119864minus10
652119864minus11
1500
046
100
000
031
000
0313
00104
050
005
013
009
Colum
nH
2corrected
05420
0738
040
000142
0237
61814
1502
3049
563
123119864minus03
267
145
823119864minus11
446119864minus11
1000
046
100
000
026
000
0259
00054
050
005
013
009
Guiochon
columns
1198673-1198674serie
s(2010)
Colum
nH
3rawdata
06270
060
903820
0245
0396
463
700
1117
4296
685
143119864minus03
163
102
447119864minus11
280119864minus11
1500
021
100
000
018
000
0177
00036
050
024
063
038
Colum
nH
4rawdata
05170
0791
040
900108
0183
433
616
1191
2639
511
162119864minus03
233
121
580119864minus11
300119864minus11
1500
021
100
000
019
000
0189
00036
050
024
059
047
Colum
nH
3corrected
06270
0638
040
000227
0378
463
942
1502
3527
563
106119864minus03
267
167
447119864minus11
280119864minus11
1500
021
100
000
021
000
0205
00036
050
024
060
038
Colum
nH
4corrected
05170
0791
040
900108
0183
433
699
1353
2639
511
143119864minus03
265
137
580119864minus11
300119864minus11
1500
021
100
000
020
000
0201
00036
050
024
059
047
Guiochon
columns
1198681-1198686
(2010)
Colum
nI 1rawdata
06580
0562
03700
0288
0457
488
741
1127
5156
784
135119864minus03
144
95390119864minus11
256119864minus11
500
021
100
000
017
000
0170
00104
050
024
065
037
Colum
nI 2rawdata
06550
0576
03770
0278
044
6873
661
1009
4745
724
151119864minus03
139
91437119864minus11
286119864minus11
1000
021
100
000
017
000
0170
00104
050
024
064
037
Colum
nI 3rawdata
06550
0588
03850
0270
0439
1209
610
931
4341
663
164119864minus03
141
92474119864minus11
310119864minus11
1500
021
100
000
017
000
0170
00104
050
024
062
037
Colum
nI 4rawdata
06430
0572
03680
0275
0435
260
787
1225
5153
801
127119864minus03
153
98367119864minus11
236119864minus11
500
030
100
000
017
000
0170
00104
050
012
032
018
Colum
nI 5rawdata
06540
0583
03810
0273
044
144
3683
1044
4531
693
146119864minus03
151
99423119864minus11
277119864minus11
1000
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 6rawdata
06550
0585
03830
0272
044
164
6665
1016
4438
678
150119864minus03
150
98434119864minus11
285119864minus11
1500
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 1corrected
06580
060
8040
000258
0430
488
988
1502
3701
563
101119864minus03
267
176
390119864minus11
256119864minus11
500
021
100
000
020
000
0196
00104
050
024
060
037
Colum
nI 2corrected
06550
0611
040
000255
0425
873
984
1502
3684
563
102119864minus03
267
175
437119864minus11
286119864minus11
1000
021
100
000
021
000
0207
00104
050
024
060
037
Colum
nI 3corrected
06550
0611
040
000255
0425
1209
984
1502
3684
563
102119864minus03
267
175
474119864minus11
310119864minus11
1500
021
100
000
022
000
0216
00104
050
024
060
037
Colum
nI 4corrected
06430
0622
040
000243
040
5260
966
1502
3617
563
104119864minus03
267
172
367119864minus11
236119864minus11
500
030
100
000
019
000
0188
00104
050
012
029
018
Colum
nI 5corrected
06540
0612
040
000254
0423
443
982
1502
3679
563
102119864minus03
267
175
423119864minus11
277119864minus11
1000
030
100
000
020
000
0204
00104
050
012
029
018
Colum
nI 6corrected
06550
0611
040
000255
0425
646
984
1502
3684
563
102119864minus03
267
175
434119864minus11
285119864minus11
1500
030
100
000
021
000
0207
00104
050
012
029
018
Guiochon
columns
1198691-1198692
(2011)
Colum
nJ 1rawdata90
∘A(120578
=00054
P)04980
0803
040
000098
0163
65654
1313
2801
563
153119864minus03
234
116126119864minus10
627119864minus11
1500
046
100
000
029
000
0287
00054
050
005
013
010
Colum
nJ 2rawdata90
∘A(120578
=00104
P)04980
0803
040
000098
0163
124
651
1307
2801
563
154119864minus03
232
116127119864minus10
630119864minus11
1500
046
100
000
029
000
0287
00104
050
005
013
010
Colum
nJ 1rawdata160
∘A(120578
=00054
P)05630
0714
04020
0161
0269
71896
1592
3099
550
112119864minus03
289
163
102119864minus10
573119864minus11
1500
046
100
000
030
000
0302
00054
050
005
012
009
Colum
nJ 2rawdata160
∘A(120578
=00104
P)05630
0714
04020
0161
0269
129
847
1504
3099
550
118119864minus03
273
154
108119864minus10
606119864minus11
1500
046
100
000
030
000
0302
00104
050
005
012
009
Colum
nJ 1corrected
04980
0803
040
000098
0163
65748
1502
2801
563
134119864minus03
267
133
126119864minus10
627119864minus11
1500
046
100
000
031
000
0307
00054
050
005
013
010
Colum
nJ 2corrected
04980
0803
040
000098
0163
124
748
1502
2801
563
134119864minus03
267
133
127119864minus10
630119864minus11
1500
046
100
000
031
000
0308
00104
050
005
013
010
Colum
nJ 1corrected
05630
0714
04020
0161
0269
71827
1470
3099
550
121119864minus03
267
150
102119864minus10
573119864minus11
1500
046
100
000
029
000
0290
00054
050
005
012
009
Colum
nJ 2corrected
05630
0714
04020
0161
0269
129
827
1470
3099
550
121119864minus03
267
150
108119864minus10
606119864minus11
1500
046
100
000
030
000
0299
00104
050
005
012
009
14 Journal of Materials
050
100150200250300350400
045 050 055 060 065 070 075 080 085
LineAll corrected data
Linear (all corrected data)Squares raw data Triangles corrected data
Pt
120576t
D seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
J1 J2 H4
H1H2
D6
B2 B3 B4
J3 J4
E2
C
H3
D5 D4
E1
D2
A1
D1
A2
y = 267xLine slope = 267
Figure 2 Summary of experimental results
0
500
1000
1500
2000
2500
20 25 30 35 40 45 50 55 60 65 70 75 80
Linear (line)
Ψ
LineAll corrected dataD seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
Line slope (P0) = 267
J1 J2
J3 J4E2
E1
A1
A2
H4 H1
C
120601
H2H3I1 I2 I3 I4 I5
F1 F2 F3 F4
D1 D2 D3 D4 D5 D6
Squares raw data Triangles corrected data
Figure 3 Summary of experimental results
Figures 2 and 3 herein are graphical representations ofTable 31015840s raw and corrected data for Guiochonrsquos columnsplotted in the same reference frames as in Figures 1(a) and1(b) respectively
As shown in the plots of the thirty-one total columnsevaluated in this data base only six columns had reportedraw data values for 119875
0close to the normThese were columns
B2 C E
1 H
4 J
1and J
2 Interestingly of the six conforming
columns 3 of the columns were packed with fully porousparticles and 3 columns with partially porous particles Itcan be seen however that when properly interpreted andadjusted all of Guiochonrsquos experimental data falls fairlywithin the norms outlined in the reference chart in Table 1In order to explore the reasons why the raw data does notconform more closely to the norms outlined in Table 1 eachof Guiochonrsquos studies included in Table 3 is evaluated in turn
Finally Table 4 herein is a list of symbols and parameterdefinitions used throughout this paper as well as their unitsof measure
91 Guiochonrsquos 1998 Study In a 1998 publication by Kohand Guiochon entitled ldquoEffect of the Column Length on theCharacteristics of the Packed Bed and the Column Efficiencyin a Dynamic Axial Compression Columnrdquo [19] Guiochonand his coauthor report the results of a study they hadcarried out on column packing using a rather novel techniquewhich was a combination of slurry packing and mechanicalcompression
The study involved the packing of a single cylinder ofdiameter 5 cm to various column lengths ranging between 1and 25 cm approximately The particles were all silica basedand coated with a reverse stationary phase C
18 However two
different categories of particles were involved One categorycontained highly irregular particles which were about 130micrometers in average diameter In addition these particleshad a high specific pore volume (11mLg) and exhibitedsignificant particle breakage under the packing conditionsinvestigated in the study Conversely the other particlecategory involved very small particles (6 micrometers) whichwere spherical in shape had a low specific pore volume(03mLg) and exhibited a greater ability to withstandparticle breakage under the studyrsquos packing conditions
The permeability data for the irregular particles isneglected herein because according to the authors theparticles exhibited significant breakage during packing andwithout an accurate value for the characteristic dimensionof the particle the data cannot be used in the Kozeny-Blake model to determine the value of 119875
0 The spherical
particles on the other hand behaved successfully under thepacking conditions chosen and yielded values calculated bythe authors for 119875
0of 619 268 226 and 229 for the four
different lengths of columns studied The data for thesecolumns is reported in our Table 3 in the rows for ColumnsB1through B
4 The higher value of 619 (the shortest column)
was rejected by the authors because they felt that either theparticles had been somewhat broken during packing or thecolumn frits had been blocked with fines Accordingly thepermeability data for this column is likewise neglected herein
The remaining three columns however were not con-sistent with respect to the authorsrsquo calculated values for 119875
0
Since the particles were identical the range of values for 1198750
of 226 to 268 must be due to experimental error The rawdata for column B
2 which produced a 119875
0value of 268 is
all self-consistent For example it generates a value for 119875119905
of 148 which is a reasonable value based upon the knownlow particle porosity for these Nova Pak particles and acorresponding value for 120576
0of 04073 which is a reasonable
value for a well-packed column However the other twocolumns columns B
3and B
4 had the much lower value of
126 for 119875119905 Accordingly in Table 3 herein corrected values
for column external porosity for columns B3and B
4are
displayed These corrected values are only slightly differentthan the reported values are within the experimental errorof the measurement and therefore in combination with the
Journal of Materials 15
Table 4 Glossary of terms
Symbol Unit dimensions (cgs) Definition119860 cm2 Column cross-sectional area119863 cm Column diameter119889119901
cm Particle spherical diameter equivalentΔ119875 gcmminus1secminus2 Pressure drop along column119889119901119898
cm Average particle diameter (measured)1198960
None Guiochonrsquos residual reciprocal permeability coefficient119875119905
None Guiochonrsquos modified permeability coefficient in Kozeny-Blake equation119871 cm Column length1198750
None Permeability coefficient in Kozeny-Blake equation119876 cm3minminus1 (exception) Volumetric fluid flow rate1199050
sec Time for elution of totally permeating unretained solute119896 cm2 General permeability coefficient in Darcy equation119870
119865cm2 Halaszrsquos permeability coefficient in Darcy equation (1974 paper)
119870 cm2 Halaszrsquos permeability coefficient in Darcy equation (1972 paper)119870
119888None Guiochonrsquos permeability coefficient in the Kozeny-Blake equation (as modified by Carmen)
Greek1205760
None External column porosity120576119894
None Internal column porosity120576119905
None Total column porosity120576119901
None Particle porosityΦ None Giddingsrsquo ratio of external and total column porosity120601 None Flow resistance parameter based on mobile phase velocity1206010
None Flow resistance parameter based on superficial velocityΩ
119901None Particle sphericity factor
120578 gcmminus1secminus1 Fluid absolute viscosity120583119894
cmsecminus1 Average fluid interstitial linear velocity120583119904
cmsecminus1 Average fluid superficial linear velocity120583119905
cmsecminus1 Average mobile phase velocityΨ None Porosity dependence term in the Kozeny-Blake equation based on mobile phase velocityΨ
0None Porosity dependence term in the Kozeny-Blake equation based on superficial velocity
raw data for column B2 are supportive of the value of 267 for
1198750
92 Guiochonrsquos 1999 Study In a 1999 publication by Farkaset al entitled ldquoValidity of Darcyrsquos Law at Low Flow Rates inLiquid Chromatographyrdquo [20] Guiochon and his coauthorsreport the results of some very exactingmeasurements whichthey made to validate Darcyrsquos law at extremely low fluidflow rates Using a column containing particles packed toa measured external column porosity of 0399 and havinga measured total column porosity of 0593 (Figure 2 in thepaper) Guiochon et al measured the column pressure dropand fluid flow rate at flow rates ranging between 0015 and05mLminwith ethylene glycol as the fluidThe authors statethat the particles in the column are spherical and they appearto have gone to extraordinary lengths to obtain an accuratemeasure of the particle diameter and of course as a result ofmodern techniques of particle size classification the particlesize distribution is very narrow
In Figure 1 of the paper the authors show a plot ofcolumn measured pressure drop in bar versus ldquofluid linearvelocityrdquo in cmsec This plot represents their measurementstaken in the empty column The only velocity that exists in
an empty column is the superficial velocity and thus theabscissa values in the plot in Figure 1 designated as ldquofluidlinear velocityrdquo represent superficial linear velocity
In Figure 2 in their paper however the authors showanother plot of measured pressure drop in bar also versusldquofluid linear velocityrdquo in cmsec The velocity on the abscissain this plot although also designated as ldquofluid linear velocityrdquodoes not equate (for permeability related purposes) to theldquofluid linear velocityrdquo in Figure 1 This is because in Figure 2the columnwas filled with porous particles and themeasuredvelocity was the mobile phase velocity Indeed the onlyvelocity used throughout the entire paper is the mobile phasevelocity (mobile phase velocity and superficial velocity areone and the same in an empty column)
In Table 1 of their paper the authors report that for thecolumn represented in Figure 2 the slope of a line achievedwhen the mobile phase velocityin units of cmsec is plottedversus pressure drop in units of bar is 1829 They also reportthat the flow resistance parameter 120601 for this column is 865In Table 3 herein it can be seen that the raw data for thiscolumn (Column C) generates an apparent value of 257 for1198750 This value is very close to Giddingsrsquo value of 267 Indeed
by referring to Figures 2 and 3 herein one can see that thereported raw data in this study without any modification
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
4 Journal of Materials
And now reconciling (1) and (3) we get
1198750(1 minus 120576
0)2
120583119904
120576301198892119901
=120583119905120576119905
1198960(119889
119901119898Ω
119901)2 (12)
4 The Flow Resistance Parameter
Before proceeding any further we digress to define anotherterm which will become important to our understanding ofGuiochonrsquos work The chromatographic literature containsa term called the ldquoflow resistance parameterrdquo [7] The fluidvelocity underlying this parameter is based upon Giddingsrsquodefinition of ldquomean fluid velocity through a cross-sectionrdquoand is found in his 1965 text book at page 207 [8]2 It is definedas follows
120601 =Δ119875119889
2
119901
120583119905120578119871
(13)
where 120601 = flow resistance parameterThe flow resistance parameter therefore is equivalent to
the reciprocal of Guiochonrsquos term 1198960
120601 =1
1198960
(14)
5 Refining the Terms For Column Porosity
In order to continue with our comparison of Guiochonrsquosteaching to that embodied in the Kozeny-Blake equation andin particular to accommodate columns packed with porousparticles we need to refine the porosity dependence term inthe Kozeny-Blake equation To accomplish this we adopt thedefinition of column porosity as taught by J Calvin Giddings[8] namely
Ψ =(1 minus 120576
119905Φ)
2
(120576119905Φ)
3 (15)
where Ψ = the porosity dependence term in Kozeny-Blake(porous particles) and
Φ =1205760
120576119905
(16)
where Φ = the ratio between the external and total columnporosities
Using thismethodology in themost general case involvingporous particles to express total column porosity (where totalcolumn porosity is not the same as external column porosity)as opposed to in the special case involving nonporous particles(where total column porosity and external column porosityare the same) the Kozeny-Blake equation (using mobilephase velocity as in (10)) may now be reexpressed as follows
Δ119875
119871=1198750(1 minus 120576
119905Φ)
2
120583119905120576119905120578
(120576119905Φ)
3
(119889119901)2
(17)
It will be appreciated that in the special case of columnspacked with nonporous particles (17) is identical to (3)because here 120576
119905= 120576
0and thus 120583
119905= 120583
119894= 120583
119904120576
0and Φ = 1
6 The Relationship between 1198960
and 1198750
We can now connect the teaching of Guiochon to thatembodied in the Kozeny-Blake equation as follows
1198750=
1
1198960Ψ
(18)
or
1198750=120601
Ψ (19)
As illustrated by (19) the term 1198750in the Kozeny-Blake
equation is the flow resistance parameter normalized for thecolumn porosity dependence term
Note that if1198750is a true constant the two parameters upon
which it depends 120601 and Ψ will vary in direct proportion toeach other This is illustrated by the data in Table 1 hereinwhich is a reference chart where particle porosity 120576
119901 is varied
from nonporous at one extreme to highly porous at the otherwhere 120576
119901= volume of the free space within the particles in
a packed columnthe total volume of all the particles in thecolumn
Figure 1(a) is a plot of the data contained in Table 1 as120601 versus Ψ Proceeding from the lower left hand corner tothe upper right hand corner of the plot particle porosityincreases It is obvious from the plot that all the data in Table 1falls on a single line with a slope of 267 which we say is theconstant value of 119875
0 As is illustrated by the plot in the range
of porosity covered by the data the values for 120601 vary fromabout 500 to about 1900 and correspondingly the values ofΨvary from approximately 2 to approximately 7 Note also thatthe upper boundary value of Ψ for packed columns is about55 with larger values related to capillaries
Finally we point out that once the value of 1198750is correctly
identified ameasured value for the flow resistance parametermay then be used to calculate the value of the externalporosity of the column 120576
0 when the latterrsquos value has not or
cannot be measured
7 Guiochonrsquos Modified PermeabilityCoefficient 119875
119905
As reflected in some of his recent publications which wewill explore in more depth later Guiochon inadvertentlyapparently confuses the term 119875
0with another parameter 119875
119905
which we identify here and define for the first time 119875119905is not
the same as 1198750because it is based upon Guiochonrsquos unique
(from a permeability perspective) teaching for fluid velocityIt is defined as follows
119875119905= 119875
0120576119905 (20)
where 119875119905= Guiochonrsquos modified permeability coefficient
It will be appreciated that Guiochonrsquos 119875119905 if used in the
Kozeny-Blake equation in place of 1198750 accommodates his use
of themobile phase velocity in away that is different from butequivalent to a direct conversion of mobile phase velocity tosuperficial velocity
Journal of Materials 5
Table1Colum
nperm
eabilityreferencetable
119870119865
Particlepo
rosity
rang
e120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119896119871
119863Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
120576119905Φ
(120576119905minus1205760)
120576119894
Δ119875119889119901
2Δ119875119889119901
2(1minus120576119905Φ)2120576119905
(1minus1205760)2
1120601
1198750120576119905
119889119901
2120576119905
(119889119901119898Ω119901)
(1minus1205760)
120583119905120578119871
120583119904120578119871
(120576119905Φ)3
(1205760)3
120601Ψ
120601po
iseUnits
Non
eNon
eNon
eNon
eNon
epsi
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
cmNon
ecm
cmgcmminus1 secminus1
cm3 m
inminus1
cmsecminus
1cm
secminus
1cm
secminus
1
Non
porous
particles
0380
1000
0380
000
0000
0160
711
1870
2662
701
141119864minus03
267
101
535119864minus10
15046
100
000100
000100
00100
038
0038
0100
0100
0390
1000
0390
000
0000
0147
653
1675
244
6627
153119864minus03
267
104
597119864minus10
15046
100
000100
000100
00100
039
0039
0100
0100
040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
100
000100
000100
00100
040
004
00100
0100
0410
1000
0410
000
0000
0124
553
1349
2071
505
181119864minus03
267
109
742119864minus10
15046
100
000100
000100
00100
041
004
10100
0100
0420
1000
0420
000
0000
0115
509
1212
1907
454
196119864minus03
267
112
825119864minus10
15046
100
000100
000100
00100
042
004
20100
0100
Lowpo
rosity
particles
0507
0750
0380
0127
0204
213
948
1870
3549
701
106119864minus03
267
135
535119864minus10
15046
100
000100
000100
00100
051
0051
0133
0100
0520
0750
0390
0130
0213
196
871
1675
3261
627
115119864minus03
267
139
597119864minus10
15046
100
000100
000100
00100
052
0052
0133
0100
0533
0750
040
00133
0222
180
801
1502
2999
563
125119864minus03
267
142
666119864minus10
15046
100
000100
000100
00100
053
0053
0133
0100
0546
0750
0410
0136
0231
166
737
1349
2760
505
136119864minus03
267
146
742119864minus10
15046
100
000100
000100
00100
055
0055
0133
0100
0560
0750
0420
0140
0241
153
679
1212
2542
454
147119864minus03
267
149
825119864minus10
15046
100
000100
000100
00100
056
0056
0133
0100
Medium
porosity
particles
0613
0620
0380
0233
0376
258
1146
1870
4293
701
872119864minus04
267
164
535119864minus10
15046
100
000100
000100
00100
061
0061
0161
0100
0629
0620
0390
0239
0392
237
1053
1675
3946
627
949119864minus04
267
168
597119864minus10
15046
100
000100
000100
00100
063
0063
0161
0100
064
50620
040
00245
040
9218
969
1502
3629
563
103119864minus03
267
172
666119864minus10
15046
100
000100
000100
00100
064
0065
0161
0100
0661
0620
0410
0251
0426
201
892
1349
3340
505
112119864minus03
267
177
742119864minus10
15046
100
000100
000100
00100
066
006
60161
0100
0677
0620
0420
0257
044
4185
821
1212
3076
454
122119864minus03
267
181
825119864minus10
15046
100
000100
000100
00100
068
006
80161
0100
Highpo
rosity
particles
0760
0500
0380
0380
0613
320
1422
1870
5325
701
703119864minus04
267
203
535119864minus10
15046
100
000100
000100
00100
076
0076
0200
0100
0780
0500
0390
0390
0639
294
1306
1675
4891
627
766119864minus04
267
208
597119864minus10
15046
100
000100
000100
00100
078
0078
0200
0100
0800
0500
040
0040
0066
7270
1202
1502
4500
563
832119864minus04
267
214
666119864minus10
15046
100
000100
000100
00100
080
0080
0200
0100
0820
0500
0410
0410
0694
249
1105
1349
4140
505
905119864minus04
267
219
742119864minus10
15046
100
000100
000100
00100
082
0082
0200
0100
0840
0500
0420
0420
0724
229
1018
1212
3814
454
982119864minus04
267
224
825119864minus10
15046
100
000100
000100
00100
084
0084
0200
0100
Capillary
1000
0380
0380
0620
1000
421
1870
1870
7005
701
535119864minus04
267
267
535119864minus10
15046
100
000100
000100
00100
100
0100
0263
0100
1000
0390
0390
0610
1000
377
1675
1675
6273
627
597119864minus04
267
267
597119864minus10
15046
100
000100
000100
00100
100
0100
0256
0100
1000
040
0040
0060
010
00338
1502
1502
5625
563
666119864minus04
267
267
666119864minus10
15046
100
000100
000100
00100
100
0100
0250
0100
1000
0410
0410
0590
1000
303
1349
1349
5051
505
742119864minus04
267
267
742119864minus10
15046
100
000100
000100
00100
100
0100
0244
0100
1000
0420
0420
0580
1000
273
1212
1212
4541
454
825119864minus04
267
267
825119864minus10
15046
100
000100
000100
00100
100
0100
0238
0100
6 Journal of Materials
0200400600800
100012001400160018002000
20 30 40 50 60 70
NonporousLow porosityMedium porosity
CapillaryLinear (line)
Nonporous Low porosity particles
Medium porosity particles
High porosity particles
Capillary
Line slope (P0) = 267
y = 267x + 0
Ψ
120601
Line
High porosity
(a)
99111123135147159171183195207219231243255267
035 045 055 065 075 085 095
LineCapillaryLinear (line)
Capillary
Line slope (P0) = 267
y = 267x + 0
Pt
120576t
Φ = 10
Φ = 075
Φ = 062
Φ = 05
Φ = 075
Φ = 062
Φ = 05
Φ = 1
(b)
99111123135147159171183195207219231243255267
000 010 020 030 040 050 060 070 080 090 100
NonporousLow porosityMedium porosityHigh porosity
120576p
y = 160x + 107
Pt
LineCapillaryLinear (line)
(c)
Figure 1 (a) The balance between 120601 and Ψ (b) Column permeability reference chart (c) Particle permeability reference chart
In the 2006 edition of their textbook [9] Guiochon andhis coauthors provide the following worked example of hispressure dropfluid flow teaching3
In Table 2 a cross-reference is established between Guio-chonrsquos teaching for columns packed with both porous andnonporous particles the Kozeny-Blake equation (asmodifiedby Carman) and for validation purposes Giddingsrsquo teachingbased upon actual permeabilitymeasurements fromhis Table53-1 in his 1965 text [8]4
In his worked example Guiochonrsquos fluid velocity 119906is found in his text as his equation (256) [9] at page61 and equates to mobile phase not superficial velocityAccordingly when applied to columns packed with porousparticles his worked example combines the contributionto fluid velocity of mass transfer by molecular diffusionin the free space within the particles with mass transferby convection in the free space between the particles Asdiscussed above if one is attempting to compare Guiochonrsquos
teaching for the value of the permeability coefficient with thatof the Kozeny-Blake equation the conversion factor is thetotal porosity of the column
In addition because Guiochonrsquos particle sizeparameter has an embedded assumption due to particleshaperoughness his worked example also commingles thecontribution to spherical particle diameter equivalent ofparticle shaperoughness with other nonparticle variablesUnfortunately his worked example is not based uponcompletely spherical particles where this would not be anissue We know this because he states that 119896
0in his worked
example equals 1 times 10minus3 which he says is the value of thepermeability coefficient for columns packed with irregularparticles
In order to move forward one must simplify If this wereto be done experimentally it would best be accomplishedby experiments with nonporous particles in which case thevariable of molecular diffusion in the pores of the particles
Journal of Materials 7
Table2Cr
oss-referencefor
term
s119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119896119871
119863Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
120576119905Φ
(120576119905minus1205760)
120576119894
Δ119875119889119901
2Δ119875119889119901
2(1minus120576119905Φ)2120576119905(1minus1205760)2
1120601
1198750120576119905
119889119901
2120576119905
(119889119901119898Ω119901)
(1minus1205760)
120583119905120578119871
120583119904120578119871
(120576119905Φ)3
(1205760)3
120601Ψ
120601po
iseUnits
Non
eNon
eNon
eNon
eNon
epsi
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
cmNon
ecm
cmgcmminus1 secminus1cm
3minminus1cm
secminus
1cm
secminus
1cm
secminus
1
Nonporousp
articles
Gidding
srsquotradition
alno
nporou
scolum
n040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
1000010
00010
0010
0399
004
00100
0100
Guiocho
nrsquostradition
alspheric
alparticle(app
arent)
0362
1000
0362
000
0000
0187
830
2293
3106
858
120119864minus03
267
97436119864minus10
15046
1000010
00010
0010
0361
0036
0100
0100
Gidd
ingrsquos
table5
3-1
5060meshglassb
eads
0422
1000
0422
000
0000
0113
500
1184
1873
444
200119864minus03
267
113
844119864minus10
15046
1000010
00010
0010
0421
004
20100
0100
5060meshglassb
eads
040
910
00040
9000
0000
0126
560
1371
2097
513
179119864minus03
267
109
729119864minus10
15046
1000010
00010
0010
040
70041
0100
0100
Guiochonrsquosteaching
ExA
irregular
particle
(app
arent)
040
010
00040
0000
0000
0225
1000
2500
2250
563
100119864minus03
444
178
400119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
ExA
correctedforp
article
irregularity
040
010
00040
0000
0000
0225
601
1502
2250
563
166119864minus03
267
107
400119864minus10
15046
078
00010
000
080010
0399
004
00100
0100
ExB
spheric
alparticle
(app
arent)
040
010
00040
0000
0000
0187
830
2075
2250
563
120119864minus03
369
148
482119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
ExB
corrected
040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
Porous
particles
Gidding
srsquotradition
alpo
rous
column
060
0066
7040
00200
0333
203
900
1500
3375
563
111119864minus03
267
160
667119864minus10
15046
100
00010
00010
0010
0599
006
00150
0100
Gidd
ingrsquos
table5
3-1
3040meshalum
ina
0803
0498
040
00403
0672
271
1204
1499
4510
562
830119864minus04
267
214
667119864minus10
15046
100
00010
00010
0010
0801
0080
0201
0100
5060meshalum
ina
0837
0499
0417
0420
0721
235
1043
1246
3906
467
959119864minus04
267
224
803119864minus10
15046
100
00010
00010
0010
0835
0084
0201
0100
6080meshchromosorbW
(5DNP)
0766
0498
0382
0384
0621
316
1404
1833
5259
687
712119864minus04
267
204
545119864minus10
15046
100
00010
00010
0010
0764
0077
0201
0100
6080meshchromosorbW
(20
DNP)
0785
0495
0389
0396
064
8300
1333
1698
4991
636
750119864minus04
267
210
589119864minus10
15046
100
00010
00010
0010
0783
0079
0202
0100
8 Journal of Materials
would be eliminated and by using only smooth sphericalparticles such as glass beads in which case the variableof particle shaperoughness would be eliminated If both ofthese things were done one would be able to directly andunambiguously determine the true value of 119875
05
Accordingly since his textbook contains a general teach-ing without any expressed restrictions with respect to particleporosity let us first assume that the column in Guiochonrsquosworked example is packed with nonporous particles Sec-ondly let us assume that the value for the external porosityof the column (which is the same as the total porosity ina column packed with nonporous particles) is the valuetraditionally assumed for a typical well-packed column thatis 04
As illustrated by example A in Table 2 this would leadto an apparent value for 119875
0of 444 which is obviously much
too high6 Using the adjusted particle size of 8 micrometers(rounded) which is calculated based on a value of 078 forΩ
119901
we derive values of 267 for 1198750and 107 for 119875
119905(the relationship
between these two values is of course equal to the totalporosity of the column 120576
119905 as is dictated by (20))
Similarly as illustrated by example B in Table 2 when thecolumn of Guiochonrsquos worked example is adapted to accom-modate spherical nonporous particles Guiochonrsquos teachingof the value for 119896
0of 12 times 10minus3 results in an apparent value
of 369 for 1198750
7 again too high [10 11] On the other hand ifwe correct the value of 119875
0 we get a column external porosity
of 03620which is too low for a typical well-packed column(120576
0= 04) [8] Moreover in the case of both corrections
Guiochonrsquos teaching overstates the pressure gradient (187 psicompared to the 135 psi which Giddingsrsquo experiments wouldgenerate) Accordingly it is only when we correct for thepressure drop that we achieve appropriate values of 04 for1205760 267 for 119875
0 and 107 for 119875
119905
As is evident from Table 2 when the porosity of columnspacked with nonporous particles varies the value of 119875
119905also
varies because it is based upon the measurement of mobilephase velocity which in turn changeswith the porosity of thecolumn Indeed for typical columns packed with nonporousparticles having external porosities close to 04 (038ndash042)[8] Table 2 reveals that the value of 119875
119905varies by as much as 6
points while the value of 1198750remains constant at 267
Having established a range of values for 119875119905based upon
columns packed with nonporous particles one can nowproceed to evaluate the range of values for 119875
119905based upon
columns packed with porous particles which is to say thatwe can now evaluate the impact of particle porosity onpermeability Among other things Table 2 shows that whenit comes to porous particles Guiochonrsquos 119875
119905is even more
variable (and thus even less useful) because the value ofthe coefficient can differ on a column-to-column basis byas much as 90 points depending on the combination of theinternal andexternal column porosities at play
Figures 1(a) and 1(b) are a plot of the data found in Table 1In this plot Guiochonrsquos modified permeability coefficient 119875
119905
is plotted against column total porosity 120576119905 Once again it will
be appreciated that (a) all data falls on a line with a slope of267 which represents the value of 119875
0and (b) the line has no
intercept which means that when 120576119905is zero 119875
119905is also zero As
shown in the plot values of the parameter Φ are highest atlow values of 119875
119905and lowest at high values of 119875
119905 Further as is
evidenced by the respective definitions for the terms119875119905and119875
0
contained herein the only time that the value of Guiochonrsquos119875119905does not differ from the value of 119875
0(267) is when the value
of 120576119905is 1 its value in a theoretical packed column devoid of
particles that is acapillaryFinally Figure 1(c) is another plot of the data in Table 1
but this time the values for 119875119905are plotted against particle
porosity 120576119901 The equation of this line does have an intercept
which represents the value of 119875119905when the particles are
nonporous It is also obvious from the equation of this linethat it is only when particle porosity is unity that is in acapillary that the value of 119875
119905is 267
It will be appreciated that since 120576119905is the sumof the internal
and external porosities of a packed column any given valuefor 119875
119905involving a column packed with porous particles can
represent many different combinations of these parametersVice versa any given value for 120576
119905can generate many different
values for 119875119905depending on the amount of particle porosity
which is present On the other hand as is pointed outearlier column internal porosity andor particle porosityplay no role in determining the permeability of a packedcolumn the only porosity parameter which is relevant forsuch purposes is column external porosity In other words asa permeability coefficient 119875
119905is not a particularly informative
concept because it still fails to provide any meaningful tie tothe underlying physical phenomena that govern the pressuredropfluid flow relationship
In the case of permeability measurements one quan-tifies only total pressure drop Pressure drop however isa commingling of the contributions of the several differ-ent parameters embodied in the pressureflow relationshipTherefore one must accurately represent the contributionof each factor before adding them together to equal thepressure drop Obviously if one mistakes the contribution ofany given element for instance particle size this will skewthe contribution of some other element when summed forinstance column porosity
Our frame of reference has been calibratedwith respect tothe interrelationships between the values of 119875
0 119875
119905 120576
119905 and 120576
119901
based on the data reported in Giddingsrsquo table 53-1 in his 1965textbook Since his data base was derived from experimentsusing nonporous particles the permeability results of whichwere extended to columns packed with fully porous particlesby using his Φ parameter the validity of which in turn wasestablished by independent means there is no uncertaintywith respect to the critical variable of external porosity 120576
0
inherent in our frame of reference It is for this reasontherefore that our frame of reference trumps any reporteddata which uses any other technique including the techniqueof inverse size exclusion chromatography by itself to establishvalues for 120576
0[12] In addition our frame of reference is
internally consistent due to the grounding of the parametersat the boundary condition of a theoretical capillary at whichpoint the values of 120576
119905and 120576
119901are unity and the values of 119875
0and
119875119905are equal In essence these interrelationships exemplify the
conservation laws of nature which are the ultimate standard
Journal of Materials 9
of measure when considering partial fractions of a singleentity
Consequently as Figures 1(b) and 1(c) illustrate if oneknows based upon other independent means the value of acolumnrsquos total porosity or its particlesrsquo porosity respectivelyone can at least verify whether results of permeability exper-iments expressed in terms of 119875
119905are reasonable Or to put
it another way if the 119875119905results reported do not conform to
what these plots tell us one should expect from columns andparticles based upon independently derived porosity valuesfor the particles under study we know that something iswrong Accordingly this is the way in which we use 119875
119905below
to analyze and evaluate Guiochonrsquos experimental work of thelast decade
8 The Origin of Guiochonrsquos Value forColumn Permeability
In 1966 Guiochon wrote a review paper entitled ldquoProblemsRaised by the Operation of Gas Chromatographic Columnsrdquo[13] In that paper he reviewed the development of the per-meability equation from Darcy to the then present Amongother things he reported that ldquoa mean 120576 of 038 is foundfor a number of packings obtained from powdered prod-ucts of various origins consisting of spherical or irregularparticles Almost all the measured values are between035 and 045rdquo (p 14) He then went on to discuss what hecalled the ldquosemiempirical Blake-Kozeny equationrdquo which hereported as 119896 = [119889
2
1199011205762][150(1 minus 120576)
2] [13 p 15 Equation
(11)] What sticks out of course is the fact that Guiochonstates here that 119875
0equals 150 the value for the permeability
coefficient championed by Ergun not the value posited byCarman To be fair we should point out that Guiochongave warning that he was uncomfortable with any particularvalue for the permeability coefficient For even though herecited the Kozeny-Blake equation with a value of 150 for itspermeability coefficient he stated that ldquothe values of 119896 givenby (this) equation are not very accurate and the deviationscan be as large as 50rdquo [13 p 15] Guiochon attributedmost ofthis uncertainty to the inability of practitioners of that era toeffectively measure the value of a columnrsquos external porositySince as he pointed out ldquo(the Kozeny-Blake) equation isvery sensitive to variations in 120576rdquo he concluded that theequation is ldquoof little userdquo and that it was better practice torely upon grosser measurements of permeability [13 p 15]Thus Guiochon at this timeframe made a conscious decisionto staywithin the relatively safe but admittedlymore obscureboundaries of Darcyrsquos teaching on column permeability
In 2006 however Guiochon emerged from under theshadow of Darcy permeability when he published a reviewpaper which purported to be a comprehensive summaryof the current state of the art in chromatography [14]In light of the fact that Guiochon and his coauthors hadvirtually contemporaneously published a new version of theirtextbook one would have expected that Guiochon wouldtake this opportunity to reaffirm his textbook teaching So itshould come as no surprise that he would open his discussionof the pressure dropfluid flow relationshipwith the following
assertion ldquothe permeability of packed beds used in liquidchromatography is in the order of 1 times 10minus3rdquo [14 p 9] Asone would expect the authority which Guiochon cites forthis seemingly familiar proposition is the 2006 edition of hisown textbook For the first time however he also providessome citations to third party sources In particular he cites a1997 chromatography handbook written by Uwe Neue ChiefScientist for Waters Corporation [15 p 9] As one can seeNeuersquos equationmdashlike Guiochonrsquosmdashdoes indeed postulate apermeability coefficient with a value of 1 times 10minus3
At this point however Guiochon abandons his tradi-tional caution about the value of the permeability coeffi-cient and marries his permeability equation to the Kozeny-Blake equation thereby endorsing not only the porositydependence term in that equation but also a specific valuefor 119875
0 Referring to his value of 1 times 10minus3 he states the
following ldquothe specific column permeability is related to thecolumn porosity through the Kozeny-Carman equation 119896
119891=
12057630180 (1 minus 120576
0)2rdquo [14 p 9] Not surprisingly Neue makes
the same connection ldquothe specific permeability depends onthe particle size 119889
119901and the interstitial porosity 120576
0of the
packed bedThe relationship is known as theKozeny-Carmanequation 119861
0= (1185)(1205763
01198892119901(1 minus 120576
0)2)rdquo [15 p 30]
Neuersquos permeability equation and Guiochonrsquos textbookpermeability equation however differ in that Neue uses thesuperficial not the mobile phase velocity Accordingly sinceall other features of their equations are the same one wouldexpect them to get different values for their permeabilitycoefficients Neue himself recognizes this when he says inhis handbook ldquothis value [1000] is also in good agreementwith our experience but values as low as 500 have beenreported in the literature8 However there are several reasonswhy different values reported by different workers do notrepresent true differences in the resistance parameter Firstlower values are obtained for porous packings if the specificpermeability is measured on the basis of the breakthroughtime of an unretained peak instead of using the flow raterdquo see[15 p 32] (when he distinguishes the ldquobreakthrough time ofan unretained peakrdquo from the ldquoflow raterdquo Neue is of coursereferring to the mobile phase velocity and the superficialvelocity resp) Consequently if one takes Neuersquos point andturns it on its head one can see that if he and Guiochonwere to get the same value for their permeability coefficients[1000] but one (Guiochon) is using mobile phase velocityto derive it and the other (Neue) uses superficial velocity toderive it this would indeed represent a ldquotrue differencerdquo intheir equations
So how can Guiochon cite both Neue and himself for theproposition in his review paper that ldquothe permeability of thepacked beds used in liquid chromatography is in the order of1times 10minus3rdquoThe answer is that at least for purposes of his reviewpaper he adopts Kozeny-Blakersquos fluid velocity convention inplace of the velocity frame he uses in his textbook Here iswhat he says ldquothis approximation [1 times 10minus3] is valid only ifthe superficial velocity is usedrdquo [14 p 9] (emphasis supplied)
As a matter of pure mathematics Guiochonrsquos permeabil-ity coefficient of 1 times 10minus3 can only be equal to 1205763
0180(1 minus 120576
0)2
if the value of 1205760is 04 the typical value for the interstitial
10 Journal of Materials
fraction of a well-packed column Guiochon confirms in hisreview paper that this is indeed what he assumes in his state-ment of the Kozeny-Blake equation [14 p 9] Accordinglyhis adoption of the value of 180 for 119875
0solely depends on the
validity of his assertion that the value of the permeabilitycoefficient is 1 times 10minus3
We already know from our discussion of Guiochonrsquostextbook teaching that when this value is associated with apermeability equation based on mobile phase velocity it iswrong As reflected in our Table 2 when Guiochonrsquos workedexample involving this figure is corrected to account for theembedded factor due to the irregularity of the particles whichhis worked example assumes the value of the permeabilitycoefficient generated by his equation is not 1 times 10minus3 it is 166times 10minus3
So where did this value of 1 times 10minus3 come from AlthoughGuiochon in his review paper cites Neuersquos handbook for thisvalue Neue himself offers absolutely no authority for it Onthe other hand Guiochon does provide one other citation inhis review paper for this value The citation is to an article byHalasz et al published in 1975 [16]
If one then goes to that article one sees that Halaszet al do indeed proffer the same equation as Neue (albeitin a somewhat different format) and yes it does containthe value of 1 times 10minus3 See [16 Equation (1)] Neverthelessthe authors fail to identify from whence they obtained thisvalue On the other hand they do make reference to apaper that Endele et al wrote a year earlier in 1974 withKlaus Unger entitled ldquoInfluence of the Particle Size (5ndash35 120583)of Spherical Silica on Column Efficiencies in High-PressureLiquid Chromatographyrdquo [17] It appears that this is wherethe body is buried for it is in this article that Halasz reportshaving actually derived a value of 1 times 10minus3 for the permeabilitycoefficient
In summarizing the results of their experiments whichincidentally are based on measurements of superficial notmobile phase velocity (see their Equation (7)) Halasz et alcalculate an average value for the permeability coefficient(which they notate as119870
119865) of 1198892
1199011080 Accordingly they state
the following ldquo[I]t is proposed to define an average particlesize 119889
119901 for columns packed with spherical silica using the
balanced density packing method by the equation 1198892119901
=
1198701198651000 = 1198651205781198711031199032120587Δ119875rdquo [17 p 382 their Equation (7)]To recapitulate it thus appears that (1) Halasz was the
source of the value of 1 times 10minus3 for the permeability coefficientwhich Guiochon adopts in his review paper (2)Halaszrsquo valuewas based upon measurements which involved sphericalparticles and (3)Halaszrsquo value was properly derived from anequation which utilized the superficial velocity as the fluidvelocity
However we still have a problem if we are going to acceptthe value of 1 times 10minus3 as a basis for calculating the constantin Kozeny-Blake we still need to know the value of Halaszrsquocolumnsrsquo external porosity For example in the absenceof measured values for the external porosity Guiochonrsquosassertion in his review paper that the value of the constantis 180 is without foundation
Unfortunately however Halasz bases his methodologyon ldquoassumedrdquo versus ldquomeasuredrdquo values for external porosityNevertheless he does provide some clues that may helpus peel back this onion First he says that his value forthe permeability coefficient applies to ldquocolumns using thebalanced density packing methodrdquo Secondly after statinghis proposed value for the coefficient we find the followingtelltale sentence ldquothe permeability of columns packed by theconventional ldquodryrdquo method are worse by a factor of about 2than those packed by the balanced density methodrdquo [17 p382] For this proposition Halasz cites yet another of his ownarticles this time a paper written by himself and M Naefein 1972 entitled ldquoInfluence of Column Parameters on PeakBroadening in High-Pressure Liquid Chromatographyrdquo [18]
When one follows this thread by going to Halaszrsquos 1972paper one finds a possible solution to the puzzle For this iswhere Halasz describes his calculations of the permeabilitycoefficient from experiments with a number of conventionaldry-packed columns containing spherical silica particlesNot surprisingly the resulting value for the permeabilitycoefficient for such columns was not 1 times 10minus3 Instead Halaszsays that his experiments with such columns resulted inthe following value for the permeability coefficient ldquo119870 sim
11988921199012000rdquo [18 p 78] In other words consistent with what
Halasz says in his 1974 article the value for the coefficient forthese dry-packed columns was indeed ldquoworse by a factor ofabout 2rdquo compared to its value for columns prepared with thebalanced density method
The obvious implication of this is that when Halasz gota value of 1 times 10minus3 for the permeability coefficient in 1974he must have been evaluating columns that were packed toan interstitial fraction considerably higher than the externalporosity of the columns that were the subject of his 1972experiments Our problem however is still not resolvedbecause Halasz did not accurately determine the externalporosity of the columns under study in either the 1972 or 1974columns
We surmise that in their 1972 paper Halasz and hiscoauthors made use of the teaching of Giddings outlinedin his 1965 textbook which at the time of their 1972 paperwas seven years old Here is what they say ldquoexperiencehas shown that in a regular packed column with the sievefractions usual for chromatography 120576
0is independent of the
particle size if the particles are more or less spherical In anacceptable approximation 120576
0is equal to 04 In those columns
packed with nonporous supports 119906V and 119906 are identicalUsing porous supports (ie silica gel alumina chromosorbPorasil and so forth) 119906 = 4119865(1205871198892
119888120576119879) where 120576
119879is the
total porosity and indicates the pore volume of the supportrdquoIt will be appreciated that 119906V stands for interstitial velocity(our 120583
119894) and 119906 stands for mobile phase velocity (our 120583
119905) The
ldquoexperiencerdquo which Halasz and his coauthors are referringto concerning ldquocolumns packed with nonporous supportsalumina and chromosorbrdquo is that recorded by Giddings inTable 53-1 in his 1965 text Continuing to establish theinterrelationships inherent in the various entities involvedin chromatographic columns with respect to both interstitialandmobile phase velocity the authors conclude ldquoformally 119906V
Journal of Materials 11
can be calculated for a porous support assuming 1205760is equal
to 04rdquo (emphasis supplied) The authors then announce themethodology underlying their frame of reference with regardto column permeability as follows ldquowe determined in all ofour columns the 119906119906V ratio to be 056 with a differentialrefractometer Consequently the total porosity 120576
119905was equal
to 0714rdquo It will be appreciated that their ratio of 119906119906V is oneand the same as GiddingsrsquoΦ parameter as utilized by him inhis 1965 textbook and as we define hereinabove
We can now identify the origin of the discrepancybetween Halasz and Giddings and by extension betweenHalasz and Guiochon (and Neue) For even though Halaszmakes use of the results in Giddingsrsquo Table 53-1 he failsto appreciate the teaching inherent therein regarding theimportance of an accurate value for column external porositywhen using porous particles In establishing his values forΦGiddings was careful to ground his values for the externalporosity of columns packed with porous particles in acomparison of their measured pressure drops with measuredpressure drops in columns packed with nonporous particleswhere the values for (external) porositywere accurate becausethey are equal to the columnsrsquo total porosity
In their 1972 paper Halasz et al used a refractometer todetect the inert solute 119899-pentane in the mobile phase of 119899-heptaneThey injected the 119899-pentane into the flowing mobilephase the flow rate of which they presumably measured witha volumetric cylinder and stop watch Since the exit timeof the 119899-pentane peak was identified by the refractometertheir measurement of holdup volume was accurate andaccordingly so was their value of 0714 for 120576
119905(there is no
uncertainty related to column total porosity when using inertsolutes)This in turn led to an accurate value for their119906 termmobile phase velocity However since they did not measurethe interstitial velocity 119906V but rather used an estimate basedupon the measured flow rate and an ldquoassumedrdquo value of 04for external porosity their calculatedestimated value for 119906Vwas arbitrary This in turn means that the value of 056 fortheirΦ ratio was likewise arbitrary
In Table 3 herein we show the results for column number1 in the 1972 study which we designate as Column A
1 Note
that the authorsrsquo raw values correspond to a value of 500 for1198750 a valuewhich is obviously far too high As is plainly evident
in our Figures 2 and 3 this column data set does not conformin any reasonable fashion to the norms in our referencecharts Accordingly the data is badly flawed In our correcteddata for Column A
1in Table 3 herein we adjust the column
external porosity to the lowest value permitted in our frame ofreference (038) This correction results in a revised value forGiddingsrsquoΦparameter of 0532 In additionwe are compelledto adjust the particle downward (to 11 micrometers approx)as is also dictated by our frame of reference We justify ourlow corrected value for external porosity and our downwardadjustment for particle size on the aggressive ldquodry-packingrdquoprocess described by the authors ldquo[T] the best results wereachieved with the following method the column (id =2mm) 25 cm long was packed with 25 equal portions of theldquobrushrdquo After adding each portion the column was vibratedand afterward tapped on the floor and also vigorously tappedwith a rod from the siderdquo Based upon a column packing
experience of the principal author herein spanningmore than30 years we believe that these packing conditions generateda packed column with a low external porosity and also one inwhich the particles experienced significant breakage
The main differences between the columns in the 1972and 1974 papers concerned the particle sizes and themethodsused to pack the columns The particle diameters in the 1972study fell in a range of about 15ndash200 micrometers while thosein the 1974 study were much smaller being in the range of 5ndash40 micrometers approximately Moreover while the columnsin the 1972 study were dry-packed the columns in the 1974study were slurry-packed (what Halasz calls ldquothe balanceddensity packing methodrdquo) The authors of the 1974 paperrecite that their permeability results shown in their Table 4correspond to values for 120601
0of 1080 and values for 120601 of 910
In Table 3 herein we show the results for the 1974 study as anaverage value in our column designated as A
2 Note that the
raw data reported by the authors which again has the built-inassumption that the column external porosity was 04 placesthe value of 119875
0at 192 In our corrected values for Column A
2
we show that a value of 267 for 1198750places the column external
porosity at a value of 043 The ratio of the value for 120601 of 910when compared to the value for 120601 in the 1972 study of 2007is about 22 (2007910 = 22) which is in good agreementwith the authorsrsquo statement that the permeability of the 1972columns was worse by a factor of about 2 Again based onthe experience of the principal author herein in the packingof hundreds of thousands of commercially sold slurry packedcolumns we conclude that the balanced density techniquedescribed by the authors in the 1974 paper is fully capable ofproducing columns with this high and external porosity
In summary we conclude that for purposes of theKozeny-Blake equation (a) the external porosity of theHalasz 1972 columns was 038 (b) the external porosity of thecolumns which he tested in 1974 was 043 and (c) Halaszrsquos1972 and 1974 studies when properly analyzed corroborateGiddingsrsquo value for 119875
0of 267 not Carmanrsquos value of 180
Although the fault for all of the misdirection about the valueof the permeability coefficient being 1 times 10minus3 therefore restsprimarily on the shoulders of Halasz one has to question whyGuiochon (andNeue) so readily adopted itWhatevermay bethe reason Guiochonrsquos support for this proposition has beenand continues to be a cause for much of the conventionalacceptance of Carmanrsquos (erroneous) figure of 180 as thecharacteristic value of the coefficient in the Kozeny-Blakeequation
9 Guiochonrsquos Experimental Data
In addition to the data from Halasz 1972 and 1974 studiesTable 3 herein contains data from a representative sampleof Guiochonrsquos personal experimental permeability investi-gations over approximately the last decade as reported inthe numerous trade journals particularly the Journal ofChromatography A In each case the table has a single rowdedicated to Guiochonrsquos reported raw data and a single rowdedicated to corrected values for each column based on thereference chart in Table 1 herein
12 Journal of MaterialsTa
ble3Summaryexperim
entald
ata
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Naefe
andHalasz
columns
1198601(19
72)
Colum
nA1rawdata
07140
0560
040
000314
0523
782007
2811
4016
563
498119864minus04
500
357
908119864minus10
648119864minus10
2500
020
100
000135
0001350
00038
100
053
132
074
Colum
nA1corrected
07140
0532
03800
0334
0539
7813
3518
705002
701
749119864minus04
267
191
908119864minus10
648119864minus10
2500
020
100
000110
000110
100038
100
053
139
074
Endelean
dHalasz
columns
1198602(19
74)
Colum
nA2rawdata
08426
0475
040
00044
30738
11910
1080
4740
563
110119864minus03
192
162
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
033
016
Colum
nA2corrected
08426
0511
04309
0412
0723
11910
1080
3411
405
110119864minus03
267
225
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
031
016
Guiochon
columns
1198612ndash1198614(19
98)
Colum
nB 2
rawdata
05639
0722
040
730157
0264
22771
1367
2931
520
130119864minus03
263
148
467119864minus10
263119864minus10
429
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 3
rawdata
05582
0704
03932
0165
0272
44764
1369
3382
606
131119864minus03
226
126
471119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 4
rawdata
05509
0710
03909
0160
0263
72784
1422
3422
621
128119864minus03
229
126
459119864minus10
253119864minus10
1328
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 3
corrected
05653
0723
040
860157
0265
44774
1369
2899
513
129119864minus03
267
151
465119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 4
corrected
05651
0723
040
830157
0265
72776
1374
2907
514
129119864minus03
267
151
464119864minus10
262119864minus10
1375
500
100
000
060
000
0600
00165
98008
020
015
Guiochon
columnC
(1999)
Colum
nCrawdata
05930
0673
03990
0194
0323
183
865
1459
3372
569
116119864minus03
257
152
109119864minus09
645119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Colum
nCcorrected
05930
0673
03990
0194
0323
190
900
1518
3372
569
111119864minus03
267
158
105119864minus09
620119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Guiochon
columns
1198631ndash1198636(2006)
Colum
nD
1rawdata
07830
0456
03570
0426
0663
86694
886
7115
909
144119864minus03
975
76334119864minus10
261119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
013
Colum
nD
2rawdata
07230
0491
03550
0368
0571
88656
907
6726
930
152119864minus03
975
70353119864minus10
255119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
014
Colum
nD
3rawdata
06850
0506
03466
0338
0518
97685
1000
7025
1026
146119864minus03
975
67338119864minus10
231119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD
4rawdata
06560
0521
03415
0315
0478
103
696
1062
7143
1089
144119864minus03
975
64332119864minus10
218119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD5rawdata
06190
0545
03371
0282
0425
109
692
1118
7100
1147
144119864minus03
975
60334119864minus10
207119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
016
Colum
nD
6rawdata
05760
0587
03383
0238
0359
107
635
1103
6516
1131
157119864minus03
975
56364119864minus10
210119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
017
Colum
nD
1corrected
07830
0542
04243
0359
0623
86907
1158
3397
434
110119864minus03
267
209
334119864minus10
261119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
013
Colum
nD
2corrected
07230
0584
04221
0301
0521
88857
1186
3212
444
117119864minus03
267
193
353119864minus10
255119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
014
Colum
nD
3corrected
06850
0603
04129
0272
0463
97896
1307
3354
490
112119864minus03
267
183
338119864minus10
231119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
015
Colum
nD
4corrected
06560
0621
040
730249
0420
103
911
1388
3411
520
110119864minus03
267
175
332119864minus10
218119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
015
Colum
nD
5corrected
06190
0650
04025
0217
0362
109
905
1462
3390
548
110119864minus03
267
165
334119864minus10
207119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
016
Colum
nD
6corrected
05760
0701
04038
0172
0289
107
831
1442
3111
540
120119864minus03
267
154
364119864minus10
210119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
017
Guiochon
columns
1198641-1198642
(2007)
Colum
nE 1
rawdata
064
500594
03830
0262
0425
356
1119
1735
4371
678
894119864minus04
256
165
804119864minus11
519119864minus11
1500
046
100
000
030
000
0300
000
41300
030
079
047
Colum
nE 2
rawdata
05400
0800
04320
0108
0190
470
1000
1853
2161
400
100119864minus03
463
250
729119864minus11
393119864minus11
1500
046
100
000
027
000
0270
000
41300
030
070
056
Colum
nE 1
corrected
064
500594
040
000245
040
8356
969
1502
3628
563
103119864minus03
267
172
804119864minus11
519119864minus11
1500
046
100
000
028
000
0279
000
41300
030
075
047
Colum
nE 2
corrected
05400
0800
04320
0108
0190
470
577
1068
2161
400
173119864minus03
267
144
729119864minus11
393119864minus11
1500
046
088
000
023
000
0205
000
41300
030
070
056
Guiochon
columns
1198651ndash1198654
(2008)
Colum
nF 1
rawdata
06350
0586
03720
0263
0419
197
725
1142
4865
766
138119864minus03
149
95399119864minus11
253119864minus11
300
021
100
000
017
000
0170
00035
100
048
129
076
Colum
nF 2
rawdata
064
200581
03730
0269
0429
361
807
1258
4863
758
124119864minus03
166
107
358119864minus11
230119864minus11
500
021
100
000
017
000
0170
00035
100
048
129
075
Colum
nF 3
rawdata
064
100588
03770
0264
0424
670
748
1166
4643
724
134119864minus03
161
103
387119864minus11
248119864minus11
1000
021
100
000
017
000
0170
00035
100
048
128
075
Colum
nF 4
rawdata
06390
0595
03800
0259
0418
1014
752
1177
4476
701
133119864minus03
168
107
384119864minus11
246119864minus11
1500
021
100
000
017
000
0170
00035
100
048
127
075
Colum
nF 1
corrected
06350
0630
040
000235
0392
197
954
1502
3572
563
105119864minus03
267
170
399119864minus11
253119864minus11
300
021
100
000
019
000
0195
00035
100
048
120
076
Colum
nF 2
corrected
064
200623
040
000242
0403
361
964
1502
3611
563
104119864minus03
267
171
358119864minus11
230119864minus11
500
021
100
000
019
000
0186
00035
100
048
120
075
Colum
nF 3
corrected
064
100624
040
000241
0402
670
963
1502
360
6563
104119864minus03
267
171
386119864minus11
248119864minus11
1000
021
100
000
019
000
0193
00035
100
048
120
075
Colum
nF 4
corrected
06390
0626
040
000239
0398
1014
960
1502
3594
563
104119864minus03
267
171
384119864minus11
246119864minus11
1500
021
100
000
019
000
0192
00035
100
048
120
075
Journal of Materials 13
Table3Con
tinued
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Guiochon
columns
1198671-1198672serie
s(2010)
Colum
nH
1rawdata
05320
0744
03960
0136
0225
120
563
1057
3125
587
178119864minus03
180
96122119864minus10
652119864minus11
1500
046
100
000
026
000
0262
00104
050
005
013
009
Colum
nH
2rawdata
05420
0686
03720
0170
0271
61747
1379
4152
766
134119864minus03
180
98823119864minus11
446119864minus11
1000
046
100
000
025
000
0248
00054
050
005
013
009
Colum
nH
1corrected
05320
0752
040
000132
0220
120
799
1502
2993
563
125119864minus03
267
142
122119864minus10
652119864minus11
1500
046
100
000
031
000
0313
00104
050
005
013
009
Colum
nH
2corrected
05420
0738
040
000142
0237
61814
1502
3049
563
123119864minus03
267
145
823119864minus11
446119864minus11
1000
046
100
000
026
000
0259
00054
050
005
013
009
Guiochon
columns
1198673-1198674serie
s(2010)
Colum
nH
3rawdata
06270
060
903820
0245
0396
463
700
1117
4296
685
143119864minus03
163
102
447119864minus11
280119864minus11
1500
021
100
000
018
000
0177
00036
050
024
063
038
Colum
nH
4rawdata
05170
0791
040
900108
0183
433
616
1191
2639
511
162119864minus03
233
121
580119864minus11
300119864minus11
1500
021
100
000
019
000
0189
00036
050
024
059
047
Colum
nH
3corrected
06270
0638
040
000227
0378
463
942
1502
3527
563
106119864minus03
267
167
447119864minus11
280119864minus11
1500
021
100
000
021
000
0205
00036
050
024
060
038
Colum
nH
4corrected
05170
0791
040
900108
0183
433
699
1353
2639
511
143119864minus03
265
137
580119864minus11
300119864minus11
1500
021
100
000
020
000
0201
00036
050
024
059
047
Guiochon
columns
1198681-1198686
(2010)
Colum
nI 1rawdata
06580
0562
03700
0288
0457
488
741
1127
5156
784
135119864minus03
144
95390119864minus11
256119864minus11
500
021
100
000
017
000
0170
00104
050
024
065
037
Colum
nI 2rawdata
06550
0576
03770
0278
044
6873
661
1009
4745
724
151119864minus03
139
91437119864minus11
286119864minus11
1000
021
100
000
017
000
0170
00104
050
024
064
037
Colum
nI 3rawdata
06550
0588
03850
0270
0439
1209
610
931
4341
663
164119864minus03
141
92474119864minus11
310119864minus11
1500
021
100
000
017
000
0170
00104
050
024
062
037
Colum
nI 4rawdata
06430
0572
03680
0275
0435
260
787
1225
5153
801
127119864minus03
153
98367119864minus11
236119864minus11
500
030
100
000
017
000
0170
00104
050
012
032
018
Colum
nI 5rawdata
06540
0583
03810
0273
044
144
3683
1044
4531
693
146119864minus03
151
99423119864minus11
277119864minus11
1000
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 6rawdata
06550
0585
03830
0272
044
164
6665
1016
4438
678
150119864minus03
150
98434119864minus11
285119864minus11
1500
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 1corrected
06580
060
8040
000258
0430
488
988
1502
3701
563
101119864minus03
267
176
390119864minus11
256119864minus11
500
021
100
000
020
000
0196
00104
050
024
060
037
Colum
nI 2corrected
06550
0611
040
000255
0425
873
984
1502
3684
563
102119864minus03
267
175
437119864minus11
286119864minus11
1000
021
100
000
021
000
0207
00104
050
024
060
037
Colum
nI 3corrected
06550
0611
040
000255
0425
1209
984
1502
3684
563
102119864minus03
267
175
474119864minus11
310119864minus11
1500
021
100
000
022
000
0216
00104
050
024
060
037
Colum
nI 4corrected
06430
0622
040
000243
040
5260
966
1502
3617
563
104119864minus03
267
172
367119864minus11
236119864minus11
500
030
100
000
019
000
0188
00104
050
012
029
018
Colum
nI 5corrected
06540
0612
040
000254
0423
443
982
1502
3679
563
102119864minus03
267
175
423119864minus11
277119864minus11
1000
030
100
000
020
000
0204
00104
050
012
029
018
Colum
nI 6corrected
06550
0611
040
000255
0425
646
984
1502
3684
563
102119864minus03
267
175
434119864minus11
285119864minus11
1500
030
100
000
021
000
0207
00104
050
012
029
018
Guiochon
columns
1198691-1198692
(2011)
Colum
nJ 1rawdata90
∘A(120578
=00054
P)04980
0803
040
000098
0163
65654
1313
2801
563
153119864minus03
234
116126119864minus10
627119864minus11
1500
046
100
000
029
000
0287
00054
050
005
013
010
Colum
nJ 2rawdata90
∘A(120578
=00104
P)04980
0803
040
000098
0163
124
651
1307
2801
563
154119864minus03
232
116127119864minus10
630119864minus11
1500
046
100
000
029
000
0287
00104
050
005
013
010
Colum
nJ 1rawdata160
∘A(120578
=00054
P)05630
0714
04020
0161
0269
71896
1592
3099
550
112119864minus03
289
163
102119864minus10
573119864minus11
1500
046
100
000
030
000
0302
00054
050
005
012
009
Colum
nJ 2rawdata160
∘A(120578
=00104
P)05630
0714
04020
0161
0269
129
847
1504
3099
550
118119864minus03
273
154
108119864minus10
606119864minus11
1500
046
100
000
030
000
0302
00104
050
005
012
009
Colum
nJ 1corrected
04980
0803
040
000098
0163
65748
1502
2801
563
134119864minus03
267
133
126119864minus10
627119864minus11
1500
046
100
000
031
000
0307
00054
050
005
013
010
Colum
nJ 2corrected
04980
0803
040
000098
0163
124
748
1502
2801
563
134119864minus03
267
133
127119864minus10
630119864minus11
1500
046
100
000
031
000
0308
00104
050
005
013
010
Colum
nJ 1corrected
05630
0714
04020
0161
0269
71827
1470
3099
550
121119864minus03
267
150
102119864minus10
573119864minus11
1500
046
100
000
029
000
0290
00054
050
005
012
009
Colum
nJ 2corrected
05630
0714
04020
0161
0269
129
827
1470
3099
550
121119864minus03
267
150
108119864minus10
606119864minus11
1500
046
100
000
030
000
0299
00104
050
005
012
009
14 Journal of Materials
050
100150200250300350400
045 050 055 060 065 070 075 080 085
LineAll corrected data
Linear (all corrected data)Squares raw data Triangles corrected data
Pt
120576t
D seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
J1 J2 H4
H1H2
D6
B2 B3 B4
J3 J4
E2
C
H3
D5 D4
E1
D2
A1
D1
A2
y = 267xLine slope = 267
Figure 2 Summary of experimental results
0
500
1000
1500
2000
2500
20 25 30 35 40 45 50 55 60 65 70 75 80
Linear (line)
Ψ
LineAll corrected dataD seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
Line slope (P0) = 267
J1 J2
J3 J4E2
E1
A1
A2
H4 H1
C
120601
H2H3I1 I2 I3 I4 I5
F1 F2 F3 F4
D1 D2 D3 D4 D5 D6
Squares raw data Triangles corrected data
Figure 3 Summary of experimental results
Figures 2 and 3 herein are graphical representations ofTable 31015840s raw and corrected data for Guiochonrsquos columnsplotted in the same reference frames as in Figures 1(a) and1(b) respectively
As shown in the plots of the thirty-one total columnsevaluated in this data base only six columns had reportedraw data values for 119875
0close to the normThese were columns
B2 C E
1 H
4 J
1and J
2 Interestingly of the six conforming
columns 3 of the columns were packed with fully porousparticles and 3 columns with partially porous particles Itcan be seen however that when properly interpreted andadjusted all of Guiochonrsquos experimental data falls fairlywithin the norms outlined in the reference chart in Table 1In order to explore the reasons why the raw data does notconform more closely to the norms outlined in Table 1 eachof Guiochonrsquos studies included in Table 3 is evaluated in turn
Finally Table 4 herein is a list of symbols and parameterdefinitions used throughout this paper as well as their unitsof measure
91 Guiochonrsquos 1998 Study In a 1998 publication by Kohand Guiochon entitled ldquoEffect of the Column Length on theCharacteristics of the Packed Bed and the Column Efficiencyin a Dynamic Axial Compression Columnrdquo [19] Guiochonand his coauthor report the results of a study they hadcarried out on column packing using a rather novel techniquewhich was a combination of slurry packing and mechanicalcompression
The study involved the packing of a single cylinder ofdiameter 5 cm to various column lengths ranging between 1and 25 cm approximately The particles were all silica basedand coated with a reverse stationary phase C
18 However two
different categories of particles were involved One categorycontained highly irregular particles which were about 130micrometers in average diameter In addition these particleshad a high specific pore volume (11mLg) and exhibitedsignificant particle breakage under the packing conditionsinvestigated in the study Conversely the other particlecategory involved very small particles (6 micrometers) whichwere spherical in shape had a low specific pore volume(03mLg) and exhibited a greater ability to withstandparticle breakage under the studyrsquos packing conditions
The permeability data for the irregular particles isneglected herein because according to the authors theparticles exhibited significant breakage during packing andwithout an accurate value for the characteristic dimensionof the particle the data cannot be used in the Kozeny-Blake model to determine the value of 119875
0 The spherical
particles on the other hand behaved successfully under thepacking conditions chosen and yielded values calculated bythe authors for 119875
0of 619 268 226 and 229 for the four
different lengths of columns studied The data for thesecolumns is reported in our Table 3 in the rows for ColumnsB1through B
4 The higher value of 619 (the shortest column)
was rejected by the authors because they felt that either theparticles had been somewhat broken during packing or thecolumn frits had been blocked with fines Accordingly thepermeability data for this column is likewise neglected herein
The remaining three columns however were not con-sistent with respect to the authorsrsquo calculated values for 119875
0
Since the particles were identical the range of values for 1198750
of 226 to 268 must be due to experimental error The rawdata for column B
2 which produced a 119875
0value of 268 is
all self-consistent For example it generates a value for 119875119905
of 148 which is a reasonable value based upon the knownlow particle porosity for these Nova Pak particles and acorresponding value for 120576
0of 04073 which is a reasonable
value for a well-packed column However the other twocolumns columns B
3and B
4 had the much lower value of
126 for 119875119905 Accordingly in Table 3 herein corrected values
for column external porosity for columns B3and B
4are
displayed These corrected values are only slightly differentthan the reported values are within the experimental errorof the measurement and therefore in combination with the
Journal of Materials 15
Table 4 Glossary of terms
Symbol Unit dimensions (cgs) Definition119860 cm2 Column cross-sectional area119863 cm Column diameter119889119901
cm Particle spherical diameter equivalentΔ119875 gcmminus1secminus2 Pressure drop along column119889119901119898
cm Average particle diameter (measured)1198960
None Guiochonrsquos residual reciprocal permeability coefficient119875119905
None Guiochonrsquos modified permeability coefficient in Kozeny-Blake equation119871 cm Column length1198750
None Permeability coefficient in Kozeny-Blake equation119876 cm3minminus1 (exception) Volumetric fluid flow rate1199050
sec Time for elution of totally permeating unretained solute119896 cm2 General permeability coefficient in Darcy equation119870
119865cm2 Halaszrsquos permeability coefficient in Darcy equation (1974 paper)
119870 cm2 Halaszrsquos permeability coefficient in Darcy equation (1972 paper)119870
119888None Guiochonrsquos permeability coefficient in the Kozeny-Blake equation (as modified by Carmen)
Greek1205760
None External column porosity120576119894
None Internal column porosity120576119905
None Total column porosity120576119901
None Particle porosityΦ None Giddingsrsquo ratio of external and total column porosity120601 None Flow resistance parameter based on mobile phase velocity1206010
None Flow resistance parameter based on superficial velocityΩ
119901None Particle sphericity factor
120578 gcmminus1secminus1 Fluid absolute viscosity120583119894
cmsecminus1 Average fluid interstitial linear velocity120583119904
cmsecminus1 Average fluid superficial linear velocity120583119905
cmsecminus1 Average mobile phase velocityΨ None Porosity dependence term in the Kozeny-Blake equation based on mobile phase velocityΨ
0None Porosity dependence term in the Kozeny-Blake equation based on superficial velocity
raw data for column B2 are supportive of the value of 267 for
1198750
92 Guiochonrsquos 1999 Study In a 1999 publication by Farkaset al entitled ldquoValidity of Darcyrsquos Law at Low Flow Rates inLiquid Chromatographyrdquo [20] Guiochon and his coauthorsreport the results of some very exactingmeasurements whichthey made to validate Darcyrsquos law at extremely low fluidflow rates Using a column containing particles packed toa measured external column porosity of 0399 and havinga measured total column porosity of 0593 (Figure 2 in thepaper) Guiochon et al measured the column pressure dropand fluid flow rate at flow rates ranging between 0015 and05mLminwith ethylene glycol as the fluidThe authors statethat the particles in the column are spherical and they appearto have gone to extraordinary lengths to obtain an accuratemeasure of the particle diameter and of course as a result ofmodern techniques of particle size classification the particlesize distribution is very narrow
In Figure 1 of the paper the authors show a plot ofcolumn measured pressure drop in bar versus ldquofluid linearvelocityrdquo in cmsec This plot represents their measurementstaken in the empty column The only velocity that exists in
an empty column is the superficial velocity and thus theabscissa values in the plot in Figure 1 designated as ldquofluidlinear velocityrdquo represent superficial linear velocity
In Figure 2 in their paper however the authors showanother plot of measured pressure drop in bar also versusldquofluid linear velocityrdquo in cmsec The velocity on the abscissain this plot although also designated as ldquofluid linear velocityrdquodoes not equate (for permeability related purposes) to theldquofluid linear velocityrdquo in Figure 1 This is because in Figure 2the columnwas filled with porous particles and themeasuredvelocity was the mobile phase velocity Indeed the onlyvelocity used throughout the entire paper is the mobile phasevelocity (mobile phase velocity and superficial velocity areone and the same in an empty column)
In Table 1 of their paper the authors report that for thecolumn represented in Figure 2 the slope of a line achievedwhen the mobile phase velocityin units of cmsec is plottedversus pressure drop in units of bar is 1829 They also reportthat the flow resistance parameter 120601 for this column is 865In Table 3 herein it can be seen that the raw data for thiscolumn (Column C) generates an apparent value of 257 for1198750 This value is very close to Giddingsrsquo value of 267 Indeed
by referring to Figures 2 and 3 herein one can see that thereported raw data in this study without any modification
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
Journal of Materials 5
Table1Colum
nperm
eabilityreferencetable
119870119865
Particlepo
rosity
rang
e120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119896119871
119863Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
120576119905Φ
(120576119905minus1205760)
120576119894
Δ119875119889119901
2Δ119875119889119901
2(1minus120576119905Φ)2120576119905
(1minus1205760)2
1120601
1198750120576119905
119889119901
2120576119905
(119889119901119898Ω119901)
(1minus1205760)
120583119905120578119871
120583119904120578119871
(120576119905Φ)3
(1205760)3
120601Ψ
120601po
iseUnits
Non
eNon
eNon
eNon
eNon
epsi
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
cmNon
ecm
cmgcmminus1 secminus1
cm3 m
inminus1
cmsecminus
1cm
secminus
1cm
secminus
1
Non
porous
particles
0380
1000
0380
000
0000
0160
711
1870
2662
701
141119864minus03
267
101
535119864minus10
15046
100
000100
000100
00100
038
0038
0100
0100
0390
1000
0390
000
0000
0147
653
1675
244
6627
153119864minus03
267
104
597119864minus10
15046
100
000100
000100
00100
039
0039
0100
0100
040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
100
000100
000100
00100
040
004
00100
0100
0410
1000
0410
000
0000
0124
553
1349
2071
505
181119864minus03
267
109
742119864minus10
15046
100
000100
000100
00100
041
004
10100
0100
0420
1000
0420
000
0000
0115
509
1212
1907
454
196119864minus03
267
112
825119864minus10
15046
100
000100
000100
00100
042
004
20100
0100
Lowpo
rosity
particles
0507
0750
0380
0127
0204
213
948
1870
3549
701
106119864minus03
267
135
535119864minus10
15046
100
000100
000100
00100
051
0051
0133
0100
0520
0750
0390
0130
0213
196
871
1675
3261
627
115119864minus03
267
139
597119864minus10
15046
100
000100
000100
00100
052
0052
0133
0100
0533
0750
040
00133
0222
180
801
1502
2999
563
125119864minus03
267
142
666119864minus10
15046
100
000100
000100
00100
053
0053
0133
0100
0546
0750
0410
0136
0231
166
737
1349
2760
505
136119864minus03
267
146
742119864minus10
15046
100
000100
000100
00100
055
0055
0133
0100
0560
0750
0420
0140
0241
153
679
1212
2542
454
147119864minus03
267
149
825119864minus10
15046
100
000100
000100
00100
056
0056
0133
0100
Medium
porosity
particles
0613
0620
0380
0233
0376
258
1146
1870
4293
701
872119864minus04
267
164
535119864minus10
15046
100
000100
000100
00100
061
0061
0161
0100
0629
0620
0390
0239
0392
237
1053
1675
3946
627
949119864minus04
267
168
597119864minus10
15046
100
000100
000100
00100
063
0063
0161
0100
064
50620
040
00245
040
9218
969
1502
3629
563
103119864minus03
267
172
666119864minus10
15046
100
000100
000100
00100
064
0065
0161
0100
0661
0620
0410
0251
0426
201
892
1349
3340
505
112119864minus03
267
177
742119864minus10
15046
100
000100
000100
00100
066
006
60161
0100
0677
0620
0420
0257
044
4185
821
1212
3076
454
122119864minus03
267
181
825119864minus10
15046
100
000100
000100
00100
068
006
80161
0100
Highpo
rosity
particles
0760
0500
0380
0380
0613
320
1422
1870
5325
701
703119864minus04
267
203
535119864minus10
15046
100
000100
000100
00100
076
0076
0200
0100
0780
0500
0390
0390
0639
294
1306
1675
4891
627
766119864minus04
267
208
597119864minus10
15046
100
000100
000100
00100
078
0078
0200
0100
0800
0500
040
0040
0066
7270
1202
1502
4500
563
832119864minus04
267
214
666119864minus10
15046
100
000100
000100
00100
080
0080
0200
0100
0820
0500
0410
0410
0694
249
1105
1349
4140
505
905119864minus04
267
219
742119864minus10
15046
100
000100
000100
00100
082
0082
0200
0100
0840
0500
0420
0420
0724
229
1018
1212
3814
454
982119864minus04
267
224
825119864minus10
15046
100
000100
000100
00100
084
0084
0200
0100
Capillary
1000
0380
0380
0620
1000
421
1870
1870
7005
701
535119864minus04
267
267
535119864minus10
15046
100
000100
000100
00100
100
0100
0263
0100
1000
0390
0390
0610
1000
377
1675
1675
6273
627
597119864minus04
267
267
597119864minus10
15046
100
000100
000100
00100
100
0100
0256
0100
1000
040
0040
0060
010
00338
1502
1502
5625
563
666119864minus04
267
267
666119864minus10
15046
100
000100
000100
00100
100
0100
0250
0100
1000
0410
0410
0590
1000
303
1349
1349
5051
505
742119864minus04
267
267
742119864minus10
15046
100
000100
000100
00100
100
0100
0244
0100
1000
0420
0420
0580
1000
273
1212
1212
4541
454
825119864minus04
267
267
825119864minus10
15046
100
000100
000100
00100
100
0100
0238
0100
6 Journal of Materials
0200400600800
100012001400160018002000
20 30 40 50 60 70
NonporousLow porosityMedium porosity
CapillaryLinear (line)
Nonporous Low porosity particles
Medium porosity particles
High porosity particles
Capillary
Line slope (P0) = 267
y = 267x + 0
Ψ
120601
Line
High porosity
(a)
99111123135147159171183195207219231243255267
035 045 055 065 075 085 095
LineCapillaryLinear (line)
Capillary
Line slope (P0) = 267
y = 267x + 0
Pt
120576t
Φ = 10
Φ = 075
Φ = 062
Φ = 05
Φ = 075
Φ = 062
Φ = 05
Φ = 1
(b)
99111123135147159171183195207219231243255267
000 010 020 030 040 050 060 070 080 090 100
NonporousLow porosityMedium porosityHigh porosity
120576p
y = 160x + 107
Pt
LineCapillaryLinear (line)
(c)
Figure 1 (a) The balance between 120601 and Ψ (b) Column permeability reference chart (c) Particle permeability reference chart
In the 2006 edition of their textbook [9] Guiochon andhis coauthors provide the following worked example of hispressure dropfluid flow teaching3
In Table 2 a cross-reference is established between Guio-chonrsquos teaching for columns packed with both porous andnonporous particles the Kozeny-Blake equation (asmodifiedby Carman) and for validation purposes Giddingsrsquo teachingbased upon actual permeabilitymeasurements fromhis Table53-1 in his 1965 text [8]4
In his worked example Guiochonrsquos fluid velocity 119906is found in his text as his equation (256) [9] at page61 and equates to mobile phase not superficial velocityAccordingly when applied to columns packed with porousparticles his worked example combines the contributionto fluid velocity of mass transfer by molecular diffusionin the free space within the particles with mass transferby convection in the free space between the particles Asdiscussed above if one is attempting to compare Guiochonrsquos
teaching for the value of the permeability coefficient with thatof the Kozeny-Blake equation the conversion factor is thetotal porosity of the column
In addition because Guiochonrsquos particle sizeparameter has an embedded assumption due to particleshaperoughness his worked example also commingles thecontribution to spherical particle diameter equivalent ofparticle shaperoughness with other nonparticle variablesUnfortunately his worked example is not based uponcompletely spherical particles where this would not be anissue We know this because he states that 119896
0in his worked
example equals 1 times 10minus3 which he says is the value of thepermeability coefficient for columns packed with irregularparticles
In order to move forward one must simplify If this wereto be done experimentally it would best be accomplishedby experiments with nonporous particles in which case thevariable of molecular diffusion in the pores of the particles
Journal of Materials 7
Table2Cr
oss-referencefor
term
s119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119896119871
119863Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
120576119905Φ
(120576119905minus1205760)
120576119894
Δ119875119889119901
2Δ119875119889119901
2(1minus120576119905Φ)2120576119905(1minus1205760)2
1120601
1198750120576119905
119889119901
2120576119905
(119889119901119898Ω119901)
(1minus1205760)
120583119905120578119871
120583119904120578119871
(120576119905Φ)3
(1205760)3
120601Ψ
120601po
iseUnits
Non
eNon
eNon
eNon
eNon
epsi
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
cmNon
ecm
cmgcmminus1 secminus1cm
3minminus1cm
secminus
1cm
secminus
1cm
secminus
1
Nonporousp
articles
Gidding
srsquotradition
alno
nporou
scolum
n040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
1000010
00010
0010
0399
004
00100
0100
Guiocho
nrsquostradition
alspheric
alparticle(app
arent)
0362
1000
0362
000
0000
0187
830
2293
3106
858
120119864minus03
267
97436119864minus10
15046
1000010
00010
0010
0361
0036
0100
0100
Gidd
ingrsquos
table5
3-1
5060meshglassb
eads
0422
1000
0422
000
0000
0113
500
1184
1873
444
200119864minus03
267
113
844119864minus10
15046
1000010
00010
0010
0421
004
20100
0100
5060meshglassb
eads
040
910
00040
9000
0000
0126
560
1371
2097
513
179119864minus03
267
109
729119864minus10
15046
1000010
00010
0010
040
70041
0100
0100
Guiochonrsquosteaching
ExA
irregular
particle
(app
arent)
040
010
00040
0000
0000
0225
1000
2500
2250
563
100119864minus03
444
178
400119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
ExA
correctedforp
article
irregularity
040
010
00040
0000
0000
0225
601
1502
2250
563
166119864minus03
267
107
400119864minus10
15046
078
00010
000
080010
0399
004
00100
0100
ExB
spheric
alparticle
(app
arent)
040
010
00040
0000
0000
0187
830
2075
2250
563
120119864minus03
369
148
482119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
ExB
corrected
040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
Porous
particles
Gidding
srsquotradition
alpo
rous
column
060
0066
7040
00200
0333
203
900
1500
3375
563
111119864minus03
267
160
667119864minus10
15046
100
00010
00010
0010
0599
006
00150
0100
Gidd
ingrsquos
table5
3-1
3040meshalum
ina
0803
0498
040
00403
0672
271
1204
1499
4510
562
830119864minus04
267
214
667119864minus10
15046
100
00010
00010
0010
0801
0080
0201
0100
5060meshalum
ina
0837
0499
0417
0420
0721
235
1043
1246
3906
467
959119864minus04
267
224
803119864minus10
15046
100
00010
00010
0010
0835
0084
0201
0100
6080meshchromosorbW
(5DNP)
0766
0498
0382
0384
0621
316
1404
1833
5259
687
712119864minus04
267
204
545119864minus10
15046
100
00010
00010
0010
0764
0077
0201
0100
6080meshchromosorbW
(20
DNP)
0785
0495
0389
0396
064
8300
1333
1698
4991
636
750119864minus04
267
210
589119864minus10
15046
100
00010
00010
0010
0783
0079
0202
0100
8 Journal of Materials
would be eliminated and by using only smooth sphericalparticles such as glass beads in which case the variableof particle shaperoughness would be eliminated If both ofthese things were done one would be able to directly andunambiguously determine the true value of 119875
05
Accordingly since his textbook contains a general teach-ing without any expressed restrictions with respect to particleporosity let us first assume that the column in Guiochonrsquosworked example is packed with nonporous particles Sec-ondly let us assume that the value for the external porosityof the column (which is the same as the total porosity ina column packed with nonporous particles) is the valuetraditionally assumed for a typical well-packed column thatis 04
As illustrated by example A in Table 2 this would leadto an apparent value for 119875
0of 444 which is obviously much
too high6 Using the adjusted particle size of 8 micrometers(rounded) which is calculated based on a value of 078 forΩ
119901
we derive values of 267 for 1198750and 107 for 119875
119905(the relationship
between these two values is of course equal to the totalporosity of the column 120576
119905 as is dictated by (20))
Similarly as illustrated by example B in Table 2 when thecolumn of Guiochonrsquos worked example is adapted to accom-modate spherical nonporous particles Guiochonrsquos teachingof the value for 119896
0of 12 times 10minus3 results in an apparent value
of 369 for 1198750
7 again too high [10 11] On the other hand ifwe correct the value of 119875
0 we get a column external porosity
of 03620which is too low for a typical well-packed column(120576
0= 04) [8] Moreover in the case of both corrections
Guiochonrsquos teaching overstates the pressure gradient (187 psicompared to the 135 psi which Giddingsrsquo experiments wouldgenerate) Accordingly it is only when we correct for thepressure drop that we achieve appropriate values of 04 for1205760 267 for 119875
0 and 107 for 119875
119905
As is evident from Table 2 when the porosity of columnspacked with nonporous particles varies the value of 119875
119905also
varies because it is based upon the measurement of mobilephase velocity which in turn changeswith the porosity of thecolumn Indeed for typical columns packed with nonporousparticles having external porosities close to 04 (038ndash042)[8] Table 2 reveals that the value of 119875
119905varies by as much as 6
points while the value of 1198750remains constant at 267
Having established a range of values for 119875119905based upon
columns packed with nonporous particles one can nowproceed to evaluate the range of values for 119875
119905based upon
columns packed with porous particles which is to say thatwe can now evaluate the impact of particle porosity onpermeability Among other things Table 2 shows that whenit comes to porous particles Guiochonrsquos 119875
119905is even more
variable (and thus even less useful) because the value ofthe coefficient can differ on a column-to-column basis byas much as 90 points depending on the combination of theinternal andexternal column porosities at play
Figures 1(a) and 1(b) are a plot of the data found in Table 1In this plot Guiochonrsquos modified permeability coefficient 119875
119905
is plotted against column total porosity 120576119905 Once again it will
be appreciated that (a) all data falls on a line with a slope of267 which represents the value of 119875
0and (b) the line has no
intercept which means that when 120576119905is zero 119875
119905is also zero As
shown in the plot values of the parameter Φ are highest atlow values of 119875
119905and lowest at high values of 119875
119905 Further as is
evidenced by the respective definitions for the terms119875119905and119875
0
contained herein the only time that the value of Guiochonrsquos119875119905does not differ from the value of 119875
0(267) is when the value
of 120576119905is 1 its value in a theoretical packed column devoid of
particles that is acapillaryFinally Figure 1(c) is another plot of the data in Table 1
but this time the values for 119875119905are plotted against particle
porosity 120576119901 The equation of this line does have an intercept
which represents the value of 119875119905when the particles are
nonporous It is also obvious from the equation of this linethat it is only when particle porosity is unity that is in acapillary that the value of 119875
119905is 267
It will be appreciated that since 120576119905is the sumof the internal
and external porosities of a packed column any given valuefor 119875
119905involving a column packed with porous particles can
represent many different combinations of these parametersVice versa any given value for 120576
119905can generate many different
values for 119875119905depending on the amount of particle porosity
which is present On the other hand as is pointed outearlier column internal porosity andor particle porosityplay no role in determining the permeability of a packedcolumn the only porosity parameter which is relevant forsuch purposes is column external porosity In other words asa permeability coefficient 119875
119905is not a particularly informative
concept because it still fails to provide any meaningful tie tothe underlying physical phenomena that govern the pressuredropfluid flow relationship
In the case of permeability measurements one quan-tifies only total pressure drop Pressure drop however isa commingling of the contributions of the several differ-ent parameters embodied in the pressureflow relationshipTherefore one must accurately represent the contributionof each factor before adding them together to equal thepressure drop Obviously if one mistakes the contribution ofany given element for instance particle size this will skewthe contribution of some other element when summed forinstance column porosity
Our frame of reference has been calibratedwith respect tothe interrelationships between the values of 119875
0 119875
119905 120576
119905 and 120576
119901
based on the data reported in Giddingsrsquo table 53-1 in his 1965textbook Since his data base was derived from experimentsusing nonporous particles the permeability results of whichwere extended to columns packed with fully porous particlesby using his Φ parameter the validity of which in turn wasestablished by independent means there is no uncertaintywith respect to the critical variable of external porosity 120576
0
inherent in our frame of reference It is for this reasontherefore that our frame of reference trumps any reporteddata which uses any other technique including the techniqueof inverse size exclusion chromatography by itself to establishvalues for 120576
0[12] In addition our frame of reference is
internally consistent due to the grounding of the parametersat the boundary condition of a theoretical capillary at whichpoint the values of 120576
119905and 120576
119901are unity and the values of 119875
0and
119875119905are equal In essence these interrelationships exemplify the
conservation laws of nature which are the ultimate standard
Journal of Materials 9
of measure when considering partial fractions of a singleentity
Consequently as Figures 1(b) and 1(c) illustrate if oneknows based upon other independent means the value of acolumnrsquos total porosity or its particlesrsquo porosity respectivelyone can at least verify whether results of permeability exper-iments expressed in terms of 119875
119905are reasonable Or to put
it another way if the 119875119905results reported do not conform to
what these plots tell us one should expect from columns andparticles based upon independently derived porosity valuesfor the particles under study we know that something iswrong Accordingly this is the way in which we use 119875
119905below
to analyze and evaluate Guiochonrsquos experimental work of thelast decade
8 The Origin of Guiochonrsquos Value forColumn Permeability
In 1966 Guiochon wrote a review paper entitled ldquoProblemsRaised by the Operation of Gas Chromatographic Columnsrdquo[13] In that paper he reviewed the development of the per-meability equation from Darcy to the then present Amongother things he reported that ldquoa mean 120576 of 038 is foundfor a number of packings obtained from powdered prod-ucts of various origins consisting of spherical or irregularparticles Almost all the measured values are between035 and 045rdquo (p 14) He then went on to discuss what hecalled the ldquosemiempirical Blake-Kozeny equationrdquo which hereported as 119896 = [119889
2
1199011205762][150(1 minus 120576)
2] [13 p 15 Equation
(11)] What sticks out of course is the fact that Guiochonstates here that 119875
0equals 150 the value for the permeability
coefficient championed by Ergun not the value posited byCarman To be fair we should point out that Guiochongave warning that he was uncomfortable with any particularvalue for the permeability coefficient For even though herecited the Kozeny-Blake equation with a value of 150 for itspermeability coefficient he stated that ldquothe values of 119896 givenby (this) equation are not very accurate and the deviationscan be as large as 50rdquo [13 p 15] Guiochon attributedmost ofthis uncertainty to the inability of practitioners of that era toeffectively measure the value of a columnrsquos external porositySince as he pointed out ldquo(the Kozeny-Blake) equation isvery sensitive to variations in 120576rdquo he concluded that theequation is ldquoof little userdquo and that it was better practice torely upon grosser measurements of permeability [13 p 15]Thus Guiochon at this timeframe made a conscious decisionto staywithin the relatively safe but admittedlymore obscureboundaries of Darcyrsquos teaching on column permeability
In 2006 however Guiochon emerged from under theshadow of Darcy permeability when he published a reviewpaper which purported to be a comprehensive summaryof the current state of the art in chromatography [14]In light of the fact that Guiochon and his coauthors hadvirtually contemporaneously published a new version of theirtextbook one would have expected that Guiochon wouldtake this opportunity to reaffirm his textbook teaching So itshould come as no surprise that he would open his discussionof the pressure dropfluid flow relationshipwith the following
assertion ldquothe permeability of packed beds used in liquidchromatography is in the order of 1 times 10minus3rdquo [14 p 9] Asone would expect the authority which Guiochon cites forthis seemingly familiar proposition is the 2006 edition of hisown textbook For the first time however he also providessome citations to third party sources In particular he cites a1997 chromatography handbook written by Uwe Neue ChiefScientist for Waters Corporation [15 p 9] As one can seeNeuersquos equationmdashlike Guiochonrsquosmdashdoes indeed postulate apermeability coefficient with a value of 1 times 10minus3
At this point however Guiochon abandons his tradi-tional caution about the value of the permeability coeffi-cient and marries his permeability equation to the Kozeny-Blake equation thereby endorsing not only the porositydependence term in that equation but also a specific valuefor 119875
0 Referring to his value of 1 times 10minus3 he states the
following ldquothe specific column permeability is related to thecolumn porosity through the Kozeny-Carman equation 119896
119891=
12057630180 (1 minus 120576
0)2rdquo [14 p 9] Not surprisingly Neue makes
the same connection ldquothe specific permeability depends onthe particle size 119889
119901and the interstitial porosity 120576
0of the
packed bedThe relationship is known as theKozeny-Carmanequation 119861
0= (1185)(1205763
01198892119901(1 minus 120576
0)2)rdquo [15 p 30]
Neuersquos permeability equation and Guiochonrsquos textbookpermeability equation however differ in that Neue uses thesuperficial not the mobile phase velocity Accordingly sinceall other features of their equations are the same one wouldexpect them to get different values for their permeabilitycoefficients Neue himself recognizes this when he says inhis handbook ldquothis value [1000] is also in good agreementwith our experience but values as low as 500 have beenreported in the literature8 However there are several reasonswhy different values reported by different workers do notrepresent true differences in the resistance parameter Firstlower values are obtained for porous packings if the specificpermeability is measured on the basis of the breakthroughtime of an unretained peak instead of using the flow raterdquo see[15 p 32] (when he distinguishes the ldquobreakthrough time ofan unretained peakrdquo from the ldquoflow raterdquo Neue is of coursereferring to the mobile phase velocity and the superficialvelocity resp) Consequently if one takes Neuersquos point andturns it on its head one can see that if he and Guiochonwere to get the same value for their permeability coefficients[1000] but one (Guiochon) is using mobile phase velocityto derive it and the other (Neue) uses superficial velocity toderive it this would indeed represent a ldquotrue differencerdquo intheir equations
So how can Guiochon cite both Neue and himself for theproposition in his review paper that ldquothe permeability of thepacked beds used in liquid chromatography is in the order of1times 10minus3rdquoThe answer is that at least for purposes of his reviewpaper he adopts Kozeny-Blakersquos fluid velocity convention inplace of the velocity frame he uses in his textbook Here iswhat he says ldquothis approximation [1 times 10minus3] is valid only ifthe superficial velocity is usedrdquo [14 p 9] (emphasis supplied)
As a matter of pure mathematics Guiochonrsquos permeabil-ity coefficient of 1 times 10minus3 can only be equal to 1205763
0180(1 minus 120576
0)2
if the value of 1205760is 04 the typical value for the interstitial
10 Journal of Materials
fraction of a well-packed column Guiochon confirms in hisreview paper that this is indeed what he assumes in his state-ment of the Kozeny-Blake equation [14 p 9] Accordinglyhis adoption of the value of 180 for 119875
0solely depends on the
validity of his assertion that the value of the permeabilitycoefficient is 1 times 10minus3
We already know from our discussion of Guiochonrsquostextbook teaching that when this value is associated with apermeability equation based on mobile phase velocity it iswrong As reflected in our Table 2 when Guiochonrsquos workedexample involving this figure is corrected to account for theembedded factor due to the irregularity of the particles whichhis worked example assumes the value of the permeabilitycoefficient generated by his equation is not 1 times 10minus3 it is 166times 10minus3
So where did this value of 1 times 10minus3 come from AlthoughGuiochon in his review paper cites Neuersquos handbook for thisvalue Neue himself offers absolutely no authority for it Onthe other hand Guiochon does provide one other citation inhis review paper for this value The citation is to an article byHalasz et al published in 1975 [16]
If one then goes to that article one sees that Halaszet al do indeed proffer the same equation as Neue (albeitin a somewhat different format) and yes it does containthe value of 1 times 10minus3 See [16 Equation (1)] Neverthelessthe authors fail to identify from whence they obtained thisvalue On the other hand they do make reference to apaper that Endele et al wrote a year earlier in 1974 withKlaus Unger entitled ldquoInfluence of the Particle Size (5ndash35 120583)of Spherical Silica on Column Efficiencies in High-PressureLiquid Chromatographyrdquo [17] It appears that this is wherethe body is buried for it is in this article that Halasz reportshaving actually derived a value of 1 times 10minus3 for the permeabilitycoefficient
In summarizing the results of their experiments whichincidentally are based on measurements of superficial notmobile phase velocity (see their Equation (7)) Halasz et alcalculate an average value for the permeability coefficient(which they notate as119870
119865) of 1198892
1199011080 Accordingly they state
the following ldquo[I]t is proposed to define an average particlesize 119889
119901 for columns packed with spherical silica using the
balanced density packing method by the equation 1198892119901
=
1198701198651000 = 1198651205781198711031199032120587Δ119875rdquo [17 p 382 their Equation (7)]To recapitulate it thus appears that (1) Halasz was the
source of the value of 1 times 10minus3 for the permeability coefficientwhich Guiochon adopts in his review paper (2)Halaszrsquo valuewas based upon measurements which involved sphericalparticles and (3)Halaszrsquo value was properly derived from anequation which utilized the superficial velocity as the fluidvelocity
However we still have a problem if we are going to acceptthe value of 1 times 10minus3 as a basis for calculating the constantin Kozeny-Blake we still need to know the value of Halaszrsquocolumnsrsquo external porosity For example in the absenceof measured values for the external porosity Guiochonrsquosassertion in his review paper that the value of the constantis 180 is without foundation
Unfortunately however Halasz bases his methodologyon ldquoassumedrdquo versus ldquomeasuredrdquo values for external porosityNevertheless he does provide some clues that may helpus peel back this onion First he says that his value forthe permeability coefficient applies to ldquocolumns using thebalanced density packing methodrdquo Secondly after statinghis proposed value for the coefficient we find the followingtelltale sentence ldquothe permeability of columns packed by theconventional ldquodryrdquo method are worse by a factor of about 2than those packed by the balanced density methodrdquo [17 p382] For this proposition Halasz cites yet another of his ownarticles this time a paper written by himself and M Naefein 1972 entitled ldquoInfluence of Column Parameters on PeakBroadening in High-Pressure Liquid Chromatographyrdquo [18]
When one follows this thread by going to Halaszrsquos 1972paper one finds a possible solution to the puzzle For this iswhere Halasz describes his calculations of the permeabilitycoefficient from experiments with a number of conventionaldry-packed columns containing spherical silica particlesNot surprisingly the resulting value for the permeabilitycoefficient for such columns was not 1 times 10minus3 Instead Halaszsays that his experiments with such columns resulted inthe following value for the permeability coefficient ldquo119870 sim
11988921199012000rdquo [18 p 78] In other words consistent with what
Halasz says in his 1974 article the value for the coefficient forthese dry-packed columns was indeed ldquoworse by a factor ofabout 2rdquo compared to its value for columns prepared with thebalanced density method
The obvious implication of this is that when Halasz gota value of 1 times 10minus3 for the permeability coefficient in 1974he must have been evaluating columns that were packed toan interstitial fraction considerably higher than the externalporosity of the columns that were the subject of his 1972experiments Our problem however is still not resolvedbecause Halasz did not accurately determine the externalporosity of the columns under study in either the 1972 or 1974columns
We surmise that in their 1972 paper Halasz and hiscoauthors made use of the teaching of Giddings outlinedin his 1965 textbook which at the time of their 1972 paperwas seven years old Here is what they say ldquoexperiencehas shown that in a regular packed column with the sievefractions usual for chromatography 120576
0is independent of the
particle size if the particles are more or less spherical In anacceptable approximation 120576
0is equal to 04 In those columns
packed with nonporous supports 119906V and 119906 are identicalUsing porous supports (ie silica gel alumina chromosorbPorasil and so forth) 119906 = 4119865(1205871198892
119888120576119879) where 120576
119879is the
total porosity and indicates the pore volume of the supportrdquoIt will be appreciated that 119906V stands for interstitial velocity(our 120583
119894) and 119906 stands for mobile phase velocity (our 120583
119905) The
ldquoexperiencerdquo which Halasz and his coauthors are referringto concerning ldquocolumns packed with nonporous supportsalumina and chromosorbrdquo is that recorded by Giddings inTable 53-1 in his 1965 text Continuing to establish theinterrelationships inherent in the various entities involvedin chromatographic columns with respect to both interstitialandmobile phase velocity the authors conclude ldquoformally 119906V
Journal of Materials 11
can be calculated for a porous support assuming 1205760is equal
to 04rdquo (emphasis supplied) The authors then announce themethodology underlying their frame of reference with regardto column permeability as follows ldquowe determined in all ofour columns the 119906119906V ratio to be 056 with a differentialrefractometer Consequently the total porosity 120576
119905was equal
to 0714rdquo It will be appreciated that their ratio of 119906119906V is oneand the same as GiddingsrsquoΦ parameter as utilized by him inhis 1965 textbook and as we define hereinabove
We can now identify the origin of the discrepancybetween Halasz and Giddings and by extension betweenHalasz and Guiochon (and Neue) For even though Halaszmakes use of the results in Giddingsrsquo Table 53-1 he failsto appreciate the teaching inherent therein regarding theimportance of an accurate value for column external porositywhen using porous particles In establishing his values forΦGiddings was careful to ground his values for the externalporosity of columns packed with porous particles in acomparison of their measured pressure drops with measuredpressure drops in columns packed with nonporous particleswhere the values for (external) porositywere accurate becausethey are equal to the columnsrsquo total porosity
In their 1972 paper Halasz et al used a refractometer todetect the inert solute 119899-pentane in the mobile phase of 119899-heptaneThey injected the 119899-pentane into the flowing mobilephase the flow rate of which they presumably measured witha volumetric cylinder and stop watch Since the exit timeof the 119899-pentane peak was identified by the refractometertheir measurement of holdup volume was accurate andaccordingly so was their value of 0714 for 120576
119905(there is no
uncertainty related to column total porosity when using inertsolutes)This in turn led to an accurate value for their119906 termmobile phase velocity However since they did not measurethe interstitial velocity 119906V but rather used an estimate basedupon the measured flow rate and an ldquoassumedrdquo value of 04for external porosity their calculatedestimated value for 119906Vwas arbitrary This in turn means that the value of 056 fortheirΦ ratio was likewise arbitrary
In Table 3 herein we show the results for column number1 in the 1972 study which we designate as Column A
1 Note
that the authorsrsquo raw values correspond to a value of 500 for1198750 a valuewhich is obviously far too high As is plainly evident
in our Figures 2 and 3 this column data set does not conformin any reasonable fashion to the norms in our referencecharts Accordingly the data is badly flawed In our correcteddata for Column A
1in Table 3 herein we adjust the column
external porosity to the lowest value permitted in our frame ofreference (038) This correction results in a revised value forGiddingsrsquoΦparameter of 0532 In additionwe are compelledto adjust the particle downward (to 11 micrometers approx)as is also dictated by our frame of reference We justify ourlow corrected value for external porosity and our downwardadjustment for particle size on the aggressive ldquodry-packingrdquoprocess described by the authors ldquo[T] the best results wereachieved with the following method the column (id =2mm) 25 cm long was packed with 25 equal portions of theldquobrushrdquo After adding each portion the column was vibratedand afterward tapped on the floor and also vigorously tappedwith a rod from the siderdquo Based upon a column packing
experience of the principal author herein spanningmore than30 years we believe that these packing conditions generateda packed column with a low external porosity and also one inwhich the particles experienced significant breakage
The main differences between the columns in the 1972and 1974 papers concerned the particle sizes and themethodsused to pack the columns The particle diameters in the 1972study fell in a range of about 15ndash200 micrometers while thosein the 1974 study were much smaller being in the range of 5ndash40 micrometers approximately Moreover while the columnsin the 1972 study were dry-packed the columns in the 1974study were slurry-packed (what Halasz calls ldquothe balanceddensity packing methodrdquo) The authors of the 1974 paperrecite that their permeability results shown in their Table 4correspond to values for 120601
0of 1080 and values for 120601 of 910
In Table 3 herein we show the results for the 1974 study as anaverage value in our column designated as A
2 Note that the
raw data reported by the authors which again has the built-inassumption that the column external porosity was 04 placesthe value of 119875
0at 192 In our corrected values for Column A
2
we show that a value of 267 for 1198750places the column external
porosity at a value of 043 The ratio of the value for 120601 of 910when compared to the value for 120601 in the 1972 study of 2007is about 22 (2007910 = 22) which is in good agreementwith the authorsrsquo statement that the permeability of the 1972columns was worse by a factor of about 2 Again based onthe experience of the principal author herein in the packingof hundreds of thousands of commercially sold slurry packedcolumns we conclude that the balanced density techniquedescribed by the authors in the 1974 paper is fully capable ofproducing columns with this high and external porosity
In summary we conclude that for purposes of theKozeny-Blake equation (a) the external porosity of theHalasz 1972 columns was 038 (b) the external porosity of thecolumns which he tested in 1974 was 043 and (c) Halaszrsquos1972 and 1974 studies when properly analyzed corroborateGiddingsrsquo value for 119875
0of 267 not Carmanrsquos value of 180
Although the fault for all of the misdirection about the valueof the permeability coefficient being 1 times 10minus3 therefore restsprimarily on the shoulders of Halasz one has to question whyGuiochon (andNeue) so readily adopted itWhatevermay bethe reason Guiochonrsquos support for this proposition has beenand continues to be a cause for much of the conventionalacceptance of Carmanrsquos (erroneous) figure of 180 as thecharacteristic value of the coefficient in the Kozeny-Blakeequation
9 Guiochonrsquos Experimental Data
In addition to the data from Halasz 1972 and 1974 studiesTable 3 herein contains data from a representative sampleof Guiochonrsquos personal experimental permeability investi-gations over approximately the last decade as reported inthe numerous trade journals particularly the Journal ofChromatography A In each case the table has a single rowdedicated to Guiochonrsquos reported raw data and a single rowdedicated to corrected values for each column based on thereference chart in Table 1 herein
12 Journal of MaterialsTa
ble3Summaryexperim
entald
ata
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Naefe
andHalasz
columns
1198601(19
72)
Colum
nA1rawdata
07140
0560
040
000314
0523
782007
2811
4016
563
498119864minus04
500
357
908119864minus10
648119864minus10
2500
020
100
000135
0001350
00038
100
053
132
074
Colum
nA1corrected
07140
0532
03800
0334
0539
7813
3518
705002
701
749119864minus04
267
191
908119864minus10
648119864minus10
2500
020
100
000110
000110
100038
100
053
139
074
Endelean
dHalasz
columns
1198602(19
74)
Colum
nA2rawdata
08426
0475
040
00044
30738
11910
1080
4740
563
110119864minus03
192
162
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
033
016
Colum
nA2corrected
08426
0511
04309
0412
0723
11910
1080
3411
405
110119864minus03
267
225
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
031
016
Guiochon
columns
1198612ndash1198614(19
98)
Colum
nB 2
rawdata
05639
0722
040
730157
0264
22771
1367
2931
520
130119864minus03
263
148
467119864minus10
263119864minus10
429
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 3
rawdata
05582
0704
03932
0165
0272
44764
1369
3382
606
131119864minus03
226
126
471119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 4
rawdata
05509
0710
03909
0160
0263
72784
1422
3422
621
128119864minus03
229
126
459119864minus10
253119864minus10
1328
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 3
corrected
05653
0723
040
860157
0265
44774
1369
2899
513
129119864minus03
267
151
465119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 4
corrected
05651
0723
040
830157
0265
72776
1374
2907
514
129119864minus03
267
151
464119864minus10
262119864minus10
1375
500
100
000
060
000
0600
00165
98008
020
015
Guiochon
columnC
(1999)
Colum
nCrawdata
05930
0673
03990
0194
0323
183
865
1459
3372
569
116119864minus03
257
152
109119864minus09
645119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Colum
nCcorrected
05930
0673
03990
0194
0323
190
900
1518
3372
569
111119864minus03
267
158
105119864minus09
620119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Guiochon
columns
1198631ndash1198636(2006)
Colum
nD
1rawdata
07830
0456
03570
0426
0663
86694
886
7115
909
144119864minus03
975
76334119864minus10
261119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
013
Colum
nD
2rawdata
07230
0491
03550
0368
0571
88656
907
6726
930
152119864minus03
975
70353119864minus10
255119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
014
Colum
nD
3rawdata
06850
0506
03466
0338
0518
97685
1000
7025
1026
146119864minus03
975
67338119864minus10
231119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD
4rawdata
06560
0521
03415
0315
0478
103
696
1062
7143
1089
144119864minus03
975
64332119864minus10
218119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD5rawdata
06190
0545
03371
0282
0425
109
692
1118
7100
1147
144119864minus03
975
60334119864minus10
207119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
016
Colum
nD
6rawdata
05760
0587
03383
0238
0359
107
635
1103
6516
1131
157119864minus03
975
56364119864minus10
210119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
017
Colum
nD
1corrected
07830
0542
04243
0359
0623
86907
1158
3397
434
110119864minus03
267
209
334119864minus10
261119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
013
Colum
nD
2corrected
07230
0584
04221
0301
0521
88857
1186
3212
444
117119864minus03
267
193
353119864minus10
255119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
014
Colum
nD
3corrected
06850
0603
04129
0272
0463
97896
1307
3354
490
112119864minus03
267
183
338119864minus10
231119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
015
Colum
nD
4corrected
06560
0621
040
730249
0420
103
911
1388
3411
520
110119864minus03
267
175
332119864minus10
218119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
015
Colum
nD
5corrected
06190
0650
04025
0217
0362
109
905
1462
3390
548
110119864minus03
267
165
334119864minus10
207119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
016
Colum
nD
6corrected
05760
0701
04038
0172
0289
107
831
1442
3111
540
120119864minus03
267
154
364119864minus10
210119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
017
Guiochon
columns
1198641-1198642
(2007)
Colum
nE 1
rawdata
064
500594
03830
0262
0425
356
1119
1735
4371
678
894119864minus04
256
165
804119864minus11
519119864minus11
1500
046
100
000
030
000
0300
000
41300
030
079
047
Colum
nE 2
rawdata
05400
0800
04320
0108
0190
470
1000
1853
2161
400
100119864minus03
463
250
729119864minus11
393119864minus11
1500
046
100
000
027
000
0270
000
41300
030
070
056
Colum
nE 1
corrected
064
500594
040
000245
040
8356
969
1502
3628
563
103119864minus03
267
172
804119864minus11
519119864minus11
1500
046
100
000
028
000
0279
000
41300
030
075
047
Colum
nE 2
corrected
05400
0800
04320
0108
0190
470
577
1068
2161
400
173119864minus03
267
144
729119864minus11
393119864minus11
1500
046
088
000
023
000
0205
000
41300
030
070
056
Guiochon
columns
1198651ndash1198654
(2008)
Colum
nF 1
rawdata
06350
0586
03720
0263
0419
197
725
1142
4865
766
138119864minus03
149
95399119864minus11
253119864minus11
300
021
100
000
017
000
0170
00035
100
048
129
076
Colum
nF 2
rawdata
064
200581
03730
0269
0429
361
807
1258
4863
758
124119864minus03
166
107
358119864minus11
230119864minus11
500
021
100
000
017
000
0170
00035
100
048
129
075
Colum
nF 3
rawdata
064
100588
03770
0264
0424
670
748
1166
4643
724
134119864minus03
161
103
387119864minus11
248119864minus11
1000
021
100
000
017
000
0170
00035
100
048
128
075
Colum
nF 4
rawdata
06390
0595
03800
0259
0418
1014
752
1177
4476
701
133119864minus03
168
107
384119864minus11
246119864minus11
1500
021
100
000
017
000
0170
00035
100
048
127
075
Colum
nF 1
corrected
06350
0630
040
000235
0392
197
954
1502
3572
563
105119864minus03
267
170
399119864minus11
253119864minus11
300
021
100
000
019
000
0195
00035
100
048
120
076
Colum
nF 2
corrected
064
200623
040
000242
0403
361
964
1502
3611
563
104119864minus03
267
171
358119864minus11
230119864minus11
500
021
100
000
019
000
0186
00035
100
048
120
075
Colum
nF 3
corrected
064
100624
040
000241
0402
670
963
1502
360
6563
104119864minus03
267
171
386119864minus11
248119864minus11
1000
021
100
000
019
000
0193
00035
100
048
120
075
Colum
nF 4
corrected
06390
0626
040
000239
0398
1014
960
1502
3594
563
104119864minus03
267
171
384119864minus11
246119864minus11
1500
021
100
000
019
000
0192
00035
100
048
120
075
Journal of Materials 13
Table3Con
tinued
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Guiochon
columns
1198671-1198672serie
s(2010)
Colum
nH
1rawdata
05320
0744
03960
0136
0225
120
563
1057
3125
587
178119864minus03
180
96122119864minus10
652119864minus11
1500
046
100
000
026
000
0262
00104
050
005
013
009
Colum
nH
2rawdata
05420
0686
03720
0170
0271
61747
1379
4152
766
134119864minus03
180
98823119864minus11
446119864minus11
1000
046
100
000
025
000
0248
00054
050
005
013
009
Colum
nH
1corrected
05320
0752
040
000132
0220
120
799
1502
2993
563
125119864minus03
267
142
122119864minus10
652119864minus11
1500
046
100
000
031
000
0313
00104
050
005
013
009
Colum
nH
2corrected
05420
0738
040
000142
0237
61814
1502
3049
563
123119864minus03
267
145
823119864minus11
446119864minus11
1000
046
100
000
026
000
0259
00054
050
005
013
009
Guiochon
columns
1198673-1198674serie
s(2010)
Colum
nH
3rawdata
06270
060
903820
0245
0396
463
700
1117
4296
685
143119864minus03
163
102
447119864minus11
280119864minus11
1500
021
100
000
018
000
0177
00036
050
024
063
038
Colum
nH
4rawdata
05170
0791
040
900108
0183
433
616
1191
2639
511
162119864minus03
233
121
580119864minus11
300119864minus11
1500
021
100
000
019
000
0189
00036
050
024
059
047
Colum
nH
3corrected
06270
0638
040
000227
0378
463
942
1502
3527
563
106119864minus03
267
167
447119864minus11
280119864minus11
1500
021
100
000
021
000
0205
00036
050
024
060
038
Colum
nH
4corrected
05170
0791
040
900108
0183
433
699
1353
2639
511
143119864minus03
265
137
580119864minus11
300119864minus11
1500
021
100
000
020
000
0201
00036
050
024
059
047
Guiochon
columns
1198681-1198686
(2010)
Colum
nI 1rawdata
06580
0562
03700
0288
0457
488
741
1127
5156
784
135119864minus03
144
95390119864minus11
256119864minus11
500
021
100
000
017
000
0170
00104
050
024
065
037
Colum
nI 2rawdata
06550
0576
03770
0278
044
6873
661
1009
4745
724
151119864minus03
139
91437119864minus11
286119864minus11
1000
021
100
000
017
000
0170
00104
050
024
064
037
Colum
nI 3rawdata
06550
0588
03850
0270
0439
1209
610
931
4341
663
164119864minus03
141
92474119864minus11
310119864minus11
1500
021
100
000
017
000
0170
00104
050
024
062
037
Colum
nI 4rawdata
06430
0572
03680
0275
0435
260
787
1225
5153
801
127119864minus03
153
98367119864minus11
236119864minus11
500
030
100
000
017
000
0170
00104
050
012
032
018
Colum
nI 5rawdata
06540
0583
03810
0273
044
144
3683
1044
4531
693
146119864minus03
151
99423119864minus11
277119864minus11
1000
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 6rawdata
06550
0585
03830
0272
044
164
6665
1016
4438
678
150119864minus03
150
98434119864minus11
285119864minus11
1500
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 1corrected
06580
060
8040
000258
0430
488
988
1502
3701
563
101119864minus03
267
176
390119864minus11
256119864minus11
500
021
100
000
020
000
0196
00104
050
024
060
037
Colum
nI 2corrected
06550
0611
040
000255
0425
873
984
1502
3684
563
102119864minus03
267
175
437119864minus11
286119864minus11
1000
021
100
000
021
000
0207
00104
050
024
060
037
Colum
nI 3corrected
06550
0611
040
000255
0425
1209
984
1502
3684
563
102119864minus03
267
175
474119864minus11
310119864minus11
1500
021
100
000
022
000
0216
00104
050
024
060
037
Colum
nI 4corrected
06430
0622
040
000243
040
5260
966
1502
3617
563
104119864minus03
267
172
367119864minus11
236119864minus11
500
030
100
000
019
000
0188
00104
050
012
029
018
Colum
nI 5corrected
06540
0612
040
000254
0423
443
982
1502
3679
563
102119864minus03
267
175
423119864minus11
277119864minus11
1000
030
100
000
020
000
0204
00104
050
012
029
018
Colum
nI 6corrected
06550
0611
040
000255
0425
646
984
1502
3684
563
102119864minus03
267
175
434119864minus11
285119864minus11
1500
030
100
000
021
000
0207
00104
050
012
029
018
Guiochon
columns
1198691-1198692
(2011)
Colum
nJ 1rawdata90
∘A(120578
=00054
P)04980
0803
040
000098
0163
65654
1313
2801
563
153119864minus03
234
116126119864minus10
627119864minus11
1500
046
100
000
029
000
0287
00054
050
005
013
010
Colum
nJ 2rawdata90
∘A(120578
=00104
P)04980
0803
040
000098
0163
124
651
1307
2801
563
154119864minus03
232
116127119864minus10
630119864minus11
1500
046
100
000
029
000
0287
00104
050
005
013
010
Colum
nJ 1rawdata160
∘A(120578
=00054
P)05630
0714
04020
0161
0269
71896
1592
3099
550
112119864minus03
289
163
102119864minus10
573119864minus11
1500
046
100
000
030
000
0302
00054
050
005
012
009
Colum
nJ 2rawdata160
∘A(120578
=00104
P)05630
0714
04020
0161
0269
129
847
1504
3099
550
118119864minus03
273
154
108119864minus10
606119864minus11
1500
046
100
000
030
000
0302
00104
050
005
012
009
Colum
nJ 1corrected
04980
0803
040
000098
0163
65748
1502
2801
563
134119864minus03
267
133
126119864minus10
627119864minus11
1500
046
100
000
031
000
0307
00054
050
005
013
010
Colum
nJ 2corrected
04980
0803
040
000098
0163
124
748
1502
2801
563
134119864minus03
267
133
127119864minus10
630119864minus11
1500
046
100
000
031
000
0308
00104
050
005
013
010
Colum
nJ 1corrected
05630
0714
04020
0161
0269
71827
1470
3099
550
121119864minus03
267
150
102119864minus10
573119864minus11
1500
046
100
000
029
000
0290
00054
050
005
012
009
Colum
nJ 2corrected
05630
0714
04020
0161
0269
129
827
1470
3099
550
121119864minus03
267
150
108119864minus10
606119864minus11
1500
046
100
000
030
000
0299
00104
050
005
012
009
14 Journal of Materials
050
100150200250300350400
045 050 055 060 065 070 075 080 085
LineAll corrected data
Linear (all corrected data)Squares raw data Triangles corrected data
Pt
120576t
D seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
J1 J2 H4
H1H2
D6
B2 B3 B4
J3 J4
E2
C
H3
D5 D4
E1
D2
A1
D1
A2
y = 267xLine slope = 267
Figure 2 Summary of experimental results
0
500
1000
1500
2000
2500
20 25 30 35 40 45 50 55 60 65 70 75 80
Linear (line)
Ψ
LineAll corrected dataD seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
Line slope (P0) = 267
J1 J2
J3 J4E2
E1
A1
A2
H4 H1
C
120601
H2H3I1 I2 I3 I4 I5
F1 F2 F3 F4
D1 D2 D3 D4 D5 D6
Squares raw data Triangles corrected data
Figure 3 Summary of experimental results
Figures 2 and 3 herein are graphical representations ofTable 31015840s raw and corrected data for Guiochonrsquos columnsplotted in the same reference frames as in Figures 1(a) and1(b) respectively
As shown in the plots of the thirty-one total columnsevaluated in this data base only six columns had reportedraw data values for 119875
0close to the normThese were columns
B2 C E
1 H
4 J
1and J
2 Interestingly of the six conforming
columns 3 of the columns were packed with fully porousparticles and 3 columns with partially porous particles Itcan be seen however that when properly interpreted andadjusted all of Guiochonrsquos experimental data falls fairlywithin the norms outlined in the reference chart in Table 1In order to explore the reasons why the raw data does notconform more closely to the norms outlined in Table 1 eachof Guiochonrsquos studies included in Table 3 is evaluated in turn
Finally Table 4 herein is a list of symbols and parameterdefinitions used throughout this paper as well as their unitsof measure
91 Guiochonrsquos 1998 Study In a 1998 publication by Kohand Guiochon entitled ldquoEffect of the Column Length on theCharacteristics of the Packed Bed and the Column Efficiencyin a Dynamic Axial Compression Columnrdquo [19] Guiochonand his coauthor report the results of a study they hadcarried out on column packing using a rather novel techniquewhich was a combination of slurry packing and mechanicalcompression
The study involved the packing of a single cylinder ofdiameter 5 cm to various column lengths ranging between 1and 25 cm approximately The particles were all silica basedand coated with a reverse stationary phase C
18 However two
different categories of particles were involved One categorycontained highly irregular particles which were about 130micrometers in average diameter In addition these particleshad a high specific pore volume (11mLg) and exhibitedsignificant particle breakage under the packing conditionsinvestigated in the study Conversely the other particlecategory involved very small particles (6 micrometers) whichwere spherical in shape had a low specific pore volume(03mLg) and exhibited a greater ability to withstandparticle breakage under the studyrsquos packing conditions
The permeability data for the irregular particles isneglected herein because according to the authors theparticles exhibited significant breakage during packing andwithout an accurate value for the characteristic dimensionof the particle the data cannot be used in the Kozeny-Blake model to determine the value of 119875
0 The spherical
particles on the other hand behaved successfully under thepacking conditions chosen and yielded values calculated bythe authors for 119875
0of 619 268 226 and 229 for the four
different lengths of columns studied The data for thesecolumns is reported in our Table 3 in the rows for ColumnsB1through B
4 The higher value of 619 (the shortest column)
was rejected by the authors because they felt that either theparticles had been somewhat broken during packing or thecolumn frits had been blocked with fines Accordingly thepermeability data for this column is likewise neglected herein
The remaining three columns however were not con-sistent with respect to the authorsrsquo calculated values for 119875
0
Since the particles were identical the range of values for 1198750
of 226 to 268 must be due to experimental error The rawdata for column B
2 which produced a 119875
0value of 268 is
all self-consistent For example it generates a value for 119875119905
of 148 which is a reasonable value based upon the knownlow particle porosity for these Nova Pak particles and acorresponding value for 120576
0of 04073 which is a reasonable
value for a well-packed column However the other twocolumns columns B
3and B
4 had the much lower value of
126 for 119875119905 Accordingly in Table 3 herein corrected values
for column external porosity for columns B3and B
4are
displayed These corrected values are only slightly differentthan the reported values are within the experimental errorof the measurement and therefore in combination with the
Journal of Materials 15
Table 4 Glossary of terms
Symbol Unit dimensions (cgs) Definition119860 cm2 Column cross-sectional area119863 cm Column diameter119889119901
cm Particle spherical diameter equivalentΔ119875 gcmminus1secminus2 Pressure drop along column119889119901119898
cm Average particle diameter (measured)1198960
None Guiochonrsquos residual reciprocal permeability coefficient119875119905
None Guiochonrsquos modified permeability coefficient in Kozeny-Blake equation119871 cm Column length1198750
None Permeability coefficient in Kozeny-Blake equation119876 cm3minminus1 (exception) Volumetric fluid flow rate1199050
sec Time for elution of totally permeating unretained solute119896 cm2 General permeability coefficient in Darcy equation119870
119865cm2 Halaszrsquos permeability coefficient in Darcy equation (1974 paper)
119870 cm2 Halaszrsquos permeability coefficient in Darcy equation (1972 paper)119870
119888None Guiochonrsquos permeability coefficient in the Kozeny-Blake equation (as modified by Carmen)
Greek1205760
None External column porosity120576119894
None Internal column porosity120576119905
None Total column porosity120576119901
None Particle porosityΦ None Giddingsrsquo ratio of external and total column porosity120601 None Flow resistance parameter based on mobile phase velocity1206010
None Flow resistance parameter based on superficial velocityΩ
119901None Particle sphericity factor
120578 gcmminus1secminus1 Fluid absolute viscosity120583119894
cmsecminus1 Average fluid interstitial linear velocity120583119904
cmsecminus1 Average fluid superficial linear velocity120583119905
cmsecminus1 Average mobile phase velocityΨ None Porosity dependence term in the Kozeny-Blake equation based on mobile phase velocityΨ
0None Porosity dependence term in the Kozeny-Blake equation based on superficial velocity
raw data for column B2 are supportive of the value of 267 for
1198750
92 Guiochonrsquos 1999 Study In a 1999 publication by Farkaset al entitled ldquoValidity of Darcyrsquos Law at Low Flow Rates inLiquid Chromatographyrdquo [20] Guiochon and his coauthorsreport the results of some very exactingmeasurements whichthey made to validate Darcyrsquos law at extremely low fluidflow rates Using a column containing particles packed toa measured external column porosity of 0399 and havinga measured total column porosity of 0593 (Figure 2 in thepaper) Guiochon et al measured the column pressure dropand fluid flow rate at flow rates ranging between 0015 and05mLminwith ethylene glycol as the fluidThe authors statethat the particles in the column are spherical and they appearto have gone to extraordinary lengths to obtain an accuratemeasure of the particle diameter and of course as a result ofmodern techniques of particle size classification the particlesize distribution is very narrow
In Figure 1 of the paper the authors show a plot ofcolumn measured pressure drop in bar versus ldquofluid linearvelocityrdquo in cmsec This plot represents their measurementstaken in the empty column The only velocity that exists in
an empty column is the superficial velocity and thus theabscissa values in the plot in Figure 1 designated as ldquofluidlinear velocityrdquo represent superficial linear velocity
In Figure 2 in their paper however the authors showanother plot of measured pressure drop in bar also versusldquofluid linear velocityrdquo in cmsec The velocity on the abscissain this plot although also designated as ldquofluid linear velocityrdquodoes not equate (for permeability related purposes) to theldquofluid linear velocityrdquo in Figure 1 This is because in Figure 2the columnwas filled with porous particles and themeasuredvelocity was the mobile phase velocity Indeed the onlyvelocity used throughout the entire paper is the mobile phasevelocity (mobile phase velocity and superficial velocity areone and the same in an empty column)
In Table 1 of their paper the authors report that for thecolumn represented in Figure 2 the slope of a line achievedwhen the mobile phase velocityin units of cmsec is plottedversus pressure drop in units of bar is 1829 They also reportthat the flow resistance parameter 120601 for this column is 865In Table 3 herein it can be seen that the raw data for thiscolumn (Column C) generates an apparent value of 257 for1198750 This value is very close to Giddingsrsquo value of 267 Indeed
by referring to Figures 2 and 3 herein one can see that thereported raw data in this study without any modification
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
6 Journal of Materials
0200400600800
100012001400160018002000
20 30 40 50 60 70
NonporousLow porosityMedium porosity
CapillaryLinear (line)
Nonporous Low porosity particles
Medium porosity particles
High porosity particles
Capillary
Line slope (P0) = 267
y = 267x + 0
Ψ
120601
Line
High porosity
(a)
99111123135147159171183195207219231243255267
035 045 055 065 075 085 095
LineCapillaryLinear (line)
Capillary
Line slope (P0) = 267
y = 267x + 0
Pt
120576t
Φ = 10
Φ = 075
Φ = 062
Φ = 05
Φ = 075
Φ = 062
Φ = 05
Φ = 1
(b)
99111123135147159171183195207219231243255267
000 010 020 030 040 050 060 070 080 090 100
NonporousLow porosityMedium porosityHigh porosity
120576p
y = 160x + 107
Pt
LineCapillaryLinear (line)
(c)
Figure 1 (a) The balance between 120601 and Ψ (b) Column permeability reference chart (c) Particle permeability reference chart
In the 2006 edition of their textbook [9] Guiochon andhis coauthors provide the following worked example of hispressure dropfluid flow teaching3
In Table 2 a cross-reference is established between Guio-chonrsquos teaching for columns packed with both porous andnonporous particles the Kozeny-Blake equation (asmodifiedby Carman) and for validation purposes Giddingsrsquo teachingbased upon actual permeabilitymeasurements fromhis Table53-1 in his 1965 text [8]4
In his worked example Guiochonrsquos fluid velocity 119906is found in his text as his equation (256) [9] at page61 and equates to mobile phase not superficial velocityAccordingly when applied to columns packed with porousparticles his worked example combines the contributionto fluid velocity of mass transfer by molecular diffusionin the free space within the particles with mass transferby convection in the free space between the particles Asdiscussed above if one is attempting to compare Guiochonrsquos
teaching for the value of the permeability coefficient with thatof the Kozeny-Blake equation the conversion factor is thetotal porosity of the column
In addition because Guiochonrsquos particle sizeparameter has an embedded assumption due to particleshaperoughness his worked example also commingles thecontribution to spherical particle diameter equivalent ofparticle shaperoughness with other nonparticle variablesUnfortunately his worked example is not based uponcompletely spherical particles where this would not be anissue We know this because he states that 119896
0in his worked
example equals 1 times 10minus3 which he says is the value of thepermeability coefficient for columns packed with irregularparticles
In order to move forward one must simplify If this wereto be done experimentally it would best be accomplishedby experiments with nonporous particles in which case thevariable of molecular diffusion in the pores of the particles
Journal of Materials 7
Table2Cr
oss-referencefor
term
s119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119896119871
119863Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
120576119905Φ
(120576119905minus1205760)
120576119894
Δ119875119889119901
2Δ119875119889119901
2(1minus120576119905Φ)2120576119905(1minus1205760)2
1120601
1198750120576119905
119889119901
2120576119905
(119889119901119898Ω119901)
(1minus1205760)
120583119905120578119871
120583119904120578119871
(120576119905Φ)3
(1205760)3
120601Ψ
120601po
iseUnits
Non
eNon
eNon
eNon
eNon
epsi
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
cmNon
ecm
cmgcmminus1 secminus1cm
3minminus1cm
secminus
1cm
secminus
1cm
secminus
1
Nonporousp
articles
Gidding
srsquotradition
alno
nporou
scolum
n040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
1000010
00010
0010
0399
004
00100
0100
Guiocho
nrsquostradition
alspheric
alparticle(app
arent)
0362
1000
0362
000
0000
0187
830
2293
3106
858
120119864minus03
267
97436119864minus10
15046
1000010
00010
0010
0361
0036
0100
0100
Gidd
ingrsquos
table5
3-1
5060meshglassb
eads
0422
1000
0422
000
0000
0113
500
1184
1873
444
200119864minus03
267
113
844119864minus10
15046
1000010
00010
0010
0421
004
20100
0100
5060meshglassb
eads
040
910
00040
9000
0000
0126
560
1371
2097
513
179119864minus03
267
109
729119864minus10
15046
1000010
00010
0010
040
70041
0100
0100
Guiochonrsquosteaching
ExA
irregular
particle
(app
arent)
040
010
00040
0000
0000
0225
1000
2500
2250
563
100119864minus03
444
178
400119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
ExA
correctedforp
article
irregularity
040
010
00040
0000
0000
0225
601
1502
2250
563
166119864minus03
267
107
400119864minus10
15046
078
00010
000
080010
0399
004
00100
0100
ExB
spheric
alparticle
(app
arent)
040
010
00040
0000
0000
0187
830
2075
2250
563
120119864minus03
369
148
482119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
ExB
corrected
040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
Porous
particles
Gidding
srsquotradition
alpo
rous
column
060
0066
7040
00200
0333
203
900
1500
3375
563
111119864minus03
267
160
667119864minus10
15046
100
00010
00010
0010
0599
006
00150
0100
Gidd
ingrsquos
table5
3-1
3040meshalum
ina
0803
0498
040
00403
0672
271
1204
1499
4510
562
830119864minus04
267
214
667119864minus10
15046
100
00010
00010
0010
0801
0080
0201
0100
5060meshalum
ina
0837
0499
0417
0420
0721
235
1043
1246
3906
467
959119864minus04
267
224
803119864minus10
15046
100
00010
00010
0010
0835
0084
0201
0100
6080meshchromosorbW
(5DNP)
0766
0498
0382
0384
0621
316
1404
1833
5259
687
712119864minus04
267
204
545119864minus10
15046
100
00010
00010
0010
0764
0077
0201
0100
6080meshchromosorbW
(20
DNP)
0785
0495
0389
0396
064
8300
1333
1698
4991
636
750119864minus04
267
210
589119864minus10
15046
100
00010
00010
0010
0783
0079
0202
0100
8 Journal of Materials
would be eliminated and by using only smooth sphericalparticles such as glass beads in which case the variableof particle shaperoughness would be eliminated If both ofthese things were done one would be able to directly andunambiguously determine the true value of 119875
05
Accordingly since his textbook contains a general teach-ing without any expressed restrictions with respect to particleporosity let us first assume that the column in Guiochonrsquosworked example is packed with nonporous particles Sec-ondly let us assume that the value for the external porosityof the column (which is the same as the total porosity ina column packed with nonporous particles) is the valuetraditionally assumed for a typical well-packed column thatis 04
As illustrated by example A in Table 2 this would leadto an apparent value for 119875
0of 444 which is obviously much
too high6 Using the adjusted particle size of 8 micrometers(rounded) which is calculated based on a value of 078 forΩ
119901
we derive values of 267 for 1198750and 107 for 119875
119905(the relationship
between these two values is of course equal to the totalporosity of the column 120576
119905 as is dictated by (20))
Similarly as illustrated by example B in Table 2 when thecolumn of Guiochonrsquos worked example is adapted to accom-modate spherical nonporous particles Guiochonrsquos teachingof the value for 119896
0of 12 times 10minus3 results in an apparent value
of 369 for 1198750
7 again too high [10 11] On the other hand ifwe correct the value of 119875
0 we get a column external porosity
of 03620which is too low for a typical well-packed column(120576
0= 04) [8] Moreover in the case of both corrections
Guiochonrsquos teaching overstates the pressure gradient (187 psicompared to the 135 psi which Giddingsrsquo experiments wouldgenerate) Accordingly it is only when we correct for thepressure drop that we achieve appropriate values of 04 for1205760 267 for 119875
0 and 107 for 119875
119905
As is evident from Table 2 when the porosity of columnspacked with nonporous particles varies the value of 119875
119905also
varies because it is based upon the measurement of mobilephase velocity which in turn changeswith the porosity of thecolumn Indeed for typical columns packed with nonporousparticles having external porosities close to 04 (038ndash042)[8] Table 2 reveals that the value of 119875
119905varies by as much as 6
points while the value of 1198750remains constant at 267
Having established a range of values for 119875119905based upon
columns packed with nonporous particles one can nowproceed to evaluate the range of values for 119875
119905based upon
columns packed with porous particles which is to say thatwe can now evaluate the impact of particle porosity onpermeability Among other things Table 2 shows that whenit comes to porous particles Guiochonrsquos 119875
119905is even more
variable (and thus even less useful) because the value ofthe coefficient can differ on a column-to-column basis byas much as 90 points depending on the combination of theinternal andexternal column porosities at play
Figures 1(a) and 1(b) are a plot of the data found in Table 1In this plot Guiochonrsquos modified permeability coefficient 119875
119905
is plotted against column total porosity 120576119905 Once again it will
be appreciated that (a) all data falls on a line with a slope of267 which represents the value of 119875
0and (b) the line has no
intercept which means that when 120576119905is zero 119875
119905is also zero As
shown in the plot values of the parameter Φ are highest atlow values of 119875
119905and lowest at high values of 119875
119905 Further as is
evidenced by the respective definitions for the terms119875119905and119875
0
contained herein the only time that the value of Guiochonrsquos119875119905does not differ from the value of 119875
0(267) is when the value
of 120576119905is 1 its value in a theoretical packed column devoid of
particles that is acapillaryFinally Figure 1(c) is another plot of the data in Table 1
but this time the values for 119875119905are plotted against particle
porosity 120576119901 The equation of this line does have an intercept
which represents the value of 119875119905when the particles are
nonporous It is also obvious from the equation of this linethat it is only when particle porosity is unity that is in acapillary that the value of 119875
119905is 267
It will be appreciated that since 120576119905is the sumof the internal
and external porosities of a packed column any given valuefor 119875
119905involving a column packed with porous particles can
represent many different combinations of these parametersVice versa any given value for 120576
119905can generate many different
values for 119875119905depending on the amount of particle porosity
which is present On the other hand as is pointed outearlier column internal porosity andor particle porosityplay no role in determining the permeability of a packedcolumn the only porosity parameter which is relevant forsuch purposes is column external porosity In other words asa permeability coefficient 119875
119905is not a particularly informative
concept because it still fails to provide any meaningful tie tothe underlying physical phenomena that govern the pressuredropfluid flow relationship
In the case of permeability measurements one quan-tifies only total pressure drop Pressure drop however isa commingling of the contributions of the several differ-ent parameters embodied in the pressureflow relationshipTherefore one must accurately represent the contributionof each factor before adding them together to equal thepressure drop Obviously if one mistakes the contribution ofany given element for instance particle size this will skewthe contribution of some other element when summed forinstance column porosity
Our frame of reference has been calibratedwith respect tothe interrelationships between the values of 119875
0 119875
119905 120576
119905 and 120576
119901
based on the data reported in Giddingsrsquo table 53-1 in his 1965textbook Since his data base was derived from experimentsusing nonporous particles the permeability results of whichwere extended to columns packed with fully porous particlesby using his Φ parameter the validity of which in turn wasestablished by independent means there is no uncertaintywith respect to the critical variable of external porosity 120576
0
inherent in our frame of reference It is for this reasontherefore that our frame of reference trumps any reporteddata which uses any other technique including the techniqueof inverse size exclusion chromatography by itself to establishvalues for 120576
0[12] In addition our frame of reference is
internally consistent due to the grounding of the parametersat the boundary condition of a theoretical capillary at whichpoint the values of 120576
119905and 120576
119901are unity and the values of 119875
0and
119875119905are equal In essence these interrelationships exemplify the
conservation laws of nature which are the ultimate standard
Journal of Materials 9
of measure when considering partial fractions of a singleentity
Consequently as Figures 1(b) and 1(c) illustrate if oneknows based upon other independent means the value of acolumnrsquos total porosity or its particlesrsquo porosity respectivelyone can at least verify whether results of permeability exper-iments expressed in terms of 119875
119905are reasonable Or to put
it another way if the 119875119905results reported do not conform to
what these plots tell us one should expect from columns andparticles based upon independently derived porosity valuesfor the particles under study we know that something iswrong Accordingly this is the way in which we use 119875
119905below
to analyze and evaluate Guiochonrsquos experimental work of thelast decade
8 The Origin of Guiochonrsquos Value forColumn Permeability
In 1966 Guiochon wrote a review paper entitled ldquoProblemsRaised by the Operation of Gas Chromatographic Columnsrdquo[13] In that paper he reviewed the development of the per-meability equation from Darcy to the then present Amongother things he reported that ldquoa mean 120576 of 038 is foundfor a number of packings obtained from powdered prod-ucts of various origins consisting of spherical or irregularparticles Almost all the measured values are between035 and 045rdquo (p 14) He then went on to discuss what hecalled the ldquosemiempirical Blake-Kozeny equationrdquo which hereported as 119896 = [119889
2
1199011205762][150(1 minus 120576)
2] [13 p 15 Equation
(11)] What sticks out of course is the fact that Guiochonstates here that 119875
0equals 150 the value for the permeability
coefficient championed by Ergun not the value posited byCarman To be fair we should point out that Guiochongave warning that he was uncomfortable with any particularvalue for the permeability coefficient For even though herecited the Kozeny-Blake equation with a value of 150 for itspermeability coefficient he stated that ldquothe values of 119896 givenby (this) equation are not very accurate and the deviationscan be as large as 50rdquo [13 p 15] Guiochon attributedmost ofthis uncertainty to the inability of practitioners of that era toeffectively measure the value of a columnrsquos external porositySince as he pointed out ldquo(the Kozeny-Blake) equation isvery sensitive to variations in 120576rdquo he concluded that theequation is ldquoof little userdquo and that it was better practice torely upon grosser measurements of permeability [13 p 15]Thus Guiochon at this timeframe made a conscious decisionto staywithin the relatively safe but admittedlymore obscureboundaries of Darcyrsquos teaching on column permeability
In 2006 however Guiochon emerged from under theshadow of Darcy permeability when he published a reviewpaper which purported to be a comprehensive summaryof the current state of the art in chromatography [14]In light of the fact that Guiochon and his coauthors hadvirtually contemporaneously published a new version of theirtextbook one would have expected that Guiochon wouldtake this opportunity to reaffirm his textbook teaching So itshould come as no surprise that he would open his discussionof the pressure dropfluid flow relationshipwith the following
assertion ldquothe permeability of packed beds used in liquidchromatography is in the order of 1 times 10minus3rdquo [14 p 9] Asone would expect the authority which Guiochon cites forthis seemingly familiar proposition is the 2006 edition of hisown textbook For the first time however he also providessome citations to third party sources In particular he cites a1997 chromatography handbook written by Uwe Neue ChiefScientist for Waters Corporation [15 p 9] As one can seeNeuersquos equationmdashlike Guiochonrsquosmdashdoes indeed postulate apermeability coefficient with a value of 1 times 10minus3
At this point however Guiochon abandons his tradi-tional caution about the value of the permeability coeffi-cient and marries his permeability equation to the Kozeny-Blake equation thereby endorsing not only the porositydependence term in that equation but also a specific valuefor 119875
0 Referring to his value of 1 times 10minus3 he states the
following ldquothe specific column permeability is related to thecolumn porosity through the Kozeny-Carman equation 119896
119891=
12057630180 (1 minus 120576
0)2rdquo [14 p 9] Not surprisingly Neue makes
the same connection ldquothe specific permeability depends onthe particle size 119889
119901and the interstitial porosity 120576
0of the
packed bedThe relationship is known as theKozeny-Carmanequation 119861
0= (1185)(1205763
01198892119901(1 minus 120576
0)2)rdquo [15 p 30]
Neuersquos permeability equation and Guiochonrsquos textbookpermeability equation however differ in that Neue uses thesuperficial not the mobile phase velocity Accordingly sinceall other features of their equations are the same one wouldexpect them to get different values for their permeabilitycoefficients Neue himself recognizes this when he says inhis handbook ldquothis value [1000] is also in good agreementwith our experience but values as low as 500 have beenreported in the literature8 However there are several reasonswhy different values reported by different workers do notrepresent true differences in the resistance parameter Firstlower values are obtained for porous packings if the specificpermeability is measured on the basis of the breakthroughtime of an unretained peak instead of using the flow raterdquo see[15 p 32] (when he distinguishes the ldquobreakthrough time ofan unretained peakrdquo from the ldquoflow raterdquo Neue is of coursereferring to the mobile phase velocity and the superficialvelocity resp) Consequently if one takes Neuersquos point andturns it on its head one can see that if he and Guiochonwere to get the same value for their permeability coefficients[1000] but one (Guiochon) is using mobile phase velocityto derive it and the other (Neue) uses superficial velocity toderive it this would indeed represent a ldquotrue differencerdquo intheir equations
So how can Guiochon cite both Neue and himself for theproposition in his review paper that ldquothe permeability of thepacked beds used in liquid chromatography is in the order of1times 10minus3rdquoThe answer is that at least for purposes of his reviewpaper he adopts Kozeny-Blakersquos fluid velocity convention inplace of the velocity frame he uses in his textbook Here iswhat he says ldquothis approximation [1 times 10minus3] is valid only ifthe superficial velocity is usedrdquo [14 p 9] (emphasis supplied)
As a matter of pure mathematics Guiochonrsquos permeabil-ity coefficient of 1 times 10minus3 can only be equal to 1205763
0180(1 minus 120576
0)2
if the value of 1205760is 04 the typical value for the interstitial
10 Journal of Materials
fraction of a well-packed column Guiochon confirms in hisreview paper that this is indeed what he assumes in his state-ment of the Kozeny-Blake equation [14 p 9] Accordinglyhis adoption of the value of 180 for 119875
0solely depends on the
validity of his assertion that the value of the permeabilitycoefficient is 1 times 10minus3
We already know from our discussion of Guiochonrsquostextbook teaching that when this value is associated with apermeability equation based on mobile phase velocity it iswrong As reflected in our Table 2 when Guiochonrsquos workedexample involving this figure is corrected to account for theembedded factor due to the irregularity of the particles whichhis worked example assumes the value of the permeabilitycoefficient generated by his equation is not 1 times 10minus3 it is 166times 10minus3
So where did this value of 1 times 10minus3 come from AlthoughGuiochon in his review paper cites Neuersquos handbook for thisvalue Neue himself offers absolutely no authority for it Onthe other hand Guiochon does provide one other citation inhis review paper for this value The citation is to an article byHalasz et al published in 1975 [16]
If one then goes to that article one sees that Halaszet al do indeed proffer the same equation as Neue (albeitin a somewhat different format) and yes it does containthe value of 1 times 10minus3 See [16 Equation (1)] Neverthelessthe authors fail to identify from whence they obtained thisvalue On the other hand they do make reference to apaper that Endele et al wrote a year earlier in 1974 withKlaus Unger entitled ldquoInfluence of the Particle Size (5ndash35 120583)of Spherical Silica on Column Efficiencies in High-PressureLiquid Chromatographyrdquo [17] It appears that this is wherethe body is buried for it is in this article that Halasz reportshaving actually derived a value of 1 times 10minus3 for the permeabilitycoefficient
In summarizing the results of their experiments whichincidentally are based on measurements of superficial notmobile phase velocity (see their Equation (7)) Halasz et alcalculate an average value for the permeability coefficient(which they notate as119870
119865) of 1198892
1199011080 Accordingly they state
the following ldquo[I]t is proposed to define an average particlesize 119889
119901 for columns packed with spherical silica using the
balanced density packing method by the equation 1198892119901
=
1198701198651000 = 1198651205781198711031199032120587Δ119875rdquo [17 p 382 their Equation (7)]To recapitulate it thus appears that (1) Halasz was the
source of the value of 1 times 10minus3 for the permeability coefficientwhich Guiochon adopts in his review paper (2)Halaszrsquo valuewas based upon measurements which involved sphericalparticles and (3)Halaszrsquo value was properly derived from anequation which utilized the superficial velocity as the fluidvelocity
However we still have a problem if we are going to acceptthe value of 1 times 10minus3 as a basis for calculating the constantin Kozeny-Blake we still need to know the value of Halaszrsquocolumnsrsquo external porosity For example in the absenceof measured values for the external porosity Guiochonrsquosassertion in his review paper that the value of the constantis 180 is without foundation
Unfortunately however Halasz bases his methodologyon ldquoassumedrdquo versus ldquomeasuredrdquo values for external porosityNevertheless he does provide some clues that may helpus peel back this onion First he says that his value forthe permeability coefficient applies to ldquocolumns using thebalanced density packing methodrdquo Secondly after statinghis proposed value for the coefficient we find the followingtelltale sentence ldquothe permeability of columns packed by theconventional ldquodryrdquo method are worse by a factor of about 2than those packed by the balanced density methodrdquo [17 p382] For this proposition Halasz cites yet another of his ownarticles this time a paper written by himself and M Naefein 1972 entitled ldquoInfluence of Column Parameters on PeakBroadening in High-Pressure Liquid Chromatographyrdquo [18]
When one follows this thread by going to Halaszrsquos 1972paper one finds a possible solution to the puzzle For this iswhere Halasz describes his calculations of the permeabilitycoefficient from experiments with a number of conventionaldry-packed columns containing spherical silica particlesNot surprisingly the resulting value for the permeabilitycoefficient for such columns was not 1 times 10minus3 Instead Halaszsays that his experiments with such columns resulted inthe following value for the permeability coefficient ldquo119870 sim
11988921199012000rdquo [18 p 78] In other words consistent with what
Halasz says in his 1974 article the value for the coefficient forthese dry-packed columns was indeed ldquoworse by a factor ofabout 2rdquo compared to its value for columns prepared with thebalanced density method
The obvious implication of this is that when Halasz gota value of 1 times 10minus3 for the permeability coefficient in 1974he must have been evaluating columns that were packed toan interstitial fraction considerably higher than the externalporosity of the columns that were the subject of his 1972experiments Our problem however is still not resolvedbecause Halasz did not accurately determine the externalporosity of the columns under study in either the 1972 or 1974columns
We surmise that in their 1972 paper Halasz and hiscoauthors made use of the teaching of Giddings outlinedin his 1965 textbook which at the time of their 1972 paperwas seven years old Here is what they say ldquoexperiencehas shown that in a regular packed column with the sievefractions usual for chromatography 120576
0is independent of the
particle size if the particles are more or less spherical In anacceptable approximation 120576
0is equal to 04 In those columns
packed with nonporous supports 119906V and 119906 are identicalUsing porous supports (ie silica gel alumina chromosorbPorasil and so forth) 119906 = 4119865(1205871198892
119888120576119879) where 120576
119879is the
total porosity and indicates the pore volume of the supportrdquoIt will be appreciated that 119906V stands for interstitial velocity(our 120583
119894) and 119906 stands for mobile phase velocity (our 120583
119905) The
ldquoexperiencerdquo which Halasz and his coauthors are referringto concerning ldquocolumns packed with nonporous supportsalumina and chromosorbrdquo is that recorded by Giddings inTable 53-1 in his 1965 text Continuing to establish theinterrelationships inherent in the various entities involvedin chromatographic columns with respect to both interstitialandmobile phase velocity the authors conclude ldquoformally 119906V
Journal of Materials 11
can be calculated for a porous support assuming 1205760is equal
to 04rdquo (emphasis supplied) The authors then announce themethodology underlying their frame of reference with regardto column permeability as follows ldquowe determined in all ofour columns the 119906119906V ratio to be 056 with a differentialrefractometer Consequently the total porosity 120576
119905was equal
to 0714rdquo It will be appreciated that their ratio of 119906119906V is oneand the same as GiddingsrsquoΦ parameter as utilized by him inhis 1965 textbook and as we define hereinabove
We can now identify the origin of the discrepancybetween Halasz and Giddings and by extension betweenHalasz and Guiochon (and Neue) For even though Halaszmakes use of the results in Giddingsrsquo Table 53-1 he failsto appreciate the teaching inherent therein regarding theimportance of an accurate value for column external porositywhen using porous particles In establishing his values forΦGiddings was careful to ground his values for the externalporosity of columns packed with porous particles in acomparison of their measured pressure drops with measuredpressure drops in columns packed with nonporous particleswhere the values for (external) porositywere accurate becausethey are equal to the columnsrsquo total porosity
In their 1972 paper Halasz et al used a refractometer todetect the inert solute 119899-pentane in the mobile phase of 119899-heptaneThey injected the 119899-pentane into the flowing mobilephase the flow rate of which they presumably measured witha volumetric cylinder and stop watch Since the exit timeof the 119899-pentane peak was identified by the refractometertheir measurement of holdup volume was accurate andaccordingly so was their value of 0714 for 120576
119905(there is no
uncertainty related to column total porosity when using inertsolutes)This in turn led to an accurate value for their119906 termmobile phase velocity However since they did not measurethe interstitial velocity 119906V but rather used an estimate basedupon the measured flow rate and an ldquoassumedrdquo value of 04for external porosity their calculatedestimated value for 119906Vwas arbitrary This in turn means that the value of 056 fortheirΦ ratio was likewise arbitrary
In Table 3 herein we show the results for column number1 in the 1972 study which we designate as Column A
1 Note
that the authorsrsquo raw values correspond to a value of 500 for1198750 a valuewhich is obviously far too high As is plainly evident
in our Figures 2 and 3 this column data set does not conformin any reasonable fashion to the norms in our referencecharts Accordingly the data is badly flawed In our correcteddata for Column A
1in Table 3 herein we adjust the column
external porosity to the lowest value permitted in our frame ofreference (038) This correction results in a revised value forGiddingsrsquoΦparameter of 0532 In additionwe are compelledto adjust the particle downward (to 11 micrometers approx)as is also dictated by our frame of reference We justify ourlow corrected value for external porosity and our downwardadjustment for particle size on the aggressive ldquodry-packingrdquoprocess described by the authors ldquo[T] the best results wereachieved with the following method the column (id =2mm) 25 cm long was packed with 25 equal portions of theldquobrushrdquo After adding each portion the column was vibratedand afterward tapped on the floor and also vigorously tappedwith a rod from the siderdquo Based upon a column packing
experience of the principal author herein spanningmore than30 years we believe that these packing conditions generateda packed column with a low external porosity and also one inwhich the particles experienced significant breakage
The main differences between the columns in the 1972and 1974 papers concerned the particle sizes and themethodsused to pack the columns The particle diameters in the 1972study fell in a range of about 15ndash200 micrometers while thosein the 1974 study were much smaller being in the range of 5ndash40 micrometers approximately Moreover while the columnsin the 1972 study were dry-packed the columns in the 1974study were slurry-packed (what Halasz calls ldquothe balanceddensity packing methodrdquo) The authors of the 1974 paperrecite that their permeability results shown in their Table 4correspond to values for 120601
0of 1080 and values for 120601 of 910
In Table 3 herein we show the results for the 1974 study as anaverage value in our column designated as A
2 Note that the
raw data reported by the authors which again has the built-inassumption that the column external porosity was 04 placesthe value of 119875
0at 192 In our corrected values for Column A
2
we show that a value of 267 for 1198750places the column external
porosity at a value of 043 The ratio of the value for 120601 of 910when compared to the value for 120601 in the 1972 study of 2007is about 22 (2007910 = 22) which is in good agreementwith the authorsrsquo statement that the permeability of the 1972columns was worse by a factor of about 2 Again based onthe experience of the principal author herein in the packingof hundreds of thousands of commercially sold slurry packedcolumns we conclude that the balanced density techniquedescribed by the authors in the 1974 paper is fully capable ofproducing columns with this high and external porosity
In summary we conclude that for purposes of theKozeny-Blake equation (a) the external porosity of theHalasz 1972 columns was 038 (b) the external porosity of thecolumns which he tested in 1974 was 043 and (c) Halaszrsquos1972 and 1974 studies when properly analyzed corroborateGiddingsrsquo value for 119875
0of 267 not Carmanrsquos value of 180
Although the fault for all of the misdirection about the valueof the permeability coefficient being 1 times 10minus3 therefore restsprimarily on the shoulders of Halasz one has to question whyGuiochon (andNeue) so readily adopted itWhatevermay bethe reason Guiochonrsquos support for this proposition has beenand continues to be a cause for much of the conventionalacceptance of Carmanrsquos (erroneous) figure of 180 as thecharacteristic value of the coefficient in the Kozeny-Blakeequation
9 Guiochonrsquos Experimental Data
In addition to the data from Halasz 1972 and 1974 studiesTable 3 herein contains data from a representative sampleof Guiochonrsquos personal experimental permeability investi-gations over approximately the last decade as reported inthe numerous trade journals particularly the Journal ofChromatography A In each case the table has a single rowdedicated to Guiochonrsquos reported raw data and a single rowdedicated to corrected values for each column based on thereference chart in Table 1 herein
12 Journal of MaterialsTa
ble3Summaryexperim
entald
ata
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Naefe
andHalasz
columns
1198601(19
72)
Colum
nA1rawdata
07140
0560
040
000314
0523
782007
2811
4016
563
498119864minus04
500
357
908119864minus10
648119864minus10
2500
020
100
000135
0001350
00038
100
053
132
074
Colum
nA1corrected
07140
0532
03800
0334
0539
7813
3518
705002
701
749119864minus04
267
191
908119864minus10
648119864minus10
2500
020
100
000110
000110
100038
100
053
139
074
Endelean
dHalasz
columns
1198602(19
74)
Colum
nA2rawdata
08426
0475
040
00044
30738
11910
1080
4740
563
110119864minus03
192
162
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
033
016
Colum
nA2corrected
08426
0511
04309
0412
0723
11910
1080
3411
405
110119864minus03
267
225
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
031
016
Guiochon
columns
1198612ndash1198614(19
98)
Colum
nB 2
rawdata
05639
0722
040
730157
0264
22771
1367
2931
520
130119864minus03
263
148
467119864minus10
263119864minus10
429
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 3
rawdata
05582
0704
03932
0165
0272
44764
1369
3382
606
131119864minus03
226
126
471119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 4
rawdata
05509
0710
03909
0160
0263
72784
1422
3422
621
128119864minus03
229
126
459119864minus10
253119864minus10
1328
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 3
corrected
05653
0723
040
860157
0265
44774
1369
2899
513
129119864minus03
267
151
465119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 4
corrected
05651
0723
040
830157
0265
72776
1374
2907
514
129119864minus03
267
151
464119864minus10
262119864minus10
1375
500
100
000
060
000
0600
00165
98008
020
015
Guiochon
columnC
(1999)
Colum
nCrawdata
05930
0673
03990
0194
0323
183
865
1459
3372
569
116119864minus03
257
152
109119864minus09
645119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Colum
nCcorrected
05930
0673
03990
0194
0323
190
900
1518
3372
569
111119864minus03
267
158
105119864minus09
620119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Guiochon
columns
1198631ndash1198636(2006)
Colum
nD
1rawdata
07830
0456
03570
0426
0663
86694
886
7115
909
144119864minus03
975
76334119864minus10
261119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
013
Colum
nD
2rawdata
07230
0491
03550
0368
0571
88656
907
6726
930
152119864minus03
975
70353119864minus10
255119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
014
Colum
nD
3rawdata
06850
0506
03466
0338
0518
97685
1000
7025
1026
146119864minus03
975
67338119864minus10
231119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD
4rawdata
06560
0521
03415
0315
0478
103
696
1062
7143
1089
144119864minus03
975
64332119864minus10
218119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD5rawdata
06190
0545
03371
0282
0425
109
692
1118
7100
1147
144119864minus03
975
60334119864minus10
207119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
016
Colum
nD
6rawdata
05760
0587
03383
0238
0359
107
635
1103
6516
1131
157119864minus03
975
56364119864minus10
210119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
017
Colum
nD
1corrected
07830
0542
04243
0359
0623
86907
1158
3397
434
110119864minus03
267
209
334119864minus10
261119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
013
Colum
nD
2corrected
07230
0584
04221
0301
0521
88857
1186
3212
444
117119864minus03
267
193
353119864minus10
255119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
014
Colum
nD
3corrected
06850
0603
04129
0272
0463
97896
1307
3354
490
112119864minus03
267
183
338119864minus10
231119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
015
Colum
nD
4corrected
06560
0621
040
730249
0420
103
911
1388
3411
520
110119864minus03
267
175
332119864minus10
218119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
015
Colum
nD
5corrected
06190
0650
04025
0217
0362
109
905
1462
3390
548
110119864minus03
267
165
334119864minus10
207119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
016
Colum
nD
6corrected
05760
0701
04038
0172
0289
107
831
1442
3111
540
120119864minus03
267
154
364119864minus10
210119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
017
Guiochon
columns
1198641-1198642
(2007)
Colum
nE 1
rawdata
064
500594
03830
0262
0425
356
1119
1735
4371
678
894119864minus04
256
165
804119864minus11
519119864minus11
1500
046
100
000
030
000
0300
000
41300
030
079
047
Colum
nE 2
rawdata
05400
0800
04320
0108
0190
470
1000
1853
2161
400
100119864minus03
463
250
729119864minus11
393119864minus11
1500
046
100
000
027
000
0270
000
41300
030
070
056
Colum
nE 1
corrected
064
500594
040
000245
040
8356
969
1502
3628
563
103119864minus03
267
172
804119864minus11
519119864minus11
1500
046
100
000
028
000
0279
000
41300
030
075
047
Colum
nE 2
corrected
05400
0800
04320
0108
0190
470
577
1068
2161
400
173119864minus03
267
144
729119864minus11
393119864minus11
1500
046
088
000
023
000
0205
000
41300
030
070
056
Guiochon
columns
1198651ndash1198654
(2008)
Colum
nF 1
rawdata
06350
0586
03720
0263
0419
197
725
1142
4865
766
138119864minus03
149
95399119864minus11
253119864minus11
300
021
100
000
017
000
0170
00035
100
048
129
076
Colum
nF 2
rawdata
064
200581
03730
0269
0429
361
807
1258
4863
758
124119864minus03
166
107
358119864minus11
230119864minus11
500
021
100
000
017
000
0170
00035
100
048
129
075
Colum
nF 3
rawdata
064
100588
03770
0264
0424
670
748
1166
4643
724
134119864minus03
161
103
387119864minus11
248119864minus11
1000
021
100
000
017
000
0170
00035
100
048
128
075
Colum
nF 4
rawdata
06390
0595
03800
0259
0418
1014
752
1177
4476
701
133119864minus03
168
107
384119864minus11
246119864minus11
1500
021
100
000
017
000
0170
00035
100
048
127
075
Colum
nF 1
corrected
06350
0630
040
000235
0392
197
954
1502
3572
563
105119864minus03
267
170
399119864minus11
253119864minus11
300
021
100
000
019
000
0195
00035
100
048
120
076
Colum
nF 2
corrected
064
200623
040
000242
0403
361
964
1502
3611
563
104119864minus03
267
171
358119864minus11
230119864minus11
500
021
100
000
019
000
0186
00035
100
048
120
075
Colum
nF 3
corrected
064
100624
040
000241
0402
670
963
1502
360
6563
104119864minus03
267
171
386119864minus11
248119864minus11
1000
021
100
000
019
000
0193
00035
100
048
120
075
Colum
nF 4
corrected
06390
0626
040
000239
0398
1014
960
1502
3594
563
104119864minus03
267
171
384119864minus11
246119864minus11
1500
021
100
000
019
000
0192
00035
100
048
120
075
Journal of Materials 13
Table3Con
tinued
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Guiochon
columns
1198671-1198672serie
s(2010)
Colum
nH
1rawdata
05320
0744
03960
0136
0225
120
563
1057
3125
587
178119864minus03
180
96122119864minus10
652119864minus11
1500
046
100
000
026
000
0262
00104
050
005
013
009
Colum
nH
2rawdata
05420
0686
03720
0170
0271
61747
1379
4152
766
134119864minus03
180
98823119864minus11
446119864minus11
1000
046
100
000
025
000
0248
00054
050
005
013
009
Colum
nH
1corrected
05320
0752
040
000132
0220
120
799
1502
2993
563
125119864minus03
267
142
122119864minus10
652119864minus11
1500
046
100
000
031
000
0313
00104
050
005
013
009
Colum
nH
2corrected
05420
0738
040
000142
0237
61814
1502
3049
563
123119864minus03
267
145
823119864minus11
446119864minus11
1000
046
100
000
026
000
0259
00054
050
005
013
009
Guiochon
columns
1198673-1198674serie
s(2010)
Colum
nH
3rawdata
06270
060
903820
0245
0396
463
700
1117
4296
685
143119864minus03
163
102
447119864minus11
280119864minus11
1500
021
100
000
018
000
0177
00036
050
024
063
038
Colum
nH
4rawdata
05170
0791
040
900108
0183
433
616
1191
2639
511
162119864minus03
233
121
580119864minus11
300119864minus11
1500
021
100
000
019
000
0189
00036
050
024
059
047
Colum
nH
3corrected
06270
0638
040
000227
0378
463
942
1502
3527
563
106119864minus03
267
167
447119864minus11
280119864minus11
1500
021
100
000
021
000
0205
00036
050
024
060
038
Colum
nH
4corrected
05170
0791
040
900108
0183
433
699
1353
2639
511
143119864minus03
265
137
580119864minus11
300119864minus11
1500
021
100
000
020
000
0201
00036
050
024
059
047
Guiochon
columns
1198681-1198686
(2010)
Colum
nI 1rawdata
06580
0562
03700
0288
0457
488
741
1127
5156
784
135119864minus03
144
95390119864minus11
256119864minus11
500
021
100
000
017
000
0170
00104
050
024
065
037
Colum
nI 2rawdata
06550
0576
03770
0278
044
6873
661
1009
4745
724
151119864minus03
139
91437119864minus11
286119864minus11
1000
021
100
000
017
000
0170
00104
050
024
064
037
Colum
nI 3rawdata
06550
0588
03850
0270
0439
1209
610
931
4341
663
164119864minus03
141
92474119864minus11
310119864minus11
1500
021
100
000
017
000
0170
00104
050
024
062
037
Colum
nI 4rawdata
06430
0572
03680
0275
0435
260
787
1225
5153
801
127119864minus03
153
98367119864minus11
236119864minus11
500
030
100
000
017
000
0170
00104
050
012
032
018
Colum
nI 5rawdata
06540
0583
03810
0273
044
144
3683
1044
4531
693
146119864minus03
151
99423119864minus11
277119864minus11
1000
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 6rawdata
06550
0585
03830
0272
044
164
6665
1016
4438
678
150119864minus03
150
98434119864minus11
285119864minus11
1500
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 1corrected
06580
060
8040
000258
0430
488
988
1502
3701
563
101119864minus03
267
176
390119864minus11
256119864minus11
500
021
100
000
020
000
0196
00104
050
024
060
037
Colum
nI 2corrected
06550
0611
040
000255
0425
873
984
1502
3684
563
102119864minus03
267
175
437119864minus11
286119864minus11
1000
021
100
000
021
000
0207
00104
050
024
060
037
Colum
nI 3corrected
06550
0611
040
000255
0425
1209
984
1502
3684
563
102119864minus03
267
175
474119864minus11
310119864minus11
1500
021
100
000
022
000
0216
00104
050
024
060
037
Colum
nI 4corrected
06430
0622
040
000243
040
5260
966
1502
3617
563
104119864minus03
267
172
367119864minus11
236119864minus11
500
030
100
000
019
000
0188
00104
050
012
029
018
Colum
nI 5corrected
06540
0612
040
000254
0423
443
982
1502
3679
563
102119864minus03
267
175
423119864minus11
277119864minus11
1000
030
100
000
020
000
0204
00104
050
012
029
018
Colum
nI 6corrected
06550
0611
040
000255
0425
646
984
1502
3684
563
102119864minus03
267
175
434119864minus11
285119864minus11
1500
030
100
000
021
000
0207
00104
050
012
029
018
Guiochon
columns
1198691-1198692
(2011)
Colum
nJ 1rawdata90
∘A(120578
=00054
P)04980
0803
040
000098
0163
65654
1313
2801
563
153119864minus03
234
116126119864minus10
627119864minus11
1500
046
100
000
029
000
0287
00054
050
005
013
010
Colum
nJ 2rawdata90
∘A(120578
=00104
P)04980
0803
040
000098
0163
124
651
1307
2801
563
154119864minus03
232
116127119864minus10
630119864minus11
1500
046
100
000
029
000
0287
00104
050
005
013
010
Colum
nJ 1rawdata160
∘A(120578
=00054
P)05630
0714
04020
0161
0269
71896
1592
3099
550
112119864minus03
289
163
102119864minus10
573119864minus11
1500
046
100
000
030
000
0302
00054
050
005
012
009
Colum
nJ 2rawdata160
∘A(120578
=00104
P)05630
0714
04020
0161
0269
129
847
1504
3099
550
118119864minus03
273
154
108119864minus10
606119864minus11
1500
046
100
000
030
000
0302
00104
050
005
012
009
Colum
nJ 1corrected
04980
0803
040
000098
0163
65748
1502
2801
563
134119864minus03
267
133
126119864minus10
627119864minus11
1500
046
100
000
031
000
0307
00054
050
005
013
010
Colum
nJ 2corrected
04980
0803
040
000098
0163
124
748
1502
2801
563
134119864minus03
267
133
127119864minus10
630119864minus11
1500
046
100
000
031
000
0308
00104
050
005
013
010
Colum
nJ 1corrected
05630
0714
04020
0161
0269
71827
1470
3099
550
121119864minus03
267
150
102119864minus10
573119864minus11
1500
046
100
000
029
000
0290
00054
050
005
012
009
Colum
nJ 2corrected
05630
0714
04020
0161
0269
129
827
1470
3099
550
121119864minus03
267
150
108119864minus10
606119864minus11
1500
046
100
000
030
000
0299
00104
050
005
012
009
14 Journal of Materials
050
100150200250300350400
045 050 055 060 065 070 075 080 085
LineAll corrected data
Linear (all corrected data)Squares raw data Triangles corrected data
Pt
120576t
D seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
J1 J2 H4
H1H2
D6
B2 B3 B4
J3 J4
E2
C
H3
D5 D4
E1
D2
A1
D1
A2
y = 267xLine slope = 267
Figure 2 Summary of experimental results
0
500
1000
1500
2000
2500
20 25 30 35 40 45 50 55 60 65 70 75 80
Linear (line)
Ψ
LineAll corrected dataD seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
Line slope (P0) = 267
J1 J2
J3 J4E2
E1
A1
A2
H4 H1
C
120601
H2H3I1 I2 I3 I4 I5
F1 F2 F3 F4
D1 D2 D3 D4 D5 D6
Squares raw data Triangles corrected data
Figure 3 Summary of experimental results
Figures 2 and 3 herein are graphical representations ofTable 31015840s raw and corrected data for Guiochonrsquos columnsplotted in the same reference frames as in Figures 1(a) and1(b) respectively
As shown in the plots of the thirty-one total columnsevaluated in this data base only six columns had reportedraw data values for 119875
0close to the normThese were columns
B2 C E
1 H
4 J
1and J
2 Interestingly of the six conforming
columns 3 of the columns were packed with fully porousparticles and 3 columns with partially porous particles Itcan be seen however that when properly interpreted andadjusted all of Guiochonrsquos experimental data falls fairlywithin the norms outlined in the reference chart in Table 1In order to explore the reasons why the raw data does notconform more closely to the norms outlined in Table 1 eachof Guiochonrsquos studies included in Table 3 is evaluated in turn
Finally Table 4 herein is a list of symbols and parameterdefinitions used throughout this paper as well as their unitsof measure
91 Guiochonrsquos 1998 Study In a 1998 publication by Kohand Guiochon entitled ldquoEffect of the Column Length on theCharacteristics of the Packed Bed and the Column Efficiencyin a Dynamic Axial Compression Columnrdquo [19] Guiochonand his coauthor report the results of a study they hadcarried out on column packing using a rather novel techniquewhich was a combination of slurry packing and mechanicalcompression
The study involved the packing of a single cylinder ofdiameter 5 cm to various column lengths ranging between 1and 25 cm approximately The particles were all silica basedand coated with a reverse stationary phase C
18 However two
different categories of particles were involved One categorycontained highly irregular particles which were about 130micrometers in average diameter In addition these particleshad a high specific pore volume (11mLg) and exhibitedsignificant particle breakage under the packing conditionsinvestigated in the study Conversely the other particlecategory involved very small particles (6 micrometers) whichwere spherical in shape had a low specific pore volume(03mLg) and exhibited a greater ability to withstandparticle breakage under the studyrsquos packing conditions
The permeability data for the irregular particles isneglected herein because according to the authors theparticles exhibited significant breakage during packing andwithout an accurate value for the characteristic dimensionof the particle the data cannot be used in the Kozeny-Blake model to determine the value of 119875
0 The spherical
particles on the other hand behaved successfully under thepacking conditions chosen and yielded values calculated bythe authors for 119875
0of 619 268 226 and 229 for the four
different lengths of columns studied The data for thesecolumns is reported in our Table 3 in the rows for ColumnsB1through B
4 The higher value of 619 (the shortest column)
was rejected by the authors because they felt that either theparticles had been somewhat broken during packing or thecolumn frits had been blocked with fines Accordingly thepermeability data for this column is likewise neglected herein
The remaining three columns however were not con-sistent with respect to the authorsrsquo calculated values for 119875
0
Since the particles were identical the range of values for 1198750
of 226 to 268 must be due to experimental error The rawdata for column B
2 which produced a 119875
0value of 268 is
all self-consistent For example it generates a value for 119875119905
of 148 which is a reasonable value based upon the knownlow particle porosity for these Nova Pak particles and acorresponding value for 120576
0of 04073 which is a reasonable
value for a well-packed column However the other twocolumns columns B
3and B
4 had the much lower value of
126 for 119875119905 Accordingly in Table 3 herein corrected values
for column external porosity for columns B3and B
4are
displayed These corrected values are only slightly differentthan the reported values are within the experimental errorof the measurement and therefore in combination with the
Journal of Materials 15
Table 4 Glossary of terms
Symbol Unit dimensions (cgs) Definition119860 cm2 Column cross-sectional area119863 cm Column diameter119889119901
cm Particle spherical diameter equivalentΔ119875 gcmminus1secminus2 Pressure drop along column119889119901119898
cm Average particle diameter (measured)1198960
None Guiochonrsquos residual reciprocal permeability coefficient119875119905
None Guiochonrsquos modified permeability coefficient in Kozeny-Blake equation119871 cm Column length1198750
None Permeability coefficient in Kozeny-Blake equation119876 cm3minminus1 (exception) Volumetric fluid flow rate1199050
sec Time for elution of totally permeating unretained solute119896 cm2 General permeability coefficient in Darcy equation119870
119865cm2 Halaszrsquos permeability coefficient in Darcy equation (1974 paper)
119870 cm2 Halaszrsquos permeability coefficient in Darcy equation (1972 paper)119870
119888None Guiochonrsquos permeability coefficient in the Kozeny-Blake equation (as modified by Carmen)
Greek1205760
None External column porosity120576119894
None Internal column porosity120576119905
None Total column porosity120576119901
None Particle porosityΦ None Giddingsrsquo ratio of external and total column porosity120601 None Flow resistance parameter based on mobile phase velocity1206010
None Flow resistance parameter based on superficial velocityΩ
119901None Particle sphericity factor
120578 gcmminus1secminus1 Fluid absolute viscosity120583119894
cmsecminus1 Average fluid interstitial linear velocity120583119904
cmsecminus1 Average fluid superficial linear velocity120583119905
cmsecminus1 Average mobile phase velocityΨ None Porosity dependence term in the Kozeny-Blake equation based on mobile phase velocityΨ
0None Porosity dependence term in the Kozeny-Blake equation based on superficial velocity
raw data for column B2 are supportive of the value of 267 for
1198750
92 Guiochonrsquos 1999 Study In a 1999 publication by Farkaset al entitled ldquoValidity of Darcyrsquos Law at Low Flow Rates inLiquid Chromatographyrdquo [20] Guiochon and his coauthorsreport the results of some very exactingmeasurements whichthey made to validate Darcyrsquos law at extremely low fluidflow rates Using a column containing particles packed toa measured external column porosity of 0399 and havinga measured total column porosity of 0593 (Figure 2 in thepaper) Guiochon et al measured the column pressure dropand fluid flow rate at flow rates ranging between 0015 and05mLminwith ethylene glycol as the fluidThe authors statethat the particles in the column are spherical and they appearto have gone to extraordinary lengths to obtain an accuratemeasure of the particle diameter and of course as a result ofmodern techniques of particle size classification the particlesize distribution is very narrow
In Figure 1 of the paper the authors show a plot ofcolumn measured pressure drop in bar versus ldquofluid linearvelocityrdquo in cmsec This plot represents their measurementstaken in the empty column The only velocity that exists in
an empty column is the superficial velocity and thus theabscissa values in the plot in Figure 1 designated as ldquofluidlinear velocityrdquo represent superficial linear velocity
In Figure 2 in their paper however the authors showanother plot of measured pressure drop in bar also versusldquofluid linear velocityrdquo in cmsec The velocity on the abscissain this plot although also designated as ldquofluid linear velocityrdquodoes not equate (for permeability related purposes) to theldquofluid linear velocityrdquo in Figure 1 This is because in Figure 2the columnwas filled with porous particles and themeasuredvelocity was the mobile phase velocity Indeed the onlyvelocity used throughout the entire paper is the mobile phasevelocity (mobile phase velocity and superficial velocity areone and the same in an empty column)
In Table 1 of their paper the authors report that for thecolumn represented in Figure 2 the slope of a line achievedwhen the mobile phase velocityin units of cmsec is plottedversus pressure drop in units of bar is 1829 They also reportthat the flow resistance parameter 120601 for this column is 865In Table 3 herein it can be seen that the raw data for thiscolumn (Column C) generates an apparent value of 257 for1198750 This value is very close to Giddingsrsquo value of 267 Indeed
by referring to Figures 2 and 3 herein one can see that thereported raw data in this study without any modification
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
Journal of Materials 7
Table2Cr
oss-referencefor
term
s119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119896119871
119863Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
120576119905Φ
(120576119905minus1205760)
120576119894
Δ119875119889119901
2Δ119875119889119901
2(1minus120576119905Φ)2120576119905(1minus1205760)2
1120601
1198750120576119905
119889119901
2120576119905
(119889119901119898Ω119901)
(1minus1205760)
120583119905120578119871
120583119904120578119871
(120576119905Φ)3
(1205760)3
120601Ψ
120601po
iseUnits
Non
eNon
eNon
eNon
eNon
epsi
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
cmNon
ecm
cmgcmminus1 secminus1cm
3minminus1cm
secminus
1cm
secminus
1cm
secminus
1
Nonporousp
articles
Gidding
srsquotradition
alno
nporou
scolum
n040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
1000010
00010
0010
0399
004
00100
0100
Guiocho
nrsquostradition
alspheric
alparticle(app
arent)
0362
1000
0362
000
0000
0187
830
2293
3106
858
120119864minus03
267
97436119864minus10
15046
1000010
00010
0010
0361
0036
0100
0100
Gidd
ingrsquos
table5
3-1
5060meshglassb
eads
0422
1000
0422
000
0000
0113
500
1184
1873
444
200119864minus03
267
113
844119864minus10
15046
1000010
00010
0010
0421
004
20100
0100
5060meshglassb
eads
040
910
00040
9000
0000
0126
560
1371
2097
513
179119864minus03
267
109
729119864minus10
15046
1000010
00010
0010
040
70041
0100
0100
Guiochonrsquosteaching
ExA
irregular
particle
(app
arent)
040
010
00040
0000
0000
0225
1000
2500
2250
563
100119864minus03
444
178
400119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
ExA
correctedforp
article
irregularity
040
010
00040
0000
0000
0225
601
1502
2250
563
166119864minus03
267
107
400119864minus10
15046
078
00010
000
080010
0399
004
00100
0100
ExB
spheric
alparticle
(app
arent)
040
010
00040
0000
0000
0187
830
2075
2250
563
120119864minus03
369
148
482119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
ExB
corrected
040
010
00040
0000
0000
0135
601
1502
2250
563
166119864minus03
267
107
666119864minus10
15046
100
00010
00010
0010
0399
004
00100
0100
Porous
particles
Gidding
srsquotradition
alpo
rous
column
060
0066
7040
00200
0333
203
900
1500
3375
563
111119864minus03
267
160
667119864minus10
15046
100
00010
00010
0010
0599
006
00150
0100
Gidd
ingrsquos
table5
3-1
3040meshalum
ina
0803
0498
040
00403
0672
271
1204
1499
4510
562
830119864minus04
267
214
667119864minus10
15046
100
00010
00010
0010
0801
0080
0201
0100
5060meshalum
ina
0837
0499
0417
0420
0721
235
1043
1246
3906
467
959119864minus04
267
224
803119864minus10
15046
100
00010
00010
0010
0835
0084
0201
0100
6080meshchromosorbW
(5DNP)
0766
0498
0382
0384
0621
316
1404
1833
5259
687
712119864minus04
267
204
545119864minus10
15046
100
00010
00010
0010
0764
0077
0201
0100
6080meshchromosorbW
(20
DNP)
0785
0495
0389
0396
064
8300
1333
1698
4991
636
750119864minus04
267
210
589119864minus10
15046
100
00010
00010
0010
0783
0079
0202
0100
8 Journal of Materials
would be eliminated and by using only smooth sphericalparticles such as glass beads in which case the variableof particle shaperoughness would be eliminated If both ofthese things were done one would be able to directly andunambiguously determine the true value of 119875
05
Accordingly since his textbook contains a general teach-ing without any expressed restrictions with respect to particleporosity let us first assume that the column in Guiochonrsquosworked example is packed with nonporous particles Sec-ondly let us assume that the value for the external porosityof the column (which is the same as the total porosity ina column packed with nonporous particles) is the valuetraditionally assumed for a typical well-packed column thatis 04
As illustrated by example A in Table 2 this would leadto an apparent value for 119875
0of 444 which is obviously much
too high6 Using the adjusted particle size of 8 micrometers(rounded) which is calculated based on a value of 078 forΩ
119901
we derive values of 267 for 1198750and 107 for 119875
119905(the relationship
between these two values is of course equal to the totalporosity of the column 120576
119905 as is dictated by (20))
Similarly as illustrated by example B in Table 2 when thecolumn of Guiochonrsquos worked example is adapted to accom-modate spherical nonporous particles Guiochonrsquos teachingof the value for 119896
0of 12 times 10minus3 results in an apparent value
of 369 for 1198750
7 again too high [10 11] On the other hand ifwe correct the value of 119875
0 we get a column external porosity
of 03620which is too low for a typical well-packed column(120576
0= 04) [8] Moreover in the case of both corrections
Guiochonrsquos teaching overstates the pressure gradient (187 psicompared to the 135 psi which Giddingsrsquo experiments wouldgenerate) Accordingly it is only when we correct for thepressure drop that we achieve appropriate values of 04 for1205760 267 for 119875
0 and 107 for 119875
119905
As is evident from Table 2 when the porosity of columnspacked with nonporous particles varies the value of 119875
119905also
varies because it is based upon the measurement of mobilephase velocity which in turn changeswith the porosity of thecolumn Indeed for typical columns packed with nonporousparticles having external porosities close to 04 (038ndash042)[8] Table 2 reveals that the value of 119875
119905varies by as much as 6
points while the value of 1198750remains constant at 267
Having established a range of values for 119875119905based upon
columns packed with nonporous particles one can nowproceed to evaluate the range of values for 119875
119905based upon
columns packed with porous particles which is to say thatwe can now evaluate the impact of particle porosity onpermeability Among other things Table 2 shows that whenit comes to porous particles Guiochonrsquos 119875
119905is even more
variable (and thus even less useful) because the value ofthe coefficient can differ on a column-to-column basis byas much as 90 points depending on the combination of theinternal andexternal column porosities at play
Figures 1(a) and 1(b) are a plot of the data found in Table 1In this plot Guiochonrsquos modified permeability coefficient 119875
119905
is plotted against column total porosity 120576119905 Once again it will
be appreciated that (a) all data falls on a line with a slope of267 which represents the value of 119875
0and (b) the line has no
intercept which means that when 120576119905is zero 119875
119905is also zero As
shown in the plot values of the parameter Φ are highest atlow values of 119875
119905and lowest at high values of 119875
119905 Further as is
evidenced by the respective definitions for the terms119875119905and119875
0
contained herein the only time that the value of Guiochonrsquos119875119905does not differ from the value of 119875
0(267) is when the value
of 120576119905is 1 its value in a theoretical packed column devoid of
particles that is acapillaryFinally Figure 1(c) is another plot of the data in Table 1
but this time the values for 119875119905are plotted against particle
porosity 120576119901 The equation of this line does have an intercept
which represents the value of 119875119905when the particles are
nonporous It is also obvious from the equation of this linethat it is only when particle porosity is unity that is in acapillary that the value of 119875
119905is 267
It will be appreciated that since 120576119905is the sumof the internal
and external porosities of a packed column any given valuefor 119875
119905involving a column packed with porous particles can
represent many different combinations of these parametersVice versa any given value for 120576
119905can generate many different
values for 119875119905depending on the amount of particle porosity
which is present On the other hand as is pointed outearlier column internal porosity andor particle porosityplay no role in determining the permeability of a packedcolumn the only porosity parameter which is relevant forsuch purposes is column external porosity In other words asa permeability coefficient 119875
119905is not a particularly informative
concept because it still fails to provide any meaningful tie tothe underlying physical phenomena that govern the pressuredropfluid flow relationship
In the case of permeability measurements one quan-tifies only total pressure drop Pressure drop however isa commingling of the contributions of the several differ-ent parameters embodied in the pressureflow relationshipTherefore one must accurately represent the contributionof each factor before adding them together to equal thepressure drop Obviously if one mistakes the contribution ofany given element for instance particle size this will skewthe contribution of some other element when summed forinstance column porosity
Our frame of reference has been calibratedwith respect tothe interrelationships between the values of 119875
0 119875
119905 120576
119905 and 120576
119901
based on the data reported in Giddingsrsquo table 53-1 in his 1965textbook Since his data base was derived from experimentsusing nonporous particles the permeability results of whichwere extended to columns packed with fully porous particlesby using his Φ parameter the validity of which in turn wasestablished by independent means there is no uncertaintywith respect to the critical variable of external porosity 120576
0
inherent in our frame of reference It is for this reasontherefore that our frame of reference trumps any reporteddata which uses any other technique including the techniqueof inverse size exclusion chromatography by itself to establishvalues for 120576
0[12] In addition our frame of reference is
internally consistent due to the grounding of the parametersat the boundary condition of a theoretical capillary at whichpoint the values of 120576
119905and 120576
119901are unity and the values of 119875
0and
119875119905are equal In essence these interrelationships exemplify the
conservation laws of nature which are the ultimate standard
Journal of Materials 9
of measure when considering partial fractions of a singleentity
Consequently as Figures 1(b) and 1(c) illustrate if oneknows based upon other independent means the value of acolumnrsquos total porosity or its particlesrsquo porosity respectivelyone can at least verify whether results of permeability exper-iments expressed in terms of 119875
119905are reasonable Or to put
it another way if the 119875119905results reported do not conform to
what these plots tell us one should expect from columns andparticles based upon independently derived porosity valuesfor the particles under study we know that something iswrong Accordingly this is the way in which we use 119875
119905below
to analyze and evaluate Guiochonrsquos experimental work of thelast decade
8 The Origin of Guiochonrsquos Value forColumn Permeability
In 1966 Guiochon wrote a review paper entitled ldquoProblemsRaised by the Operation of Gas Chromatographic Columnsrdquo[13] In that paper he reviewed the development of the per-meability equation from Darcy to the then present Amongother things he reported that ldquoa mean 120576 of 038 is foundfor a number of packings obtained from powdered prod-ucts of various origins consisting of spherical or irregularparticles Almost all the measured values are between035 and 045rdquo (p 14) He then went on to discuss what hecalled the ldquosemiempirical Blake-Kozeny equationrdquo which hereported as 119896 = [119889
2
1199011205762][150(1 minus 120576)
2] [13 p 15 Equation
(11)] What sticks out of course is the fact that Guiochonstates here that 119875
0equals 150 the value for the permeability
coefficient championed by Ergun not the value posited byCarman To be fair we should point out that Guiochongave warning that he was uncomfortable with any particularvalue for the permeability coefficient For even though herecited the Kozeny-Blake equation with a value of 150 for itspermeability coefficient he stated that ldquothe values of 119896 givenby (this) equation are not very accurate and the deviationscan be as large as 50rdquo [13 p 15] Guiochon attributedmost ofthis uncertainty to the inability of practitioners of that era toeffectively measure the value of a columnrsquos external porositySince as he pointed out ldquo(the Kozeny-Blake) equation isvery sensitive to variations in 120576rdquo he concluded that theequation is ldquoof little userdquo and that it was better practice torely upon grosser measurements of permeability [13 p 15]Thus Guiochon at this timeframe made a conscious decisionto staywithin the relatively safe but admittedlymore obscureboundaries of Darcyrsquos teaching on column permeability
In 2006 however Guiochon emerged from under theshadow of Darcy permeability when he published a reviewpaper which purported to be a comprehensive summaryof the current state of the art in chromatography [14]In light of the fact that Guiochon and his coauthors hadvirtually contemporaneously published a new version of theirtextbook one would have expected that Guiochon wouldtake this opportunity to reaffirm his textbook teaching So itshould come as no surprise that he would open his discussionof the pressure dropfluid flow relationshipwith the following
assertion ldquothe permeability of packed beds used in liquidchromatography is in the order of 1 times 10minus3rdquo [14 p 9] Asone would expect the authority which Guiochon cites forthis seemingly familiar proposition is the 2006 edition of hisown textbook For the first time however he also providessome citations to third party sources In particular he cites a1997 chromatography handbook written by Uwe Neue ChiefScientist for Waters Corporation [15 p 9] As one can seeNeuersquos equationmdashlike Guiochonrsquosmdashdoes indeed postulate apermeability coefficient with a value of 1 times 10minus3
At this point however Guiochon abandons his tradi-tional caution about the value of the permeability coeffi-cient and marries his permeability equation to the Kozeny-Blake equation thereby endorsing not only the porositydependence term in that equation but also a specific valuefor 119875
0 Referring to his value of 1 times 10minus3 he states the
following ldquothe specific column permeability is related to thecolumn porosity through the Kozeny-Carman equation 119896
119891=
12057630180 (1 minus 120576
0)2rdquo [14 p 9] Not surprisingly Neue makes
the same connection ldquothe specific permeability depends onthe particle size 119889
119901and the interstitial porosity 120576
0of the
packed bedThe relationship is known as theKozeny-Carmanequation 119861
0= (1185)(1205763
01198892119901(1 minus 120576
0)2)rdquo [15 p 30]
Neuersquos permeability equation and Guiochonrsquos textbookpermeability equation however differ in that Neue uses thesuperficial not the mobile phase velocity Accordingly sinceall other features of their equations are the same one wouldexpect them to get different values for their permeabilitycoefficients Neue himself recognizes this when he says inhis handbook ldquothis value [1000] is also in good agreementwith our experience but values as low as 500 have beenreported in the literature8 However there are several reasonswhy different values reported by different workers do notrepresent true differences in the resistance parameter Firstlower values are obtained for porous packings if the specificpermeability is measured on the basis of the breakthroughtime of an unretained peak instead of using the flow raterdquo see[15 p 32] (when he distinguishes the ldquobreakthrough time ofan unretained peakrdquo from the ldquoflow raterdquo Neue is of coursereferring to the mobile phase velocity and the superficialvelocity resp) Consequently if one takes Neuersquos point andturns it on its head one can see that if he and Guiochonwere to get the same value for their permeability coefficients[1000] but one (Guiochon) is using mobile phase velocityto derive it and the other (Neue) uses superficial velocity toderive it this would indeed represent a ldquotrue differencerdquo intheir equations
So how can Guiochon cite both Neue and himself for theproposition in his review paper that ldquothe permeability of thepacked beds used in liquid chromatography is in the order of1times 10minus3rdquoThe answer is that at least for purposes of his reviewpaper he adopts Kozeny-Blakersquos fluid velocity convention inplace of the velocity frame he uses in his textbook Here iswhat he says ldquothis approximation [1 times 10minus3] is valid only ifthe superficial velocity is usedrdquo [14 p 9] (emphasis supplied)
As a matter of pure mathematics Guiochonrsquos permeabil-ity coefficient of 1 times 10minus3 can only be equal to 1205763
0180(1 minus 120576
0)2
if the value of 1205760is 04 the typical value for the interstitial
10 Journal of Materials
fraction of a well-packed column Guiochon confirms in hisreview paper that this is indeed what he assumes in his state-ment of the Kozeny-Blake equation [14 p 9] Accordinglyhis adoption of the value of 180 for 119875
0solely depends on the
validity of his assertion that the value of the permeabilitycoefficient is 1 times 10minus3
We already know from our discussion of Guiochonrsquostextbook teaching that when this value is associated with apermeability equation based on mobile phase velocity it iswrong As reflected in our Table 2 when Guiochonrsquos workedexample involving this figure is corrected to account for theembedded factor due to the irregularity of the particles whichhis worked example assumes the value of the permeabilitycoefficient generated by his equation is not 1 times 10minus3 it is 166times 10minus3
So where did this value of 1 times 10minus3 come from AlthoughGuiochon in his review paper cites Neuersquos handbook for thisvalue Neue himself offers absolutely no authority for it Onthe other hand Guiochon does provide one other citation inhis review paper for this value The citation is to an article byHalasz et al published in 1975 [16]
If one then goes to that article one sees that Halaszet al do indeed proffer the same equation as Neue (albeitin a somewhat different format) and yes it does containthe value of 1 times 10minus3 See [16 Equation (1)] Neverthelessthe authors fail to identify from whence they obtained thisvalue On the other hand they do make reference to apaper that Endele et al wrote a year earlier in 1974 withKlaus Unger entitled ldquoInfluence of the Particle Size (5ndash35 120583)of Spherical Silica on Column Efficiencies in High-PressureLiquid Chromatographyrdquo [17] It appears that this is wherethe body is buried for it is in this article that Halasz reportshaving actually derived a value of 1 times 10minus3 for the permeabilitycoefficient
In summarizing the results of their experiments whichincidentally are based on measurements of superficial notmobile phase velocity (see their Equation (7)) Halasz et alcalculate an average value for the permeability coefficient(which they notate as119870
119865) of 1198892
1199011080 Accordingly they state
the following ldquo[I]t is proposed to define an average particlesize 119889
119901 for columns packed with spherical silica using the
balanced density packing method by the equation 1198892119901
=
1198701198651000 = 1198651205781198711031199032120587Δ119875rdquo [17 p 382 their Equation (7)]To recapitulate it thus appears that (1) Halasz was the
source of the value of 1 times 10minus3 for the permeability coefficientwhich Guiochon adopts in his review paper (2)Halaszrsquo valuewas based upon measurements which involved sphericalparticles and (3)Halaszrsquo value was properly derived from anequation which utilized the superficial velocity as the fluidvelocity
However we still have a problem if we are going to acceptthe value of 1 times 10minus3 as a basis for calculating the constantin Kozeny-Blake we still need to know the value of Halaszrsquocolumnsrsquo external porosity For example in the absenceof measured values for the external porosity Guiochonrsquosassertion in his review paper that the value of the constantis 180 is without foundation
Unfortunately however Halasz bases his methodologyon ldquoassumedrdquo versus ldquomeasuredrdquo values for external porosityNevertheless he does provide some clues that may helpus peel back this onion First he says that his value forthe permeability coefficient applies to ldquocolumns using thebalanced density packing methodrdquo Secondly after statinghis proposed value for the coefficient we find the followingtelltale sentence ldquothe permeability of columns packed by theconventional ldquodryrdquo method are worse by a factor of about 2than those packed by the balanced density methodrdquo [17 p382] For this proposition Halasz cites yet another of his ownarticles this time a paper written by himself and M Naefein 1972 entitled ldquoInfluence of Column Parameters on PeakBroadening in High-Pressure Liquid Chromatographyrdquo [18]
When one follows this thread by going to Halaszrsquos 1972paper one finds a possible solution to the puzzle For this iswhere Halasz describes his calculations of the permeabilitycoefficient from experiments with a number of conventionaldry-packed columns containing spherical silica particlesNot surprisingly the resulting value for the permeabilitycoefficient for such columns was not 1 times 10minus3 Instead Halaszsays that his experiments with such columns resulted inthe following value for the permeability coefficient ldquo119870 sim
11988921199012000rdquo [18 p 78] In other words consistent with what
Halasz says in his 1974 article the value for the coefficient forthese dry-packed columns was indeed ldquoworse by a factor ofabout 2rdquo compared to its value for columns prepared with thebalanced density method
The obvious implication of this is that when Halasz gota value of 1 times 10minus3 for the permeability coefficient in 1974he must have been evaluating columns that were packed toan interstitial fraction considerably higher than the externalporosity of the columns that were the subject of his 1972experiments Our problem however is still not resolvedbecause Halasz did not accurately determine the externalporosity of the columns under study in either the 1972 or 1974columns
We surmise that in their 1972 paper Halasz and hiscoauthors made use of the teaching of Giddings outlinedin his 1965 textbook which at the time of their 1972 paperwas seven years old Here is what they say ldquoexperiencehas shown that in a regular packed column with the sievefractions usual for chromatography 120576
0is independent of the
particle size if the particles are more or less spherical In anacceptable approximation 120576
0is equal to 04 In those columns
packed with nonporous supports 119906V and 119906 are identicalUsing porous supports (ie silica gel alumina chromosorbPorasil and so forth) 119906 = 4119865(1205871198892
119888120576119879) where 120576
119879is the
total porosity and indicates the pore volume of the supportrdquoIt will be appreciated that 119906V stands for interstitial velocity(our 120583
119894) and 119906 stands for mobile phase velocity (our 120583
119905) The
ldquoexperiencerdquo which Halasz and his coauthors are referringto concerning ldquocolumns packed with nonporous supportsalumina and chromosorbrdquo is that recorded by Giddings inTable 53-1 in his 1965 text Continuing to establish theinterrelationships inherent in the various entities involvedin chromatographic columns with respect to both interstitialandmobile phase velocity the authors conclude ldquoformally 119906V
Journal of Materials 11
can be calculated for a porous support assuming 1205760is equal
to 04rdquo (emphasis supplied) The authors then announce themethodology underlying their frame of reference with regardto column permeability as follows ldquowe determined in all ofour columns the 119906119906V ratio to be 056 with a differentialrefractometer Consequently the total porosity 120576
119905was equal
to 0714rdquo It will be appreciated that their ratio of 119906119906V is oneand the same as GiddingsrsquoΦ parameter as utilized by him inhis 1965 textbook and as we define hereinabove
We can now identify the origin of the discrepancybetween Halasz and Giddings and by extension betweenHalasz and Guiochon (and Neue) For even though Halaszmakes use of the results in Giddingsrsquo Table 53-1 he failsto appreciate the teaching inherent therein regarding theimportance of an accurate value for column external porositywhen using porous particles In establishing his values forΦGiddings was careful to ground his values for the externalporosity of columns packed with porous particles in acomparison of their measured pressure drops with measuredpressure drops in columns packed with nonporous particleswhere the values for (external) porositywere accurate becausethey are equal to the columnsrsquo total porosity
In their 1972 paper Halasz et al used a refractometer todetect the inert solute 119899-pentane in the mobile phase of 119899-heptaneThey injected the 119899-pentane into the flowing mobilephase the flow rate of which they presumably measured witha volumetric cylinder and stop watch Since the exit timeof the 119899-pentane peak was identified by the refractometertheir measurement of holdup volume was accurate andaccordingly so was their value of 0714 for 120576
119905(there is no
uncertainty related to column total porosity when using inertsolutes)This in turn led to an accurate value for their119906 termmobile phase velocity However since they did not measurethe interstitial velocity 119906V but rather used an estimate basedupon the measured flow rate and an ldquoassumedrdquo value of 04for external porosity their calculatedestimated value for 119906Vwas arbitrary This in turn means that the value of 056 fortheirΦ ratio was likewise arbitrary
In Table 3 herein we show the results for column number1 in the 1972 study which we designate as Column A
1 Note
that the authorsrsquo raw values correspond to a value of 500 for1198750 a valuewhich is obviously far too high As is plainly evident
in our Figures 2 and 3 this column data set does not conformin any reasonable fashion to the norms in our referencecharts Accordingly the data is badly flawed In our correcteddata for Column A
1in Table 3 herein we adjust the column
external porosity to the lowest value permitted in our frame ofreference (038) This correction results in a revised value forGiddingsrsquoΦparameter of 0532 In additionwe are compelledto adjust the particle downward (to 11 micrometers approx)as is also dictated by our frame of reference We justify ourlow corrected value for external porosity and our downwardadjustment for particle size on the aggressive ldquodry-packingrdquoprocess described by the authors ldquo[T] the best results wereachieved with the following method the column (id =2mm) 25 cm long was packed with 25 equal portions of theldquobrushrdquo After adding each portion the column was vibratedand afterward tapped on the floor and also vigorously tappedwith a rod from the siderdquo Based upon a column packing
experience of the principal author herein spanningmore than30 years we believe that these packing conditions generateda packed column with a low external porosity and also one inwhich the particles experienced significant breakage
The main differences between the columns in the 1972and 1974 papers concerned the particle sizes and themethodsused to pack the columns The particle diameters in the 1972study fell in a range of about 15ndash200 micrometers while thosein the 1974 study were much smaller being in the range of 5ndash40 micrometers approximately Moreover while the columnsin the 1972 study were dry-packed the columns in the 1974study were slurry-packed (what Halasz calls ldquothe balanceddensity packing methodrdquo) The authors of the 1974 paperrecite that their permeability results shown in their Table 4correspond to values for 120601
0of 1080 and values for 120601 of 910
In Table 3 herein we show the results for the 1974 study as anaverage value in our column designated as A
2 Note that the
raw data reported by the authors which again has the built-inassumption that the column external porosity was 04 placesthe value of 119875
0at 192 In our corrected values for Column A
2
we show that a value of 267 for 1198750places the column external
porosity at a value of 043 The ratio of the value for 120601 of 910when compared to the value for 120601 in the 1972 study of 2007is about 22 (2007910 = 22) which is in good agreementwith the authorsrsquo statement that the permeability of the 1972columns was worse by a factor of about 2 Again based onthe experience of the principal author herein in the packingof hundreds of thousands of commercially sold slurry packedcolumns we conclude that the balanced density techniquedescribed by the authors in the 1974 paper is fully capable ofproducing columns with this high and external porosity
In summary we conclude that for purposes of theKozeny-Blake equation (a) the external porosity of theHalasz 1972 columns was 038 (b) the external porosity of thecolumns which he tested in 1974 was 043 and (c) Halaszrsquos1972 and 1974 studies when properly analyzed corroborateGiddingsrsquo value for 119875
0of 267 not Carmanrsquos value of 180
Although the fault for all of the misdirection about the valueof the permeability coefficient being 1 times 10minus3 therefore restsprimarily on the shoulders of Halasz one has to question whyGuiochon (andNeue) so readily adopted itWhatevermay bethe reason Guiochonrsquos support for this proposition has beenand continues to be a cause for much of the conventionalacceptance of Carmanrsquos (erroneous) figure of 180 as thecharacteristic value of the coefficient in the Kozeny-Blakeequation
9 Guiochonrsquos Experimental Data
In addition to the data from Halasz 1972 and 1974 studiesTable 3 herein contains data from a representative sampleof Guiochonrsquos personal experimental permeability investi-gations over approximately the last decade as reported inthe numerous trade journals particularly the Journal ofChromatography A In each case the table has a single rowdedicated to Guiochonrsquos reported raw data and a single rowdedicated to corrected values for each column based on thereference chart in Table 1 herein
12 Journal of MaterialsTa
ble3Summaryexperim
entald
ata
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Naefe
andHalasz
columns
1198601(19
72)
Colum
nA1rawdata
07140
0560
040
000314
0523
782007
2811
4016
563
498119864minus04
500
357
908119864minus10
648119864minus10
2500
020
100
000135
0001350
00038
100
053
132
074
Colum
nA1corrected
07140
0532
03800
0334
0539
7813
3518
705002
701
749119864minus04
267
191
908119864minus10
648119864minus10
2500
020
100
000110
000110
100038
100
053
139
074
Endelean
dHalasz
columns
1198602(19
74)
Colum
nA2rawdata
08426
0475
040
00044
30738
11910
1080
4740
563
110119864minus03
192
162
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
033
016
Colum
nA2corrected
08426
0511
04309
0412
0723
11910
1080
3411
405
110119864minus03
267
225
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
031
016
Guiochon
columns
1198612ndash1198614(19
98)
Colum
nB 2
rawdata
05639
0722
040
730157
0264
22771
1367
2931
520
130119864minus03
263
148
467119864minus10
263119864minus10
429
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 3
rawdata
05582
0704
03932
0165
0272
44764
1369
3382
606
131119864minus03
226
126
471119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 4
rawdata
05509
0710
03909
0160
0263
72784
1422
3422
621
128119864minus03
229
126
459119864minus10
253119864minus10
1328
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 3
corrected
05653
0723
040
860157
0265
44774
1369
2899
513
129119864minus03
267
151
465119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 4
corrected
05651
0723
040
830157
0265
72776
1374
2907
514
129119864minus03
267
151
464119864minus10
262119864minus10
1375
500
100
000
060
000
0600
00165
98008
020
015
Guiochon
columnC
(1999)
Colum
nCrawdata
05930
0673
03990
0194
0323
183
865
1459
3372
569
116119864minus03
257
152
109119864minus09
645119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Colum
nCcorrected
05930
0673
03990
0194
0323
190
900
1518
3372
569
111119864minus03
267
158
105119864minus09
620119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Guiochon
columns
1198631ndash1198636(2006)
Colum
nD
1rawdata
07830
0456
03570
0426
0663
86694
886
7115
909
144119864minus03
975
76334119864minus10
261119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
013
Colum
nD
2rawdata
07230
0491
03550
0368
0571
88656
907
6726
930
152119864minus03
975
70353119864minus10
255119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
014
Colum
nD
3rawdata
06850
0506
03466
0338
0518
97685
1000
7025
1026
146119864minus03
975
67338119864minus10
231119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD
4rawdata
06560
0521
03415
0315
0478
103
696
1062
7143
1089
144119864minus03
975
64332119864minus10
218119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD5rawdata
06190
0545
03371
0282
0425
109
692
1118
7100
1147
144119864minus03
975
60334119864minus10
207119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
016
Colum
nD
6rawdata
05760
0587
03383
0238
0359
107
635
1103
6516
1131
157119864minus03
975
56364119864minus10
210119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
017
Colum
nD
1corrected
07830
0542
04243
0359
0623
86907
1158
3397
434
110119864minus03
267
209
334119864minus10
261119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
013
Colum
nD
2corrected
07230
0584
04221
0301
0521
88857
1186
3212
444
117119864minus03
267
193
353119864minus10
255119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
014
Colum
nD
3corrected
06850
0603
04129
0272
0463
97896
1307
3354
490
112119864minus03
267
183
338119864minus10
231119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
015
Colum
nD
4corrected
06560
0621
040
730249
0420
103
911
1388
3411
520
110119864minus03
267
175
332119864minus10
218119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
015
Colum
nD
5corrected
06190
0650
04025
0217
0362
109
905
1462
3390
548
110119864minus03
267
165
334119864minus10
207119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
016
Colum
nD
6corrected
05760
0701
04038
0172
0289
107
831
1442
3111
540
120119864minus03
267
154
364119864minus10
210119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
017
Guiochon
columns
1198641-1198642
(2007)
Colum
nE 1
rawdata
064
500594
03830
0262
0425
356
1119
1735
4371
678
894119864minus04
256
165
804119864minus11
519119864minus11
1500
046
100
000
030
000
0300
000
41300
030
079
047
Colum
nE 2
rawdata
05400
0800
04320
0108
0190
470
1000
1853
2161
400
100119864minus03
463
250
729119864minus11
393119864minus11
1500
046
100
000
027
000
0270
000
41300
030
070
056
Colum
nE 1
corrected
064
500594
040
000245
040
8356
969
1502
3628
563
103119864minus03
267
172
804119864minus11
519119864minus11
1500
046
100
000
028
000
0279
000
41300
030
075
047
Colum
nE 2
corrected
05400
0800
04320
0108
0190
470
577
1068
2161
400
173119864minus03
267
144
729119864minus11
393119864minus11
1500
046
088
000
023
000
0205
000
41300
030
070
056
Guiochon
columns
1198651ndash1198654
(2008)
Colum
nF 1
rawdata
06350
0586
03720
0263
0419
197
725
1142
4865
766
138119864minus03
149
95399119864minus11
253119864minus11
300
021
100
000
017
000
0170
00035
100
048
129
076
Colum
nF 2
rawdata
064
200581
03730
0269
0429
361
807
1258
4863
758
124119864minus03
166
107
358119864minus11
230119864minus11
500
021
100
000
017
000
0170
00035
100
048
129
075
Colum
nF 3
rawdata
064
100588
03770
0264
0424
670
748
1166
4643
724
134119864minus03
161
103
387119864minus11
248119864minus11
1000
021
100
000
017
000
0170
00035
100
048
128
075
Colum
nF 4
rawdata
06390
0595
03800
0259
0418
1014
752
1177
4476
701
133119864minus03
168
107
384119864minus11
246119864minus11
1500
021
100
000
017
000
0170
00035
100
048
127
075
Colum
nF 1
corrected
06350
0630
040
000235
0392
197
954
1502
3572
563
105119864minus03
267
170
399119864minus11
253119864minus11
300
021
100
000
019
000
0195
00035
100
048
120
076
Colum
nF 2
corrected
064
200623
040
000242
0403
361
964
1502
3611
563
104119864minus03
267
171
358119864minus11
230119864minus11
500
021
100
000
019
000
0186
00035
100
048
120
075
Colum
nF 3
corrected
064
100624
040
000241
0402
670
963
1502
360
6563
104119864minus03
267
171
386119864minus11
248119864minus11
1000
021
100
000
019
000
0193
00035
100
048
120
075
Colum
nF 4
corrected
06390
0626
040
000239
0398
1014
960
1502
3594
563
104119864minus03
267
171
384119864minus11
246119864minus11
1500
021
100
000
019
000
0192
00035
100
048
120
075
Journal of Materials 13
Table3Con
tinued
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Guiochon
columns
1198671-1198672serie
s(2010)
Colum
nH
1rawdata
05320
0744
03960
0136
0225
120
563
1057
3125
587
178119864minus03
180
96122119864minus10
652119864minus11
1500
046
100
000
026
000
0262
00104
050
005
013
009
Colum
nH
2rawdata
05420
0686
03720
0170
0271
61747
1379
4152
766
134119864minus03
180
98823119864minus11
446119864minus11
1000
046
100
000
025
000
0248
00054
050
005
013
009
Colum
nH
1corrected
05320
0752
040
000132
0220
120
799
1502
2993
563
125119864minus03
267
142
122119864minus10
652119864minus11
1500
046
100
000
031
000
0313
00104
050
005
013
009
Colum
nH
2corrected
05420
0738
040
000142
0237
61814
1502
3049
563
123119864minus03
267
145
823119864minus11
446119864minus11
1000
046
100
000
026
000
0259
00054
050
005
013
009
Guiochon
columns
1198673-1198674serie
s(2010)
Colum
nH
3rawdata
06270
060
903820
0245
0396
463
700
1117
4296
685
143119864minus03
163
102
447119864minus11
280119864minus11
1500
021
100
000
018
000
0177
00036
050
024
063
038
Colum
nH
4rawdata
05170
0791
040
900108
0183
433
616
1191
2639
511
162119864minus03
233
121
580119864minus11
300119864minus11
1500
021
100
000
019
000
0189
00036
050
024
059
047
Colum
nH
3corrected
06270
0638
040
000227
0378
463
942
1502
3527
563
106119864minus03
267
167
447119864minus11
280119864minus11
1500
021
100
000
021
000
0205
00036
050
024
060
038
Colum
nH
4corrected
05170
0791
040
900108
0183
433
699
1353
2639
511
143119864minus03
265
137
580119864minus11
300119864minus11
1500
021
100
000
020
000
0201
00036
050
024
059
047
Guiochon
columns
1198681-1198686
(2010)
Colum
nI 1rawdata
06580
0562
03700
0288
0457
488
741
1127
5156
784
135119864minus03
144
95390119864minus11
256119864minus11
500
021
100
000
017
000
0170
00104
050
024
065
037
Colum
nI 2rawdata
06550
0576
03770
0278
044
6873
661
1009
4745
724
151119864minus03
139
91437119864minus11
286119864minus11
1000
021
100
000
017
000
0170
00104
050
024
064
037
Colum
nI 3rawdata
06550
0588
03850
0270
0439
1209
610
931
4341
663
164119864minus03
141
92474119864minus11
310119864minus11
1500
021
100
000
017
000
0170
00104
050
024
062
037
Colum
nI 4rawdata
06430
0572
03680
0275
0435
260
787
1225
5153
801
127119864minus03
153
98367119864minus11
236119864minus11
500
030
100
000
017
000
0170
00104
050
012
032
018
Colum
nI 5rawdata
06540
0583
03810
0273
044
144
3683
1044
4531
693
146119864minus03
151
99423119864minus11
277119864minus11
1000
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 6rawdata
06550
0585
03830
0272
044
164
6665
1016
4438
678
150119864minus03
150
98434119864minus11
285119864minus11
1500
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 1corrected
06580
060
8040
000258
0430
488
988
1502
3701
563
101119864minus03
267
176
390119864minus11
256119864minus11
500
021
100
000
020
000
0196
00104
050
024
060
037
Colum
nI 2corrected
06550
0611
040
000255
0425
873
984
1502
3684
563
102119864minus03
267
175
437119864minus11
286119864minus11
1000
021
100
000
021
000
0207
00104
050
024
060
037
Colum
nI 3corrected
06550
0611
040
000255
0425
1209
984
1502
3684
563
102119864minus03
267
175
474119864minus11
310119864minus11
1500
021
100
000
022
000
0216
00104
050
024
060
037
Colum
nI 4corrected
06430
0622
040
000243
040
5260
966
1502
3617
563
104119864minus03
267
172
367119864minus11
236119864minus11
500
030
100
000
019
000
0188
00104
050
012
029
018
Colum
nI 5corrected
06540
0612
040
000254
0423
443
982
1502
3679
563
102119864minus03
267
175
423119864minus11
277119864minus11
1000
030
100
000
020
000
0204
00104
050
012
029
018
Colum
nI 6corrected
06550
0611
040
000255
0425
646
984
1502
3684
563
102119864minus03
267
175
434119864minus11
285119864minus11
1500
030
100
000
021
000
0207
00104
050
012
029
018
Guiochon
columns
1198691-1198692
(2011)
Colum
nJ 1rawdata90
∘A(120578
=00054
P)04980
0803
040
000098
0163
65654
1313
2801
563
153119864minus03
234
116126119864minus10
627119864minus11
1500
046
100
000
029
000
0287
00054
050
005
013
010
Colum
nJ 2rawdata90
∘A(120578
=00104
P)04980
0803
040
000098
0163
124
651
1307
2801
563
154119864minus03
232
116127119864minus10
630119864minus11
1500
046
100
000
029
000
0287
00104
050
005
013
010
Colum
nJ 1rawdata160
∘A(120578
=00054
P)05630
0714
04020
0161
0269
71896
1592
3099
550
112119864minus03
289
163
102119864minus10
573119864minus11
1500
046
100
000
030
000
0302
00054
050
005
012
009
Colum
nJ 2rawdata160
∘A(120578
=00104
P)05630
0714
04020
0161
0269
129
847
1504
3099
550
118119864minus03
273
154
108119864minus10
606119864minus11
1500
046
100
000
030
000
0302
00104
050
005
012
009
Colum
nJ 1corrected
04980
0803
040
000098
0163
65748
1502
2801
563
134119864minus03
267
133
126119864minus10
627119864minus11
1500
046
100
000
031
000
0307
00054
050
005
013
010
Colum
nJ 2corrected
04980
0803
040
000098
0163
124
748
1502
2801
563
134119864minus03
267
133
127119864minus10
630119864minus11
1500
046
100
000
031
000
0308
00104
050
005
013
010
Colum
nJ 1corrected
05630
0714
04020
0161
0269
71827
1470
3099
550
121119864minus03
267
150
102119864minus10
573119864minus11
1500
046
100
000
029
000
0290
00054
050
005
012
009
Colum
nJ 2corrected
05630
0714
04020
0161
0269
129
827
1470
3099
550
121119864minus03
267
150
108119864minus10
606119864minus11
1500
046
100
000
030
000
0299
00104
050
005
012
009
14 Journal of Materials
050
100150200250300350400
045 050 055 060 065 070 075 080 085
LineAll corrected data
Linear (all corrected data)Squares raw data Triangles corrected data
Pt
120576t
D seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
J1 J2 H4
H1H2
D6
B2 B3 B4
J3 J4
E2
C
H3
D5 D4
E1
D2
A1
D1
A2
y = 267xLine slope = 267
Figure 2 Summary of experimental results
0
500
1000
1500
2000
2500
20 25 30 35 40 45 50 55 60 65 70 75 80
Linear (line)
Ψ
LineAll corrected dataD seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
Line slope (P0) = 267
J1 J2
J3 J4E2
E1
A1
A2
H4 H1
C
120601
H2H3I1 I2 I3 I4 I5
F1 F2 F3 F4
D1 D2 D3 D4 D5 D6
Squares raw data Triangles corrected data
Figure 3 Summary of experimental results
Figures 2 and 3 herein are graphical representations ofTable 31015840s raw and corrected data for Guiochonrsquos columnsplotted in the same reference frames as in Figures 1(a) and1(b) respectively
As shown in the plots of the thirty-one total columnsevaluated in this data base only six columns had reportedraw data values for 119875
0close to the normThese were columns
B2 C E
1 H
4 J
1and J
2 Interestingly of the six conforming
columns 3 of the columns were packed with fully porousparticles and 3 columns with partially porous particles Itcan be seen however that when properly interpreted andadjusted all of Guiochonrsquos experimental data falls fairlywithin the norms outlined in the reference chart in Table 1In order to explore the reasons why the raw data does notconform more closely to the norms outlined in Table 1 eachof Guiochonrsquos studies included in Table 3 is evaluated in turn
Finally Table 4 herein is a list of symbols and parameterdefinitions used throughout this paper as well as their unitsof measure
91 Guiochonrsquos 1998 Study In a 1998 publication by Kohand Guiochon entitled ldquoEffect of the Column Length on theCharacteristics of the Packed Bed and the Column Efficiencyin a Dynamic Axial Compression Columnrdquo [19] Guiochonand his coauthor report the results of a study they hadcarried out on column packing using a rather novel techniquewhich was a combination of slurry packing and mechanicalcompression
The study involved the packing of a single cylinder ofdiameter 5 cm to various column lengths ranging between 1and 25 cm approximately The particles were all silica basedand coated with a reverse stationary phase C
18 However two
different categories of particles were involved One categorycontained highly irregular particles which were about 130micrometers in average diameter In addition these particleshad a high specific pore volume (11mLg) and exhibitedsignificant particle breakage under the packing conditionsinvestigated in the study Conversely the other particlecategory involved very small particles (6 micrometers) whichwere spherical in shape had a low specific pore volume(03mLg) and exhibited a greater ability to withstandparticle breakage under the studyrsquos packing conditions
The permeability data for the irregular particles isneglected herein because according to the authors theparticles exhibited significant breakage during packing andwithout an accurate value for the characteristic dimensionof the particle the data cannot be used in the Kozeny-Blake model to determine the value of 119875
0 The spherical
particles on the other hand behaved successfully under thepacking conditions chosen and yielded values calculated bythe authors for 119875
0of 619 268 226 and 229 for the four
different lengths of columns studied The data for thesecolumns is reported in our Table 3 in the rows for ColumnsB1through B
4 The higher value of 619 (the shortest column)
was rejected by the authors because they felt that either theparticles had been somewhat broken during packing or thecolumn frits had been blocked with fines Accordingly thepermeability data for this column is likewise neglected herein
The remaining three columns however were not con-sistent with respect to the authorsrsquo calculated values for 119875
0
Since the particles were identical the range of values for 1198750
of 226 to 268 must be due to experimental error The rawdata for column B
2 which produced a 119875
0value of 268 is
all self-consistent For example it generates a value for 119875119905
of 148 which is a reasonable value based upon the knownlow particle porosity for these Nova Pak particles and acorresponding value for 120576
0of 04073 which is a reasonable
value for a well-packed column However the other twocolumns columns B
3and B
4 had the much lower value of
126 for 119875119905 Accordingly in Table 3 herein corrected values
for column external porosity for columns B3and B
4are
displayed These corrected values are only slightly differentthan the reported values are within the experimental errorof the measurement and therefore in combination with the
Journal of Materials 15
Table 4 Glossary of terms
Symbol Unit dimensions (cgs) Definition119860 cm2 Column cross-sectional area119863 cm Column diameter119889119901
cm Particle spherical diameter equivalentΔ119875 gcmminus1secminus2 Pressure drop along column119889119901119898
cm Average particle diameter (measured)1198960
None Guiochonrsquos residual reciprocal permeability coefficient119875119905
None Guiochonrsquos modified permeability coefficient in Kozeny-Blake equation119871 cm Column length1198750
None Permeability coefficient in Kozeny-Blake equation119876 cm3minminus1 (exception) Volumetric fluid flow rate1199050
sec Time for elution of totally permeating unretained solute119896 cm2 General permeability coefficient in Darcy equation119870
119865cm2 Halaszrsquos permeability coefficient in Darcy equation (1974 paper)
119870 cm2 Halaszrsquos permeability coefficient in Darcy equation (1972 paper)119870
119888None Guiochonrsquos permeability coefficient in the Kozeny-Blake equation (as modified by Carmen)
Greek1205760
None External column porosity120576119894
None Internal column porosity120576119905
None Total column porosity120576119901
None Particle porosityΦ None Giddingsrsquo ratio of external and total column porosity120601 None Flow resistance parameter based on mobile phase velocity1206010
None Flow resistance parameter based on superficial velocityΩ
119901None Particle sphericity factor
120578 gcmminus1secminus1 Fluid absolute viscosity120583119894
cmsecminus1 Average fluid interstitial linear velocity120583119904
cmsecminus1 Average fluid superficial linear velocity120583119905
cmsecminus1 Average mobile phase velocityΨ None Porosity dependence term in the Kozeny-Blake equation based on mobile phase velocityΨ
0None Porosity dependence term in the Kozeny-Blake equation based on superficial velocity
raw data for column B2 are supportive of the value of 267 for
1198750
92 Guiochonrsquos 1999 Study In a 1999 publication by Farkaset al entitled ldquoValidity of Darcyrsquos Law at Low Flow Rates inLiquid Chromatographyrdquo [20] Guiochon and his coauthorsreport the results of some very exactingmeasurements whichthey made to validate Darcyrsquos law at extremely low fluidflow rates Using a column containing particles packed toa measured external column porosity of 0399 and havinga measured total column porosity of 0593 (Figure 2 in thepaper) Guiochon et al measured the column pressure dropand fluid flow rate at flow rates ranging between 0015 and05mLminwith ethylene glycol as the fluidThe authors statethat the particles in the column are spherical and they appearto have gone to extraordinary lengths to obtain an accuratemeasure of the particle diameter and of course as a result ofmodern techniques of particle size classification the particlesize distribution is very narrow
In Figure 1 of the paper the authors show a plot ofcolumn measured pressure drop in bar versus ldquofluid linearvelocityrdquo in cmsec This plot represents their measurementstaken in the empty column The only velocity that exists in
an empty column is the superficial velocity and thus theabscissa values in the plot in Figure 1 designated as ldquofluidlinear velocityrdquo represent superficial linear velocity
In Figure 2 in their paper however the authors showanother plot of measured pressure drop in bar also versusldquofluid linear velocityrdquo in cmsec The velocity on the abscissain this plot although also designated as ldquofluid linear velocityrdquodoes not equate (for permeability related purposes) to theldquofluid linear velocityrdquo in Figure 1 This is because in Figure 2the columnwas filled with porous particles and themeasuredvelocity was the mobile phase velocity Indeed the onlyvelocity used throughout the entire paper is the mobile phasevelocity (mobile phase velocity and superficial velocity areone and the same in an empty column)
In Table 1 of their paper the authors report that for thecolumn represented in Figure 2 the slope of a line achievedwhen the mobile phase velocityin units of cmsec is plottedversus pressure drop in units of bar is 1829 They also reportthat the flow resistance parameter 120601 for this column is 865In Table 3 herein it can be seen that the raw data for thiscolumn (Column C) generates an apparent value of 257 for1198750 This value is very close to Giddingsrsquo value of 267 Indeed
by referring to Figures 2 and 3 herein one can see that thereported raw data in this study without any modification
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
8 Journal of Materials
would be eliminated and by using only smooth sphericalparticles such as glass beads in which case the variableof particle shaperoughness would be eliminated If both ofthese things were done one would be able to directly andunambiguously determine the true value of 119875
05
Accordingly since his textbook contains a general teach-ing without any expressed restrictions with respect to particleporosity let us first assume that the column in Guiochonrsquosworked example is packed with nonporous particles Sec-ondly let us assume that the value for the external porosityof the column (which is the same as the total porosity ina column packed with nonporous particles) is the valuetraditionally assumed for a typical well-packed column thatis 04
As illustrated by example A in Table 2 this would leadto an apparent value for 119875
0of 444 which is obviously much
too high6 Using the adjusted particle size of 8 micrometers(rounded) which is calculated based on a value of 078 forΩ
119901
we derive values of 267 for 1198750and 107 for 119875
119905(the relationship
between these two values is of course equal to the totalporosity of the column 120576
119905 as is dictated by (20))
Similarly as illustrated by example B in Table 2 when thecolumn of Guiochonrsquos worked example is adapted to accom-modate spherical nonporous particles Guiochonrsquos teachingof the value for 119896
0of 12 times 10minus3 results in an apparent value
of 369 for 1198750
7 again too high [10 11] On the other hand ifwe correct the value of 119875
0 we get a column external porosity
of 03620which is too low for a typical well-packed column(120576
0= 04) [8] Moreover in the case of both corrections
Guiochonrsquos teaching overstates the pressure gradient (187 psicompared to the 135 psi which Giddingsrsquo experiments wouldgenerate) Accordingly it is only when we correct for thepressure drop that we achieve appropriate values of 04 for1205760 267 for 119875
0 and 107 for 119875
119905
As is evident from Table 2 when the porosity of columnspacked with nonporous particles varies the value of 119875
119905also
varies because it is based upon the measurement of mobilephase velocity which in turn changeswith the porosity of thecolumn Indeed for typical columns packed with nonporousparticles having external porosities close to 04 (038ndash042)[8] Table 2 reveals that the value of 119875
119905varies by as much as 6
points while the value of 1198750remains constant at 267
Having established a range of values for 119875119905based upon
columns packed with nonporous particles one can nowproceed to evaluate the range of values for 119875
119905based upon
columns packed with porous particles which is to say thatwe can now evaluate the impact of particle porosity onpermeability Among other things Table 2 shows that whenit comes to porous particles Guiochonrsquos 119875
119905is even more
variable (and thus even less useful) because the value ofthe coefficient can differ on a column-to-column basis byas much as 90 points depending on the combination of theinternal andexternal column porosities at play
Figures 1(a) and 1(b) are a plot of the data found in Table 1In this plot Guiochonrsquos modified permeability coefficient 119875
119905
is plotted against column total porosity 120576119905 Once again it will
be appreciated that (a) all data falls on a line with a slope of267 which represents the value of 119875
0and (b) the line has no
intercept which means that when 120576119905is zero 119875
119905is also zero As
shown in the plot values of the parameter Φ are highest atlow values of 119875
119905and lowest at high values of 119875
119905 Further as is
evidenced by the respective definitions for the terms119875119905and119875
0
contained herein the only time that the value of Guiochonrsquos119875119905does not differ from the value of 119875
0(267) is when the value
of 120576119905is 1 its value in a theoretical packed column devoid of
particles that is acapillaryFinally Figure 1(c) is another plot of the data in Table 1
but this time the values for 119875119905are plotted against particle
porosity 120576119901 The equation of this line does have an intercept
which represents the value of 119875119905when the particles are
nonporous It is also obvious from the equation of this linethat it is only when particle porosity is unity that is in acapillary that the value of 119875
119905is 267
It will be appreciated that since 120576119905is the sumof the internal
and external porosities of a packed column any given valuefor 119875
119905involving a column packed with porous particles can
represent many different combinations of these parametersVice versa any given value for 120576
119905can generate many different
values for 119875119905depending on the amount of particle porosity
which is present On the other hand as is pointed outearlier column internal porosity andor particle porosityplay no role in determining the permeability of a packedcolumn the only porosity parameter which is relevant forsuch purposes is column external porosity In other words asa permeability coefficient 119875
119905is not a particularly informative
concept because it still fails to provide any meaningful tie tothe underlying physical phenomena that govern the pressuredropfluid flow relationship
In the case of permeability measurements one quan-tifies only total pressure drop Pressure drop however isa commingling of the contributions of the several differ-ent parameters embodied in the pressureflow relationshipTherefore one must accurately represent the contributionof each factor before adding them together to equal thepressure drop Obviously if one mistakes the contribution ofany given element for instance particle size this will skewthe contribution of some other element when summed forinstance column porosity
Our frame of reference has been calibratedwith respect tothe interrelationships between the values of 119875
0 119875
119905 120576
119905 and 120576
119901
based on the data reported in Giddingsrsquo table 53-1 in his 1965textbook Since his data base was derived from experimentsusing nonporous particles the permeability results of whichwere extended to columns packed with fully porous particlesby using his Φ parameter the validity of which in turn wasestablished by independent means there is no uncertaintywith respect to the critical variable of external porosity 120576
0
inherent in our frame of reference It is for this reasontherefore that our frame of reference trumps any reporteddata which uses any other technique including the techniqueof inverse size exclusion chromatography by itself to establishvalues for 120576
0[12] In addition our frame of reference is
internally consistent due to the grounding of the parametersat the boundary condition of a theoretical capillary at whichpoint the values of 120576
119905and 120576
119901are unity and the values of 119875
0and
119875119905are equal In essence these interrelationships exemplify the
conservation laws of nature which are the ultimate standard
Journal of Materials 9
of measure when considering partial fractions of a singleentity
Consequently as Figures 1(b) and 1(c) illustrate if oneknows based upon other independent means the value of acolumnrsquos total porosity or its particlesrsquo porosity respectivelyone can at least verify whether results of permeability exper-iments expressed in terms of 119875
119905are reasonable Or to put
it another way if the 119875119905results reported do not conform to
what these plots tell us one should expect from columns andparticles based upon independently derived porosity valuesfor the particles under study we know that something iswrong Accordingly this is the way in which we use 119875
119905below
to analyze and evaluate Guiochonrsquos experimental work of thelast decade
8 The Origin of Guiochonrsquos Value forColumn Permeability
In 1966 Guiochon wrote a review paper entitled ldquoProblemsRaised by the Operation of Gas Chromatographic Columnsrdquo[13] In that paper he reviewed the development of the per-meability equation from Darcy to the then present Amongother things he reported that ldquoa mean 120576 of 038 is foundfor a number of packings obtained from powdered prod-ucts of various origins consisting of spherical or irregularparticles Almost all the measured values are between035 and 045rdquo (p 14) He then went on to discuss what hecalled the ldquosemiempirical Blake-Kozeny equationrdquo which hereported as 119896 = [119889
2
1199011205762][150(1 minus 120576)
2] [13 p 15 Equation
(11)] What sticks out of course is the fact that Guiochonstates here that 119875
0equals 150 the value for the permeability
coefficient championed by Ergun not the value posited byCarman To be fair we should point out that Guiochongave warning that he was uncomfortable with any particularvalue for the permeability coefficient For even though herecited the Kozeny-Blake equation with a value of 150 for itspermeability coefficient he stated that ldquothe values of 119896 givenby (this) equation are not very accurate and the deviationscan be as large as 50rdquo [13 p 15] Guiochon attributedmost ofthis uncertainty to the inability of practitioners of that era toeffectively measure the value of a columnrsquos external porositySince as he pointed out ldquo(the Kozeny-Blake) equation isvery sensitive to variations in 120576rdquo he concluded that theequation is ldquoof little userdquo and that it was better practice torely upon grosser measurements of permeability [13 p 15]Thus Guiochon at this timeframe made a conscious decisionto staywithin the relatively safe but admittedlymore obscureboundaries of Darcyrsquos teaching on column permeability
In 2006 however Guiochon emerged from under theshadow of Darcy permeability when he published a reviewpaper which purported to be a comprehensive summaryof the current state of the art in chromatography [14]In light of the fact that Guiochon and his coauthors hadvirtually contemporaneously published a new version of theirtextbook one would have expected that Guiochon wouldtake this opportunity to reaffirm his textbook teaching So itshould come as no surprise that he would open his discussionof the pressure dropfluid flow relationshipwith the following
assertion ldquothe permeability of packed beds used in liquidchromatography is in the order of 1 times 10minus3rdquo [14 p 9] Asone would expect the authority which Guiochon cites forthis seemingly familiar proposition is the 2006 edition of hisown textbook For the first time however he also providessome citations to third party sources In particular he cites a1997 chromatography handbook written by Uwe Neue ChiefScientist for Waters Corporation [15 p 9] As one can seeNeuersquos equationmdashlike Guiochonrsquosmdashdoes indeed postulate apermeability coefficient with a value of 1 times 10minus3
At this point however Guiochon abandons his tradi-tional caution about the value of the permeability coeffi-cient and marries his permeability equation to the Kozeny-Blake equation thereby endorsing not only the porositydependence term in that equation but also a specific valuefor 119875
0 Referring to his value of 1 times 10minus3 he states the
following ldquothe specific column permeability is related to thecolumn porosity through the Kozeny-Carman equation 119896
119891=
12057630180 (1 minus 120576
0)2rdquo [14 p 9] Not surprisingly Neue makes
the same connection ldquothe specific permeability depends onthe particle size 119889
119901and the interstitial porosity 120576
0of the
packed bedThe relationship is known as theKozeny-Carmanequation 119861
0= (1185)(1205763
01198892119901(1 minus 120576
0)2)rdquo [15 p 30]
Neuersquos permeability equation and Guiochonrsquos textbookpermeability equation however differ in that Neue uses thesuperficial not the mobile phase velocity Accordingly sinceall other features of their equations are the same one wouldexpect them to get different values for their permeabilitycoefficients Neue himself recognizes this when he says inhis handbook ldquothis value [1000] is also in good agreementwith our experience but values as low as 500 have beenreported in the literature8 However there are several reasonswhy different values reported by different workers do notrepresent true differences in the resistance parameter Firstlower values are obtained for porous packings if the specificpermeability is measured on the basis of the breakthroughtime of an unretained peak instead of using the flow raterdquo see[15 p 32] (when he distinguishes the ldquobreakthrough time ofan unretained peakrdquo from the ldquoflow raterdquo Neue is of coursereferring to the mobile phase velocity and the superficialvelocity resp) Consequently if one takes Neuersquos point andturns it on its head one can see that if he and Guiochonwere to get the same value for their permeability coefficients[1000] but one (Guiochon) is using mobile phase velocityto derive it and the other (Neue) uses superficial velocity toderive it this would indeed represent a ldquotrue differencerdquo intheir equations
So how can Guiochon cite both Neue and himself for theproposition in his review paper that ldquothe permeability of thepacked beds used in liquid chromatography is in the order of1times 10minus3rdquoThe answer is that at least for purposes of his reviewpaper he adopts Kozeny-Blakersquos fluid velocity convention inplace of the velocity frame he uses in his textbook Here iswhat he says ldquothis approximation [1 times 10minus3] is valid only ifthe superficial velocity is usedrdquo [14 p 9] (emphasis supplied)
As a matter of pure mathematics Guiochonrsquos permeabil-ity coefficient of 1 times 10minus3 can only be equal to 1205763
0180(1 minus 120576
0)2
if the value of 1205760is 04 the typical value for the interstitial
10 Journal of Materials
fraction of a well-packed column Guiochon confirms in hisreview paper that this is indeed what he assumes in his state-ment of the Kozeny-Blake equation [14 p 9] Accordinglyhis adoption of the value of 180 for 119875
0solely depends on the
validity of his assertion that the value of the permeabilitycoefficient is 1 times 10minus3
We already know from our discussion of Guiochonrsquostextbook teaching that when this value is associated with apermeability equation based on mobile phase velocity it iswrong As reflected in our Table 2 when Guiochonrsquos workedexample involving this figure is corrected to account for theembedded factor due to the irregularity of the particles whichhis worked example assumes the value of the permeabilitycoefficient generated by his equation is not 1 times 10minus3 it is 166times 10minus3
So where did this value of 1 times 10minus3 come from AlthoughGuiochon in his review paper cites Neuersquos handbook for thisvalue Neue himself offers absolutely no authority for it Onthe other hand Guiochon does provide one other citation inhis review paper for this value The citation is to an article byHalasz et al published in 1975 [16]
If one then goes to that article one sees that Halaszet al do indeed proffer the same equation as Neue (albeitin a somewhat different format) and yes it does containthe value of 1 times 10minus3 See [16 Equation (1)] Neverthelessthe authors fail to identify from whence they obtained thisvalue On the other hand they do make reference to apaper that Endele et al wrote a year earlier in 1974 withKlaus Unger entitled ldquoInfluence of the Particle Size (5ndash35 120583)of Spherical Silica on Column Efficiencies in High-PressureLiquid Chromatographyrdquo [17] It appears that this is wherethe body is buried for it is in this article that Halasz reportshaving actually derived a value of 1 times 10minus3 for the permeabilitycoefficient
In summarizing the results of their experiments whichincidentally are based on measurements of superficial notmobile phase velocity (see their Equation (7)) Halasz et alcalculate an average value for the permeability coefficient(which they notate as119870
119865) of 1198892
1199011080 Accordingly they state
the following ldquo[I]t is proposed to define an average particlesize 119889
119901 for columns packed with spherical silica using the
balanced density packing method by the equation 1198892119901
=
1198701198651000 = 1198651205781198711031199032120587Δ119875rdquo [17 p 382 their Equation (7)]To recapitulate it thus appears that (1) Halasz was the
source of the value of 1 times 10minus3 for the permeability coefficientwhich Guiochon adopts in his review paper (2)Halaszrsquo valuewas based upon measurements which involved sphericalparticles and (3)Halaszrsquo value was properly derived from anequation which utilized the superficial velocity as the fluidvelocity
However we still have a problem if we are going to acceptthe value of 1 times 10minus3 as a basis for calculating the constantin Kozeny-Blake we still need to know the value of Halaszrsquocolumnsrsquo external porosity For example in the absenceof measured values for the external porosity Guiochonrsquosassertion in his review paper that the value of the constantis 180 is without foundation
Unfortunately however Halasz bases his methodologyon ldquoassumedrdquo versus ldquomeasuredrdquo values for external porosityNevertheless he does provide some clues that may helpus peel back this onion First he says that his value forthe permeability coefficient applies to ldquocolumns using thebalanced density packing methodrdquo Secondly after statinghis proposed value for the coefficient we find the followingtelltale sentence ldquothe permeability of columns packed by theconventional ldquodryrdquo method are worse by a factor of about 2than those packed by the balanced density methodrdquo [17 p382] For this proposition Halasz cites yet another of his ownarticles this time a paper written by himself and M Naefein 1972 entitled ldquoInfluence of Column Parameters on PeakBroadening in High-Pressure Liquid Chromatographyrdquo [18]
When one follows this thread by going to Halaszrsquos 1972paper one finds a possible solution to the puzzle For this iswhere Halasz describes his calculations of the permeabilitycoefficient from experiments with a number of conventionaldry-packed columns containing spherical silica particlesNot surprisingly the resulting value for the permeabilitycoefficient for such columns was not 1 times 10minus3 Instead Halaszsays that his experiments with such columns resulted inthe following value for the permeability coefficient ldquo119870 sim
11988921199012000rdquo [18 p 78] In other words consistent with what
Halasz says in his 1974 article the value for the coefficient forthese dry-packed columns was indeed ldquoworse by a factor ofabout 2rdquo compared to its value for columns prepared with thebalanced density method
The obvious implication of this is that when Halasz gota value of 1 times 10minus3 for the permeability coefficient in 1974he must have been evaluating columns that were packed toan interstitial fraction considerably higher than the externalporosity of the columns that were the subject of his 1972experiments Our problem however is still not resolvedbecause Halasz did not accurately determine the externalporosity of the columns under study in either the 1972 or 1974columns
We surmise that in their 1972 paper Halasz and hiscoauthors made use of the teaching of Giddings outlinedin his 1965 textbook which at the time of their 1972 paperwas seven years old Here is what they say ldquoexperiencehas shown that in a regular packed column with the sievefractions usual for chromatography 120576
0is independent of the
particle size if the particles are more or less spherical In anacceptable approximation 120576
0is equal to 04 In those columns
packed with nonporous supports 119906V and 119906 are identicalUsing porous supports (ie silica gel alumina chromosorbPorasil and so forth) 119906 = 4119865(1205871198892
119888120576119879) where 120576
119879is the
total porosity and indicates the pore volume of the supportrdquoIt will be appreciated that 119906V stands for interstitial velocity(our 120583
119894) and 119906 stands for mobile phase velocity (our 120583
119905) The
ldquoexperiencerdquo which Halasz and his coauthors are referringto concerning ldquocolumns packed with nonporous supportsalumina and chromosorbrdquo is that recorded by Giddings inTable 53-1 in his 1965 text Continuing to establish theinterrelationships inherent in the various entities involvedin chromatographic columns with respect to both interstitialandmobile phase velocity the authors conclude ldquoformally 119906V
Journal of Materials 11
can be calculated for a porous support assuming 1205760is equal
to 04rdquo (emphasis supplied) The authors then announce themethodology underlying their frame of reference with regardto column permeability as follows ldquowe determined in all ofour columns the 119906119906V ratio to be 056 with a differentialrefractometer Consequently the total porosity 120576
119905was equal
to 0714rdquo It will be appreciated that their ratio of 119906119906V is oneand the same as GiddingsrsquoΦ parameter as utilized by him inhis 1965 textbook and as we define hereinabove
We can now identify the origin of the discrepancybetween Halasz and Giddings and by extension betweenHalasz and Guiochon (and Neue) For even though Halaszmakes use of the results in Giddingsrsquo Table 53-1 he failsto appreciate the teaching inherent therein regarding theimportance of an accurate value for column external porositywhen using porous particles In establishing his values forΦGiddings was careful to ground his values for the externalporosity of columns packed with porous particles in acomparison of their measured pressure drops with measuredpressure drops in columns packed with nonporous particleswhere the values for (external) porositywere accurate becausethey are equal to the columnsrsquo total porosity
In their 1972 paper Halasz et al used a refractometer todetect the inert solute 119899-pentane in the mobile phase of 119899-heptaneThey injected the 119899-pentane into the flowing mobilephase the flow rate of which they presumably measured witha volumetric cylinder and stop watch Since the exit timeof the 119899-pentane peak was identified by the refractometertheir measurement of holdup volume was accurate andaccordingly so was their value of 0714 for 120576
119905(there is no
uncertainty related to column total porosity when using inertsolutes)This in turn led to an accurate value for their119906 termmobile phase velocity However since they did not measurethe interstitial velocity 119906V but rather used an estimate basedupon the measured flow rate and an ldquoassumedrdquo value of 04for external porosity their calculatedestimated value for 119906Vwas arbitrary This in turn means that the value of 056 fortheirΦ ratio was likewise arbitrary
In Table 3 herein we show the results for column number1 in the 1972 study which we designate as Column A
1 Note
that the authorsrsquo raw values correspond to a value of 500 for1198750 a valuewhich is obviously far too high As is plainly evident
in our Figures 2 and 3 this column data set does not conformin any reasonable fashion to the norms in our referencecharts Accordingly the data is badly flawed In our correcteddata for Column A
1in Table 3 herein we adjust the column
external porosity to the lowest value permitted in our frame ofreference (038) This correction results in a revised value forGiddingsrsquoΦparameter of 0532 In additionwe are compelledto adjust the particle downward (to 11 micrometers approx)as is also dictated by our frame of reference We justify ourlow corrected value for external porosity and our downwardadjustment for particle size on the aggressive ldquodry-packingrdquoprocess described by the authors ldquo[T] the best results wereachieved with the following method the column (id =2mm) 25 cm long was packed with 25 equal portions of theldquobrushrdquo After adding each portion the column was vibratedand afterward tapped on the floor and also vigorously tappedwith a rod from the siderdquo Based upon a column packing
experience of the principal author herein spanningmore than30 years we believe that these packing conditions generateda packed column with a low external porosity and also one inwhich the particles experienced significant breakage
The main differences between the columns in the 1972and 1974 papers concerned the particle sizes and themethodsused to pack the columns The particle diameters in the 1972study fell in a range of about 15ndash200 micrometers while thosein the 1974 study were much smaller being in the range of 5ndash40 micrometers approximately Moreover while the columnsin the 1972 study were dry-packed the columns in the 1974study were slurry-packed (what Halasz calls ldquothe balanceddensity packing methodrdquo) The authors of the 1974 paperrecite that their permeability results shown in their Table 4correspond to values for 120601
0of 1080 and values for 120601 of 910
In Table 3 herein we show the results for the 1974 study as anaverage value in our column designated as A
2 Note that the
raw data reported by the authors which again has the built-inassumption that the column external porosity was 04 placesthe value of 119875
0at 192 In our corrected values for Column A
2
we show that a value of 267 for 1198750places the column external
porosity at a value of 043 The ratio of the value for 120601 of 910when compared to the value for 120601 in the 1972 study of 2007is about 22 (2007910 = 22) which is in good agreementwith the authorsrsquo statement that the permeability of the 1972columns was worse by a factor of about 2 Again based onthe experience of the principal author herein in the packingof hundreds of thousands of commercially sold slurry packedcolumns we conclude that the balanced density techniquedescribed by the authors in the 1974 paper is fully capable ofproducing columns with this high and external porosity
In summary we conclude that for purposes of theKozeny-Blake equation (a) the external porosity of theHalasz 1972 columns was 038 (b) the external porosity of thecolumns which he tested in 1974 was 043 and (c) Halaszrsquos1972 and 1974 studies when properly analyzed corroborateGiddingsrsquo value for 119875
0of 267 not Carmanrsquos value of 180
Although the fault for all of the misdirection about the valueof the permeability coefficient being 1 times 10minus3 therefore restsprimarily on the shoulders of Halasz one has to question whyGuiochon (andNeue) so readily adopted itWhatevermay bethe reason Guiochonrsquos support for this proposition has beenand continues to be a cause for much of the conventionalacceptance of Carmanrsquos (erroneous) figure of 180 as thecharacteristic value of the coefficient in the Kozeny-Blakeequation
9 Guiochonrsquos Experimental Data
In addition to the data from Halasz 1972 and 1974 studiesTable 3 herein contains data from a representative sampleof Guiochonrsquos personal experimental permeability investi-gations over approximately the last decade as reported inthe numerous trade journals particularly the Journal ofChromatography A In each case the table has a single rowdedicated to Guiochonrsquos reported raw data and a single rowdedicated to corrected values for each column based on thereference chart in Table 1 herein
12 Journal of MaterialsTa
ble3Summaryexperim
entald
ata
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Naefe
andHalasz
columns
1198601(19
72)
Colum
nA1rawdata
07140
0560
040
000314
0523
782007
2811
4016
563
498119864minus04
500
357
908119864minus10
648119864minus10
2500
020
100
000135
0001350
00038
100
053
132
074
Colum
nA1corrected
07140
0532
03800
0334
0539
7813
3518
705002
701
749119864minus04
267
191
908119864minus10
648119864minus10
2500
020
100
000110
000110
100038
100
053
139
074
Endelean
dHalasz
columns
1198602(19
74)
Colum
nA2rawdata
08426
0475
040
00044
30738
11910
1080
4740
563
110119864minus03
192
162
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
033
016
Colum
nA2corrected
08426
0511
04309
0412
0723
11910
1080
3411
405
110119864minus03
267
225
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
031
016
Guiochon
columns
1198612ndash1198614(19
98)
Colum
nB 2
rawdata
05639
0722
040
730157
0264
22771
1367
2931
520
130119864minus03
263
148
467119864minus10
263119864minus10
429
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 3
rawdata
05582
0704
03932
0165
0272
44764
1369
3382
606
131119864minus03
226
126
471119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 4
rawdata
05509
0710
03909
0160
0263
72784
1422
3422
621
128119864minus03
229
126
459119864minus10
253119864minus10
1328
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 3
corrected
05653
0723
040
860157
0265
44774
1369
2899
513
129119864minus03
267
151
465119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 4
corrected
05651
0723
040
830157
0265
72776
1374
2907
514
129119864minus03
267
151
464119864minus10
262119864minus10
1375
500
100
000
060
000
0600
00165
98008
020
015
Guiochon
columnC
(1999)
Colum
nCrawdata
05930
0673
03990
0194
0323
183
865
1459
3372
569
116119864minus03
257
152
109119864minus09
645119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Colum
nCcorrected
05930
0673
03990
0194
0323
190
900
1518
3372
569
111119864minus03
267
158
105119864minus09
620119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Guiochon
columns
1198631ndash1198636(2006)
Colum
nD
1rawdata
07830
0456
03570
0426
0663
86694
886
7115
909
144119864minus03
975
76334119864minus10
261119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
013
Colum
nD
2rawdata
07230
0491
03550
0368
0571
88656
907
6726
930
152119864minus03
975
70353119864minus10
255119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
014
Colum
nD
3rawdata
06850
0506
03466
0338
0518
97685
1000
7025
1026
146119864minus03
975
67338119864minus10
231119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD
4rawdata
06560
0521
03415
0315
0478
103
696
1062
7143
1089
144119864minus03
975
64332119864minus10
218119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD5rawdata
06190
0545
03371
0282
0425
109
692
1118
7100
1147
144119864minus03
975
60334119864minus10
207119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
016
Colum
nD
6rawdata
05760
0587
03383
0238
0359
107
635
1103
6516
1131
157119864minus03
975
56364119864minus10
210119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
017
Colum
nD
1corrected
07830
0542
04243
0359
0623
86907
1158
3397
434
110119864minus03
267
209
334119864minus10
261119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
013
Colum
nD
2corrected
07230
0584
04221
0301
0521
88857
1186
3212
444
117119864minus03
267
193
353119864minus10
255119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
014
Colum
nD
3corrected
06850
0603
04129
0272
0463
97896
1307
3354
490
112119864minus03
267
183
338119864minus10
231119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
015
Colum
nD
4corrected
06560
0621
040
730249
0420
103
911
1388
3411
520
110119864minus03
267
175
332119864minus10
218119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
015
Colum
nD
5corrected
06190
0650
04025
0217
0362
109
905
1462
3390
548
110119864minus03
267
165
334119864minus10
207119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
016
Colum
nD
6corrected
05760
0701
04038
0172
0289
107
831
1442
3111
540
120119864minus03
267
154
364119864minus10
210119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
017
Guiochon
columns
1198641-1198642
(2007)
Colum
nE 1
rawdata
064
500594
03830
0262
0425
356
1119
1735
4371
678
894119864minus04
256
165
804119864minus11
519119864minus11
1500
046
100
000
030
000
0300
000
41300
030
079
047
Colum
nE 2
rawdata
05400
0800
04320
0108
0190
470
1000
1853
2161
400
100119864minus03
463
250
729119864minus11
393119864minus11
1500
046
100
000
027
000
0270
000
41300
030
070
056
Colum
nE 1
corrected
064
500594
040
000245
040
8356
969
1502
3628
563
103119864minus03
267
172
804119864minus11
519119864minus11
1500
046
100
000
028
000
0279
000
41300
030
075
047
Colum
nE 2
corrected
05400
0800
04320
0108
0190
470
577
1068
2161
400
173119864minus03
267
144
729119864minus11
393119864minus11
1500
046
088
000
023
000
0205
000
41300
030
070
056
Guiochon
columns
1198651ndash1198654
(2008)
Colum
nF 1
rawdata
06350
0586
03720
0263
0419
197
725
1142
4865
766
138119864minus03
149
95399119864minus11
253119864minus11
300
021
100
000
017
000
0170
00035
100
048
129
076
Colum
nF 2
rawdata
064
200581
03730
0269
0429
361
807
1258
4863
758
124119864minus03
166
107
358119864minus11
230119864minus11
500
021
100
000
017
000
0170
00035
100
048
129
075
Colum
nF 3
rawdata
064
100588
03770
0264
0424
670
748
1166
4643
724
134119864minus03
161
103
387119864minus11
248119864minus11
1000
021
100
000
017
000
0170
00035
100
048
128
075
Colum
nF 4
rawdata
06390
0595
03800
0259
0418
1014
752
1177
4476
701
133119864minus03
168
107
384119864minus11
246119864minus11
1500
021
100
000
017
000
0170
00035
100
048
127
075
Colum
nF 1
corrected
06350
0630
040
000235
0392
197
954
1502
3572
563
105119864minus03
267
170
399119864minus11
253119864minus11
300
021
100
000
019
000
0195
00035
100
048
120
076
Colum
nF 2
corrected
064
200623
040
000242
0403
361
964
1502
3611
563
104119864minus03
267
171
358119864minus11
230119864minus11
500
021
100
000
019
000
0186
00035
100
048
120
075
Colum
nF 3
corrected
064
100624
040
000241
0402
670
963
1502
360
6563
104119864minus03
267
171
386119864minus11
248119864minus11
1000
021
100
000
019
000
0193
00035
100
048
120
075
Colum
nF 4
corrected
06390
0626
040
000239
0398
1014
960
1502
3594
563
104119864minus03
267
171
384119864minus11
246119864minus11
1500
021
100
000
019
000
0192
00035
100
048
120
075
Journal of Materials 13
Table3Con
tinued
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Guiochon
columns
1198671-1198672serie
s(2010)
Colum
nH
1rawdata
05320
0744
03960
0136
0225
120
563
1057
3125
587
178119864minus03
180
96122119864minus10
652119864minus11
1500
046
100
000
026
000
0262
00104
050
005
013
009
Colum
nH
2rawdata
05420
0686
03720
0170
0271
61747
1379
4152
766
134119864minus03
180
98823119864minus11
446119864minus11
1000
046
100
000
025
000
0248
00054
050
005
013
009
Colum
nH
1corrected
05320
0752
040
000132
0220
120
799
1502
2993
563
125119864minus03
267
142
122119864minus10
652119864minus11
1500
046
100
000
031
000
0313
00104
050
005
013
009
Colum
nH
2corrected
05420
0738
040
000142
0237
61814
1502
3049
563
123119864minus03
267
145
823119864minus11
446119864minus11
1000
046
100
000
026
000
0259
00054
050
005
013
009
Guiochon
columns
1198673-1198674serie
s(2010)
Colum
nH
3rawdata
06270
060
903820
0245
0396
463
700
1117
4296
685
143119864minus03
163
102
447119864minus11
280119864minus11
1500
021
100
000
018
000
0177
00036
050
024
063
038
Colum
nH
4rawdata
05170
0791
040
900108
0183
433
616
1191
2639
511
162119864minus03
233
121
580119864minus11
300119864minus11
1500
021
100
000
019
000
0189
00036
050
024
059
047
Colum
nH
3corrected
06270
0638
040
000227
0378
463
942
1502
3527
563
106119864minus03
267
167
447119864minus11
280119864minus11
1500
021
100
000
021
000
0205
00036
050
024
060
038
Colum
nH
4corrected
05170
0791
040
900108
0183
433
699
1353
2639
511
143119864minus03
265
137
580119864minus11
300119864minus11
1500
021
100
000
020
000
0201
00036
050
024
059
047
Guiochon
columns
1198681-1198686
(2010)
Colum
nI 1rawdata
06580
0562
03700
0288
0457
488
741
1127
5156
784
135119864minus03
144
95390119864minus11
256119864minus11
500
021
100
000
017
000
0170
00104
050
024
065
037
Colum
nI 2rawdata
06550
0576
03770
0278
044
6873
661
1009
4745
724
151119864minus03
139
91437119864minus11
286119864minus11
1000
021
100
000
017
000
0170
00104
050
024
064
037
Colum
nI 3rawdata
06550
0588
03850
0270
0439
1209
610
931
4341
663
164119864minus03
141
92474119864minus11
310119864minus11
1500
021
100
000
017
000
0170
00104
050
024
062
037
Colum
nI 4rawdata
06430
0572
03680
0275
0435
260
787
1225
5153
801
127119864minus03
153
98367119864minus11
236119864minus11
500
030
100
000
017
000
0170
00104
050
012
032
018
Colum
nI 5rawdata
06540
0583
03810
0273
044
144
3683
1044
4531
693
146119864minus03
151
99423119864minus11
277119864minus11
1000
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 6rawdata
06550
0585
03830
0272
044
164
6665
1016
4438
678
150119864minus03
150
98434119864minus11
285119864minus11
1500
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 1corrected
06580
060
8040
000258
0430
488
988
1502
3701
563
101119864minus03
267
176
390119864minus11
256119864minus11
500
021
100
000
020
000
0196
00104
050
024
060
037
Colum
nI 2corrected
06550
0611
040
000255
0425
873
984
1502
3684
563
102119864minus03
267
175
437119864minus11
286119864minus11
1000
021
100
000
021
000
0207
00104
050
024
060
037
Colum
nI 3corrected
06550
0611
040
000255
0425
1209
984
1502
3684
563
102119864minus03
267
175
474119864minus11
310119864minus11
1500
021
100
000
022
000
0216
00104
050
024
060
037
Colum
nI 4corrected
06430
0622
040
000243
040
5260
966
1502
3617
563
104119864minus03
267
172
367119864minus11
236119864minus11
500
030
100
000
019
000
0188
00104
050
012
029
018
Colum
nI 5corrected
06540
0612
040
000254
0423
443
982
1502
3679
563
102119864minus03
267
175
423119864minus11
277119864minus11
1000
030
100
000
020
000
0204
00104
050
012
029
018
Colum
nI 6corrected
06550
0611
040
000255
0425
646
984
1502
3684
563
102119864minus03
267
175
434119864minus11
285119864minus11
1500
030
100
000
021
000
0207
00104
050
012
029
018
Guiochon
columns
1198691-1198692
(2011)
Colum
nJ 1rawdata90
∘A(120578
=00054
P)04980
0803
040
000098
0163
65654
1313
2801
563
153119864minus03
234
116126119864minus10
627119864minus11
1500
046
100
000
029
000
0287
00054
050
005
013
010
Colum
nJ 2rawdata90
∘A(120578
=00104
P)04980
0803
040
000098
0163
124
651
1307
2801
563
154119864minus03
232
116127119864minus10
630119864minus11
1500
046
100
000
029
000
0287
00104
050
005
013
010
Colum
nJ 1rawdata160
∘A(120578
=00054
P)05630
0714
04020
0161
0269
71896
1592
3099
550
112119864minus03
289
163
102119864minus10
573119864minus11
1500
046
100
000
030
000
0302
00054
050
005
012
009
Colum
nJ 2rawdata160
∘A(120578
=00104
P)05630
0714
04020
0161
0269
129
847
1504
3099
550
118119864minus03
273
154
108119864minus10
606119864minus11
1500
046
100
000
030
000
0302
00104
050
005
012
009
Colum
nJ 1corrected
04980
0803
040
000098
0163
65748
1502
2801
563
134119864minus03
267
133
126119864minus10
627119864minus11
1500
046
100
000
031
000
0307
00054
050
005
013
010
Colum
nJ 2corrected
04980
0803
040
000098
0163
124
748
1502
2801
563
134119864minus03
267
133
127119864minus10
630119864minus11
1500
046
100
000
031
000
0308
00104
050
005
013
010
Colum
nJ 1corrected
05630
0714
04020
0161
0269
71827
1470
3099
550
121119864minus03
267
150
102119864minus10
573119864minus11
1500
046
100
000
029
000
0290
00054
050
005
012
009
Colum
nJ 2corrected
05630
0714
04020
0161
0269
129
827
1470
3099
550
121119864minus03
267
150
108119864minus10
606119864minus11
1500
046
100
000
030
000
0299
00104
050
005
012
009
14 Journal of Materials
050
100150200250300350400
045 050 055 060 065 070 075 080 085
LineAll corrected data
Linear (all corrected data)Squares raw data Triangles corrected data
Pt
120576t
D seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
J1 J2 H4
H1H2
D6
B2 B3 B4
J3 J4
E2
C
H3
D5 D4
E1
D2
A1
D1
A2
y = 267xLine slope = 267
Figure 2 Summary of experimental results
0
500
1000
1500
2000
2500
20 25 30 35 40 45 50 55 60 65 70 75 80
Linear (line)
Ψ
LineAll corrected dataD seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
Line slope (P0) = 267
J1 J2
J3 J4E2
E1
A1
A2
H4 H1
C
120601
H2H3I1 I2 I3 I4 I5
F1 F2 F3 F4
D1 D2 D3 D4 D5 D6
Squares raw data Triangles corrected data
Figure 3 Summary of experimental results
Figures 2 and 3 herein are graphical representations ofTable 31015840s raw and corrected data for Guiochonrsquos columnsplotted in the same reference frames as in Figures 1(a) and1(b) respectively
As shown in the plots of the thirty-one total columnsevaluated in this data base only six columns had reportedraw data values for 119875
0close to the normThese were columns
B2 C E
1 H
4 J
1and J
2 Interestingly of the six conforming
columns 3 of the columns were packed with fully porousparticles and 3 columns with partially porous particles Itcan be seen however that when properly interpreted andadjusted all of Guiochonrsquos experimental data falls fairlywithin the norms outlined in the reference chart in Table 1In order to explore the reasons why the raw data does notconform more closely to the norms outlined in Table 1 eachof Guiochonrsquos studies included in Table 3 is evaluated in turn
Finally Table 4 herein is a list of symbols and parameterdefinitions used throughout this paper as well as their unitsof measure
91 Guiochonrsquos 1998 Study In a 1998 publication by Kohand Guiochon entitled ldquoEffect of the Column Length on theCharacteristics of the Packed Bed and the Column Efficiencyin a Dynamic Axial Compression Columnrdquo [19] Guiochonand his coauthor report the results of a study they hadcarried out on column packing using a rather novel techniquewhich was a combination of slurry packing and mechanicalcompression
The study involved the packing of a single cylinder ofdiameter 5 cm to various column lengths ranging between 1and 25 cm approximately The particles were all silica basedand coated with a reverse stationary phase C
18 However two
different categories of particles were involved One categorycontained highly irregular particles which were about 130micrometers in average diameter In addition these particleshad a high specific pore volume (11mLg) and exhibitedsignificant particle breakage under the packing conditionsinvestigated in the study Conversely the other particlecategory involved very small particles (6 micrometers) whichwere spherical in shape had a low specific pore volume(03mLg) and exhibited a greater ability to withstandparticle breakage under the studyrsquos packing conditions
The permeability data for the irregular particles isneglected herein because according to the authors theparticles exhibited significant breakage during packing andwithout an accurate value for the characteristic dimensionof the particle the data cannot be used in the Kozeny-Blake model to determine the value of 119875
0 The spherical
particles on the other hand behaved successfully under thepacking conditions chosen and yielded values calculated bythe authors for 119875
0of 619 268 226 and 229 for the four
different lengths of columns studied The data for thesecolumns is reported in our Table 3 in the rows for ColumnsB1through B
4 The higher value of 619 (the shortest column)
was rejected by the authors because they felt that either theparticles had been somewhat broken during packing or thecolumn frits had been blocked with fines Accordingly thepermeability data for this column is likewise neglected herein
The remaining three columns however were not con-sistent with respect to the authorsrsquo calculated values for 119875
0
Since the particles were identical the range of values for 1198750
of 226 to 268 must be due to experimental error The rawdata for column B
2 which produced a 119875
0value of 268 is
all self-consistent For example it generates a value for 119875119905
of 148 which is a reasonable value based upon the knownlow particle porosity for these Nova Pak particles and acorresponding value for 120576
0of 04073 which is a reasonable
value for a well-packed column However the other twocolumns columns B
3and B
4 had the much lower value of
126 for 119875119905 Accordingly in Table 3 herein corrected values
for column external porosity for columns B3and B
4are
displayed These corrected values are only slightly differentthan the reported values are within the experimental errorof the measurement and therefore in combination with the
Journal of Materials 15
Table 4 Glossary of terms
Symbol Unit dimensions (cgs) Definition119860 cm2 Column cross-sectional area119863 cm Column diameter119889119901
cm Particle spherical diameter equivalentΔ119875 gcmminus1secminus2 Pressure drop along column119889119901119898
cm Average particle diameter (measured)1198960
None Guiochonrsquos residual reciprocal permeability coefficient119875119905
None Guiochonrsquos modified permeability coefficient in Kozeny-Blake equation119871 cm Column length1198750
None Permeability coefficient in Kozeny-Blake equation119876 cm3minminus1 (exception) Volumetric fluid flow rate1199050
sec Time for elution of totally permeating unretained solute119896 cm2 General permeability coefficient in Darcy equation119870
119865cm2 Halaszrsquos permeability coefficient in Darcy equation (1974 paper)
119870 cm2 Halaszrsquos permeability coefficient in Darcy equation (1972 paper)119870
119888None Guiochonrsquos permeability coefficient in the Kozeny-Blake equation (as modified by Carmen)
Greek1205760
None External column porosity120576119894
None Internal column porosity120576119905
None Total column porosity120576119901
None Particle porosityΦ None Giddingsrsquo ratio of external and total column porosity120601 None Flow resistance parameter based on mobile phase velocity1206010
None Flow resistance parameter based on superficial velocityΩ
119901None Particle sphericity factor
120578 gcmminus1secminus1 Fluid absolute viscosity120583119894
cmsecminus1 Average fluid interstitial linear velocity120583119904
cmsecminus1 Average fluid superficial linear velocity120583119905
cmsecminus1 Average mobile phase velocityΨ None Porosity dependence term in the Kozeny-Blake equation based on mobile phase velocityΨ
0None Porosity dependence term in the Kozeny-Blake equation based on superficial velocity
raw data for column B2 are supportive of the value of 267 for
1198750
92 Guiochonrsquos 1999 Study In a 1999 publication by Farkaset al entitled ldquoValidity of Darcyrsquos Law at Low Flow Rates inLiquid Chromatographyrdquo [20] Guiochon and his coauthorsreport the results of some very exactingmeasurements whichthey made to validate Darcyrsquos law at extremely low fluidflow rates Using a column containing particles packed toa measured external column porosity of 0399 and havinga measured total column porosity of 0593 (Figure 2 in thepaper) Guiochon et al measured the column pressure dropand fluid flow rate at flow rates ranging between 0015 and05mLminwith ethylene glycol as the fluidThe authors statethat the particles in the column are spherical and they appearto have gone to extraordinary lengths to obtain an accuratemeasure of the particle diameter and of course as a result ofmodern techniques of particle size classification the particlesize distribution is very narrow
In Figure 1 of the paper the authors show a plot ofcolumn measured pressure drop in bar versus ldquofluid linearvelocityrdquo in cmsec This plot represents their measurementstaken in the empty column The only velocity that exists in
an empty column is the superficial velocity and thus theabscissa values in the plot in Figure 1 designated as ldquofluidlinear velocityrdquo represent superficial linear velocity
In Figure 2 in their paper however the authors showanother plot of measured pressure drop in bar also versusldquofluid linear velocityrdquo in cmsec The velocity on the abscissain this plot although also designated as ldquofluid linear velocityrdquodoes not equate (for permeability related purposes) to theldquofluid linear velocityrdquo in Figure 1 This is because in Figure 2the columnwas filled with porous particles and themeasuredvelocity was the mobile phase velocity Indeed the onlyvelocity used throughout the entire paper is the mobile phasevelocity (mobile phase velocity and superficial velocity areone and the same in an empty column)
In Table 1 of their paper the authors report that for thecolumn represented in Figure 2 the slope of a line achievedwhen the mobile phase velocityin units of cmsec is plottedversus pressure drop in units of bar is 1829 They also reportthat the flow resistance parameter 120601 for this column is 865In Table 3 herein it can be seen that the raw data for thiscolumn (Column C) generates an apparent value of 257 for1198750 This value is very close to Giddingsrsquo value of 267 Indeed
by referring to Figures 2 and 3 herein one can see that thereported raw data in this study without any modification
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
Journal of Materials 9
of measure when considering partial fractions of a singleentity
Consequently as Figures 1(b) and 1(c) illustrate if oneknows based upon other independent means the value of acolumnrsquos total porosity or its particlesrsquo porosity respectivelyone can at least verify whether results of permeability exper-iments expressed in terms of 119875
119905are reasonable Or to put
it another way if the 119875119905results reported do not conform to
what these plots tell us one should expect from columns andparticles based upon independently derived porosity valuesfor the particles under study we know that something iswrong Accordingly this is the way in which we use 119875
119905below
to analyze and evaluate Guiochonrsquos experimental work of thelast decade
8 The Origin of Guiochonrsquos Value forColumn Permeability
In 1966 Guiochon wrote a review paper entitled ldquoProblemsRaised by the Operation of Gas Chromatographic Columnsrdquo[13] In that paper he reviewed the development of the per-meability equation from Darcy to the then present Amongother things he reported that ldquoa mean 120576 of 038 is foundfor a number of packings obtained from powdered prod-ucts of various origins consisting of spherical or irregularparticles Almost all the measured values are between035 and 045rdquo (p 14) He then went on to discuss what hecalled the ldquosemiempirical Blake-Kozeny equationrdquo which hereported as 119896 = [119889
2
1199011205762][150(1 minus 120576)
2] [13 p 15 Equation
(11)] What sticks out of course is the fact that Guiochonstates here that 119875
0equals 150 the value for the permeability
coefficient championed by Ergun not the value posited byCarman To be fair we should point out that Guiochongave warning that he was uncomfortable with any particularvalue for the permeability coefficient For even though herecited the Kozeny-Blake equation with a value of 150 for itspermeability coefficient he stated that ldquothe values of 119896 givenby (this) equation are not very accurate and the deviationscan be as large as 50rdquo [13 p 15] Guiochon attributedmost ofthis uncertainty to the inability of practitioners of that era toeffectively measure the value of a columnrsquos external porositySince as he pointed out ldquo(the Kozeny-Blake) equation isvery sensitive to variations in 120576rdquo he concluded that theequation is ldquoof little userdquo and that it was better practice torely upon grosser measurements of permeability [13 p 15]Thus Guiochon at this timeframe made a conscious decisionto staywithin the relatively safe but admittedlymore obscureboundaries of Darcyrsquos teaching on column permeability
In 2006 however Guiochon emerged from under theshadow of Darcy permeability when he published a reviewpaper which purported to be a comprehensive summaryof the current state of the art in chromatography [14]In light of the fact that Guiochon and his coauthors hadvirtually contemporaneously published a new version of theirtextbook one would have expected that Guiochon wouldtake this opportunity to reaffirm his textbook teaching So itshould come as no surprise that he would open his discussionof the pressure dropfluid flow relationshipwith the following
assertion ldquothe permeability of packed beds used in liquidchromatography is in the order of 1 times 10minus3rdquo [14 p 9] Asone would expect the authority which Guiochon cites forthis seemingly familiar proposition is the 2006 edition of hisown textbook For the first time however he also providessome citations to third party sources In particular he cites a1997 chromatography handbook written by Uwe Neue ChiefScientist for Waters Corporation [15 p 9] As one can seeNeuersquos equationmdashlike Guiochonrsquosmdashdoes indeed postulate apermeability coefficient with a value of 1 times 10minus3
At this point however Guiochon abandons his tradi-tional caution about the value of the permeability coeffi-cient and marries his permeability equation to the Kozeny-Blake equation thereby endorsing not only the porositydependence term in that equation but also a specific valuefor 119875
0 Referring to his value of 1 times 10minus3 he states the
following ldquothe specific column permeability is related to thecolumn porosity through the Kozeny-Carman equation 119896
119891=
12057630180 (1 minus 120576
0)2rdquo [14 p 9] Not surprisingly Neue makes
the same connection ldquothe specific permeability depends onthe particle size 119889
119901and the interstitial porosity 120576
0of the
packed bedThe relationship is known as theKozeny-Carmanequation 119861
0= (1185)(1205763
01198892119901(1 minus 120576
0)2)rdquo [15 p 30]
Neuersquos permeability equation and Guiochonrsquos textbookpermeability equation however differ in that Neue uses thesuperficial not the mobile phase velocity Accordingly sinceall other features of their equations are the same one wouldexpect them to get different values for their permeabilitycoefficients Neue himself recognizes this when he says inhis handbook ldquothis value [1000] is also in good agreementwith our experience but values as low as 500 have beenreported in the literature8 However there are several reasonswhy different values reported by different workers do notrepresent true differences in the resistance parameter Firstlower values are obtained for porous packings if the specificpermeability is measured on the basis of the breakthroughtime of an unretained peak instead of using the flow raterdquo see[15 p 32] (when he distinguishes the ldquobreakthrough time ofan unretained peakrdquo from the ldquoflow raterdquo Neue is of coursereferring to the mobile phase velocity and the superficialvelocity resp) Consequently if one takes Neuersquos point andturns it on its head one can see that if he and Guiochonwere to get the same value for their permeability coefficients[1000] but one (Guiochon) is using mobile phase velocityto derive it and the other (Neue) uses superficial velocity toderive it this would indeed represent a ldquotrue differencerdquo intheir equations
So how can Guiochon cite both Neue and himself for theproposition in his review paper that ldquothe permeability of thepacked beds used in liquid chromatography is in the order of1times 10minus3rdquoThe answer is that at least for purposes of his reviewpaper he adopts Kozeny-Blakersquos fluid velocity convention inplace of the velocity frame he uses in his textbook Here iswhat he says ldquothis approximation [1 times 10minus3] is valid only ifthe superficial velocity is usedrdquo [14 p 9] (emphasis supplied)
As a matter of pure mathematics Guiochonrsquos permeabil-ity coefficient of 1 times 10minus3 can only be equal to 1205763
0180(1 minus 120576
0)2
if the value of 1205760is 04 the typical value for the interstitial
10 Journal of Materials
fraction of a well-packed column Guiochon confirms in hisreview paper that this is indeed what he assumes in his state-ment of the Kozeny-Blake equation [14 p 9] Accordinglyhis adoption of the value of 180 for 119875
0solely depends on the
validity of his assertion that the value of the permeabilitycoefficient is 1 times 10minus3
We already know from our discussion of Guiochonrsquostextbook teaching that when this value is associated with apermeability equation based on mobile phase velocity it iswrong As reflected in our Table 2 when Guiochonrsquos workedexample involving this figure is corrected to account for theembedded factor due to the irregularity of the particles whichhis worked example assumes the value of the permeabilitycoefficient generated by his equation is not 1 times 10minus3 it is 166times 10minus3
So where did this value of 1 times 10minus3 come from AlthoughGuiochon in his review paper cites Neuersquos handbook for thisvalue Neue himself offers absolutely no authority for it Onthe other hand Guiochon does provide one other citation inhis review paper for this value The citation is to an article byHalasz et al published in 1975 [16]
If one then goes to that article one sees that Halaszet al do indeed proffer the same equation as Neue (albeitin a somewhat different format) and yes it does containthe value of 1 times 10minus3 See [16 Equation (1)] Neverthelessthe authors fail to identify from whence they obtained thisvalue On the other hand they do make reference to apaper that Endele et al wrote a year earlier in 1974 withKlaus Unger entitled ldquoInfluence of the Particle Size (5ndash35 120583)of Spherical Silica on Column Efficiencies in High-PressureLiquid Chromatographyrdquo [17] It appears that this is wherethe body is buried for it is in this article that Halasz reportshaving actually derived a value of 1 times 10minus3 for the permeabilitycoefficient
In summarizing the results of their experiments whichincidentally are based on measurements of superficial notmobile phase velocity (see their Equation (7)) Halasz et alcalculate an average value for the permeability coefficient(which they notate as119870
119865) of 1198892
1199011080 Accordingly they state
the following ldquo[I]t is proposed to define an average particlesize 119889
119901 for columns packed with spherical silica using the
balanced density packing method by the equation 1198892119901
=
1198701198651000 = 1198651205781198711031199032120587Δ119875rdquo [17 p 382 their Equation (7)]To recapitulate it thus appears that (1) Halasz was the
source of the value of 1 times 10minus3 for the permeability coefficientwhich Guiochon adopts in his review paper (2)Halaszrsquo valuewas based upon measurements which involved sphericalparticles and (3)Halaszrsquo value was properly derived from anequation which utilized the superficial velocity as the fluidvelocity
However we still have a problem if we are going to acceptthe value of 1 times 10minus3 as a basis for calculating the constantin Kozeny-Blake we still need to know the value of Halaszrsquocolumnsrsquo external porosity For example in the absenceof measured values for the external porosity Guiochonrsquosassertion in his review paper that the value of the constantis 180 is without foundation
Unfortunately however Halasz bases his methodologyon ldquoassumedrdquo versus ldquomeasuredrdquo values for external porosityNevertheless he does provide some clues that may helpus peel back this onion First he says that his value forthe permeability coefficient applies to ldquocolumns using thebalanced density packing methodrdquo Secondly after statinghis proposed value for the coefficient we find the followingtelltale sentence ldquothe permeability of columns packed by theconventional ldquodryrdquo method are worse by a factor of about 2than those packed by the balanced density methodrdquo [17 p382] For this proposition Halasz cites yet another of his ownarticles this time a paper written by himself and M Naefein 1972 entitled ldquoInfluence of Column Parameters on PeakBroadening in High-Pressure Liquid Chromatographyrdquo [18]
When one follows this thread by going to Halaszrsquos 1972paper one finds a possible solution to the puzzle For this iswhere Halasz describes his calculations of the permeabilitycoefficient from experiments with a number of conventionaldry-packed columns containing spherical silica particlesNot surprisingly the resulting value for the permeabilitycoefficient for such columns was not 1 times 10minus3 Instead Halaszsays that his experiments with such columns resulted inthe following value for the permeability coefficient ldquo119870 sim
11988921199012000rdquo [18 p 78] In other words consistent with what
Halasz says in his 1974 article the value for the coefficient forthese dry-packed columns was indeed ldquoworse by a factor ofabout 2rdquo compared to its value for columns prepared with thebalanced density method
The obvious implication of this is that when Halasz gota value of 1 times 10minus3 for the permeability coefficient in 1974he must have been evaluating columns that were packed toan interstitial fraction considerably higher than the externalporosity of the columns that were the subject of his 1972experiments Our problem however is still not resolvedbecause Halasz did not accurately determine the externalporosity of the columns under study in either the 1972 or 1974columns
We surmise that in their 1972 paper Halasz and hiscoauthors made use of the teaching of Giddings outlinedin his 1965 textbook which at the time of their 1972 paperwas seven years old Here is what they say ldquoexperiencehas shown that in a regular packed column with the sievefractions usual for chromatography 120576
0is independent of the
particle size if the particles are more or less spherical In anacceptable approximation 120576
0is equal to 04 In those columns
packed with nonporous supports 119906V and 119906 are identicalUsing porous supports (ie silica gel alumina chromosorbPorasil and so forth) 119906 = 4119865(1205871198892
119888120576119879) where 120576
119879is the
total porosity and indicates the pore volume of the supportrdquoIt will be appreciated that 119906V stands for interstitial velocity(our 120583
119894) and 119906 stands for mobile phase velocity (our 120583
119905) The
ldquoexperiencerdquo which Halasz and his coauthors are referringto concerning ldquocolumns packed with nonporous supportsalumina and chromosorbrdquo is that recorded by Giddings inTable 53-1 in his 1965 text Continuing to establish theinterrelationships inherent in the various entities involvedin chromatographic columns with respect to both interstitialandmobile phase velocity the authors conclude ldquoformally 119906V
Journal of Materials 11
can be calculated for a porous support assuming 1205760is equal
to 04rdquo (emphasis supplied) The authors then announce themethodology underlying their frame of reference with regardto column permeability as follows ldquowe determined in all ofour columns the 119906119906V ratio to be 056 with a differentialrefractometer Consequently the total porosity 120576
119905was equal
to 0714rdquo It will be appreciated that their ratio of 119906119906V is oneand the same as GiddingsrsquoΦ parameter as utilized by him inhis 1965 textbook and as we define hereinabove
We can now identify the origin of the discrepancybetween Halasz and Giddings and by extension betweenHalasz and Guiochon (and Neue) For even though Halaszmakes use of the results in Giddingsrsquo Table 53-1 he failsto appreciate the teaching inherent therein regarding theimportance of an accurate value for column external porositywhen using porous particles In establishing his values forΦGiddings was careful to ground his values for the externalporosity of columns packed with porous particles in acomparison of their measured pressure drops with measuredpressure drops in columns packed with nonporous particleswhere the values for (external) porositywere accurate becausethey are equal to the columnsrsquo total porosity
In their 1972 paper Halasz et al used a refractometer todetect the inert solute 119899-pentane in the mobile phase of 119899-heptaneThey injected the 119899-pentane into the flowing mobilephase the flow rate of which they presumably measured witha volumetric cylinder and stop watch Since the exit timeof the 119899-pentane peak was identified by the refractometertheir measurement of holdup volume was accurate andaccordingly so was their value of 0714 for 120576
119905(there is no
uncertainty related to column total porosity when using inertsolutes)This in turn led to an accurate value for their119906 termmobile phase velocity However since they did not measurethe interstitial velocity 119906V but rather used an estimate basedupon the measured flow rate and an ldquoassumedrdquo value of 04for external porosity their calculatedestimated value for 119906Vwas arbitrary This in turn means that the value of 056 fortheirΦ ratio was likewise arbitrary
In Table 3 herein we show the results for column number1 in the 1972 study which we designate as Column A
1 Note
that the authorsrsquo raw values correspond to a value of 500 for1198750 a valuewhich is obviously far too high As is plainly evident
in our Figures 2 and 3 this column data set does not conformin any reasonable fashion to the norms in our referencecharts Accordingly the data is badly flawed In our correcteddata for Column A
1in Table 3 herein we adjust the column
external porosity to the lowest value permitted in our frame ofreference (038) This correction results in a revised value forGiddingsrsquoΦparameter of 0532 In additionwe are compelledto adjust the particle downward (to 11 micrometers approx)as is also dictated by our frame of reference We justify ourlow corrected value for external porosity and our downwardadjustment for particle size on the aggressive ldquodry-packingrdquoprocess described by the authors ldquo[T] the best results wereachieved with the following method the column (id =2mm) 25 cm long was packed with 25 equal portions of theldquobrushrdquo After adding each portion the column was vibratedand afterward tapped on the floor and also vigorously tappedwith a rod from the siderdquo Based upon a column packing
experience of the principal author herein spanningmore than30 years we believe that these packing conditions generateda packed column with a low external porosity and also one inwhich the particles experienced significant breakage
The main differences between the columns in the 1972and 1974 papers concerned the particle sizes and themethodsused to pack the columns The particle diameters in the 1972study fell in a range of about 15ndash200 micrometers while thosein the 1974 study were much smaller being in the range of 5ndash40 micrometers approximately Moreover while the columnsin the 1972 study were dry-packed the columns in the 1974study were slurry-packed (what Halasz calls ldquothe balanceddensity packing methodrdquo) The authors of the 1974 paperrecite that their permeability results shown in their Table 4correspond to values for 120601
0of 1080 and values for 120601 of 910
In Table 3 herein we show the results for the 1974 study as anaverage value in our column designated as A
2 Note that the
raw data reported by the authors which again has the built-inassumption that the column external porosity was 04 placesthe value of 119875
0at 192 In our corrected values for Column A
2
we show that a value of 267 for 1198750places the column external
porosity at a value of 043 The ratio of the value for 120601 of 910when compared to the value for 120601 in the 1972 study of 2007is about 22 (2007910 = 22) which is in good agreementwith the authorsrsquo statement that the permeability of the 1972columns was worse by a factor of about 2 Again based onthe experience of the principal author herein in the packingof hundreds of thousands of commercially sold slurry packedcolumns we conclude that the balanced density techniquedescribed by the authors in the 1974 paper is fully capable ofproducing columns with this high and external porosity
In summary we conclude that for purposes of theKozeny-Blake equation (a) the external porosity of theHalasz 1972 columns was 038 (b) the external porosity of thecolumns which he tested in 1974 was 043 and (c) Halaszrsquos1972 and 1974 studies when properly analyzed corroborateGiddingsrsquo value for 119875
0of 267 not Carmanrsquos value of 180
Although the fault for all of the misdirection about the valueof the permeability coefficient being 1 times 10minus3 therefore restsprimarily on the shoulders of Halasz one has to question whyGuiochon (andNeue) so readily adopted itWhatevermay bethe reason Guiochonrsquos support for this proposition has beenand continues to be a cause for much of the conventionalacceptance of Carmanrsquos (erroneous) figure of 180 as thecharacteristic value of the coefficient in the Kozeny-Blakeequation
9 Guiochonrsquos Experimental Data
In addition to the data from Halasz 1972 and 1974 studiesTable 3 herein contains data from a representative sampleof Guiochonrsquos personal experimental permeability investi-gations over approximately the last decade as reported inthe numerous trade journals particularly the Journal ofChromatography A In each case the table has a single rowdedicated to Guiochonrsquos reported raw data and a single rowdedicated to corrected values for each column based on thereference chart in Table 1 herein
12 Journal of MaterialsTa
ble3Summaryexperim
entald
ata
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Naefe
andHalasz
columns
1198601(19
72)
Colum
nA1rawdata
07140
0560
040
000314
0523
782007
2811
4016
563
498119864minus04
500
357
908119864minus10
648119864minus10
2500
020
100
000135
0001350
00038
100
053
132
074
Colum
nA1corrected
07140
0532
03800
0334
0539
7813
3518
705002
701
749119864minus04
267
191
908119864minus10
648119864minus10
2500
020
100
000110
000110
100038
100
053
139
074
Endelean
dHalasz
columns
1198602(19
74)
Colum
nA2rawdata
08426
0475
040
00044
30738
11910
1080
4740
563
110119864minus03
192
162
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
033
016
Colum
nA2corrected
08426
0511
04309
0412
0723
11910
1080
3411
405
110119864minus03
267
225
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
031
016
Guiochon
columns
1198612ndash1198614(19
98)
Colum
nB 2
rawdata
05639
0722
040
730157
0264
22771
1367
2931
520
130119864minus03
263
148
467119864minus10
263119864minus10
429
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 3
rawdata
05582
0704
03932
0165
0272
44764
1369
3382
606
131119864minus03
226
126
471119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 4
rawdata
05509
0710
03909
0160
0263
72784
1422
3422
621
128119864minus03
229
126
459119864minus10
253119864minus10
1328
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 3
corrected
05653
0723
040
860157
0265
44774
1369
2899
513
129119864minus03
267
151
465119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 4
corrected
05651
0723
040
830157
0265
72776
1374
2907
514
129119864minus03
267
151
464119864minus10
262119864minus10
1375
500
100
000
060
000
0600
00165
98008
020
015
Guiochon
columnC
(1999)
Colum
nCrawdata
05930
0673
03990
0194
0323
183
865
1459
3372
569
116119864minus03
257
152
109119864minus09
645119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Colum
nCcorrected
05930
0673
03990
0194
0323
190
900
1518
3372
569
111119864minus03
267
158
105119864minus09
620119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Guiochon
columns
1198631ndash1198636(2006)
Colum
nD
1rawdata
07830
0456
03570
0426
0663
86694
886
7115
909
144119864minus03
975
76334119864minus10
261119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
013
Colum
nD
2rawdata
07230
0491
03550
0368
0571
88656
907
6726
930
152119864minus03
975
70353119864minus10
255119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
014
Colum
nD
3rawdata
06850
0506
03466
0338
0518
97685
1000
7025
1026
146119864minus03
975
67338119864minus10
231119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD
4rawdata
06560
0521
03415
0315
0478
103
696
1062
7143
1089
144119864minus03
975
64332119864minus10
218119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD5rawdata
06190
0545
03371
0282
0425
109
692
1118
7100
1147
144119864minus03
975
60334119864minus10
207119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
016
Colum
nD
6rawdata
05760
0587
03383
0238
0359
107
635
1103
6516
1131
157119864minus03
975
56364119864minus10
210119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
017
Colum
nD
1corrected
07830
0542
04243
0359
0623
86907
1158
3397
434
110119864minus03
267
209
334119864minus10
261119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
013
Colum
nD
2corrected
07230
0584
04221
0301
0521
88857
1186
3212
444
117119864minus03
267
193
353119864minus10
255119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
014
Colum
nD
3corrected
06850
0603
04129
0272
0463
97896
1307
3354
490
112119864minus03
267
183
338119864minus10
231119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
015
Colum
nD
4corrected
06560
0621
040
730249
0420
103
911
1388
3411
520
110119864minus03
267
175
332119864minus10
218119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
015
Colum
nD
5corrected
06190
0650
04025
0217
0362
109
905
1462
3390
548
110119864minus03
267
165
334119864minus10
207119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
016
Colum
nD
6corrected
05760
0701
04038
0172
0289
107
831
1442
3111
540
120119864minus03
267
154
364119864minus10
210119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
017
Guiochon
columns
1198641-1198642
(2007)
Colum
nE 1
rawdata
064
500594
03830
0262
0425
356
1119
1735
4371
678
894119864minus04
256
165
804119864minus11
519119864minus11
1500
046
100
000
030
000
0300
000
41300
030
079
047
Colum
nE 2
rawdata
05400
0800
04320
0108
0190
470
1000
1853
2161
400
100119864minus03
463
250
729119864minus11
393119864minus11
1500
046
100
000
027
000
0270
000
41300
030
070
056
Colum
nE 1
corrected
064
500594
040
000245
040
8356
969
1502
3628
563
103119864minus03
267
172
804119864minus11
519119864minus11
1500
046
100
000
028
000
0279
000
41300
030
075
047
Colum
nE 2
corrected
05400
0800
04320
0108
0190
470
577
1068
2161
400
173119864minus03
267
144
729119864minus11
393119864minus11
1500
046
088
000
023
000
0205
000
41300
030
070
056
Guiochon
columns
1198651ndash1198654
(2008)
Colum
nF 1
rawdata
06350
0586
03720
0263
0419
197
725
1142
4865
766
138119864minus03
149
95399119864minus11
253119864minus11
300
021
100
000
017
000
0170
00035
100
048
129
076
Colum
nF 2
rawdata
064
200581
03730
0269
0429
361
807
1258
4863
758
124119864minus03
166
107
358119864minus11
230119864minus11
500
021
100
000
017
000
0170
00035
100
048
129
075
Colum
nF 3
rawdata
064
100588
03770
0264
0424
670
748
1166
4643
724
134119864minus03
161
103
387119864minus11
248119864minus11
1000
021
100
000
017
000
0170
00035
100
048
128
075
Colum
nF 4
rawdata
06390
0595
03800
0259
0418
1014
752
1177
4476
701
133119864minus03
168
107
384119864minus11
246119864minus11
1500
021
100
000
017
000
0170
00035
100
048
127
075
Colum
nF 1
corrected
06350
0630
040
000235
0392
197
954
1502
3572
563
105119864minus03
267
170
399119864minus11
253119864minus11
300
021
100
000
019
000
0195
00035
100
048
120
076
Colum
nF 2
corrected
064
200623
040
000242
0403
361
964
1502
3611
563
104119864minus03
267
171
358119864minus11
230119864minus11
500
021
100
000
019
000
0186
00035
100
048
120
075
Colum
nF 3
corrected
064
100624
040
000241
0402
670
963
1502
360
6563
104119864minus03
267
171
386119864minus11
248119864minus11
1000
021
100
000
019
000
0193
00035
100
048
120
075
Colum
nF 4
corrected
06390
0626
040
000239
0398
1014
960
1502
3594
563
104119864minus03
267
171
384119864minus11
246119864minus11
1500
021
100
000
019
000
0192
00035
100
048
120
075
Journal of Materials 13
Table3Con
tinued
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Guiochon
columns
1198671-1198672serie
s(2010)
Colum
nH
1rawdata
05320
0744
03960
0136
0225
120
563
1057
3125
587
178119864minus03
180
96122119864minus10
652119864minus11
1500
046
100
000
026
000
0262
00104
050
005
013
009
Colum
nH
2rawdata
05420
0686
03720
0170
0271
61747
1379
4152
766
134119864minus03
180
98823119864minus11
446119864minus11
1000
046
100
000
025
000
0248
00054
050
005
013
009
Colum
nH
1corrected
05320
0752
040
000132
0220
120
799
1502
2993
563
125119864minus03
267
142
122119864minus10
652119864minus11
1500
046
100
000
031
000
0313
00104
050
005
013
009
Colum
nH
2corrected
05420
0738
040
000142
0237
61814
1502
3049
563
123119864minus03
267
145
823119864minus11
446119864minus11
1000
046
100
000
026
000
0259
00054
050
005
013
009
Guiochon
columns
1198673-1198674serie
s(2010)
Colum
nH
3rawdata
06270
060
903820
0245
0396
463
700
1117
4296
685
143119864minus03
163
102
447119864minus11
280119864minus11
1500
021
100
000
018
000
0177
00036
050
024
063
038
Colum
nH
4rawdata
05170
0791
040
900108
0183
433
616
1191
2639
511
162119864minus03
233
121
580119864minus11
300119864minus11
1500
021
100
000
019
000
0189
00036
050
024
059
047
Colum
nH
3corrected
06270
0638
040
000227
0378
463
942
1502
3527
563
106119864minus03
267
167
447119864minus11
280119864minus11
1500
021
100
000
021
000
0205
00036
050
024
060
038
Colum
nH
4corrected
05170
0791
040
900108
0183
433
699
1353
2639
511
143119864minus03
265
137
580119864minus11
300119864minus11
1500
021
100
000
020
000
0201
00036
050
024
059
047
Guiochon
columns
1198681-1198686
(2010)
Colum
nI 1rawdata
06580
0562
03700
0288
0457
488
741
1127
5156
784
135119864minus03
144
95390119864minus11
256119864minus11
500
021
100
000
017
000
0170
00104
050
024
065
037
Colum
nI 2rawdata
06550
0576
03770
0278
044
6873
661
1009
4745
724
151119864minus03
139
91437119864minus11
286119864minus11
1000
021
100
000
017
000
0170
00104
050
024
064
037
Colum
nI 3rawdata
06550
0588
03850
0270
0439
1209
610
931
4341
663
164119864minus03
141
92474119864minus11
310119864minus11
1500
021
100
000
017
000
0170
00104
050
024
062
037
Colum
nI 4rawdata
06430
0572
03680
0275
0435
260
787
1225
5153
801
127119864minus03
153
98367119864minus11
236119864minus11
500
030
100
000
017
000
0170
00104
050
012
032
018
Colum
nI 5rawdata
06540
0583
03810
0273
044
144
3683
1044
4531
693
146119864minus03
151
99423119864minus11
277119864minus11
1000
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 6rawdata
06550
0585
03830
0272
044
164
6665
1016
4438
678
150119864minus03
150
98434119864minus11
285119864minus11
1500
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 1corrected
06580
060
8040
000258
0430
488
988
1502
3701
563
101119864minus03
267
176
390119864minus11
256119864minus11
500
021
100
000
020
000
0196
00104
050
024
060
037
Colum
nI 2corrected
06550
0611
040
000255
0425
873
984
1502
3684
563
102119864minus03
267
175
437119864minus11
286119864minus11
1000
021
100
000
021
000
0207
00104
050
024
060
037
Colum
nI 3corrected
06550
0611
040
000255
0425
1209
984
1502
3684
563
102119864minus03
267
175
474119864minus11
310119864minus11
1500
021
100
000
022
000
0216
00104
050
024
060
037
Colum
nI 4corrected
06430
0622
040
000243
040
5260
966
1502
3617
563
104119864minus03
267
172
367119864minus11
236119864minus11
500
030
100
000
019
000
0188
00104
050
012
029
018
Colum
nI 5corrected
06540
0612
040
000254
0423
443
982
1502
3679
563
102119864minus03
267
175
423119864minus11
277119864minus11
1000
030
100
000
020
000
0204
00104
050
012
029
018
Colum
nI 6corrected
06550
0611
040
000255
0425
646
984
1502
3684
563
102119864minus03
267
175
434119864minus11
285119864minus11
1500
030
100
000
021
000
0207
00104
050
012
029
018
Guiochon
columns
1198691-1198692
(2011)
Colum
nJ 1rawdata90
∘A(120578
=00054
P)04980
0803
040
000098
0163
65654
1313
2801
563
153119864minus03
234
116126119864minus10
627119864minus11
1500
046
100
000
029
000
0287
00054
050
005
013
010
Colum
nJ 2rawdata90
∘A(120578
=00104
P)04980
0803
040
000098
0163
124
651
1307
2801
563
154119864minus03
232
116127119864minus10
630119864minus11
1500
046
100
000
029
000
0287
00104
050
005
013
010
Colum
nJ 1rawdata160
∘A(120578
=00054
P)05630
0714
04020
0161
0269
71896
1592
3099
550
112119864minus03
289
163
102119864minus10
573119864minus11
1500
046
100
000
030
000
0302
00054
050
005
012
009
Colum
nJ 2rawdata160
∘A(120578
=00104
P)05630
0714
04020
0161
0269
129
847
1504
3099
550
118119864minus03
273
154
108119864minus10
606119864minus11
1500
046
100
000
030
000
0302
00104
050
005
012
009
Colum
nJ 1corrected
04980
0803
040
000098
0163
65748
1502
2801
563
134119864minus03
267
133
126119864minus10
627119864minus11
1500
046
100
000
031
000
0307
00054
050
005
013
010
Colum
nJ 2corrected
04980
0803
040
000098
0163
124
748
1502
2801
563
134119864minus03
267
133
127119864minus10
630119864minus11
1500
046
100
000
031
000
0308
00104
050
005
013
010
Colum
nJ 1corrected
05630
0714
04020
0161
0269
71827
1470
3099
550
121119864minus03
267
150
102119864minus10
573119864minus11
1500
046
100
000
029
000
0290
00054
050
005
012
009
Colum
nJ 2corrected
05630
0714
04020
0161
0269
129
827
1470
3099
550
121119864minus03
267
150
108119864minus10
606119864minus11
1500
046
100
000
030
000
0299
00104
050
005
012
009
14 Journal of Materials
050
100150200250300350400
045 050 055 060 065 070 075 080 085
LineAll corrected data
Linear (all corrected data)Squares raw data Triangles corrected data
Pt
120576t
D seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
J1 J2 H4
H1H2
D6
B2 B3 B4
J3 J4
E2
C
H3
D5 D4
E1
D2
A1
D1
A2
y = 267xLine slope = 267
Figure 2 Summary of experimental results
0
500
1000
1500
2000
2500
20 25 30 35 40 45 50 55 60 65 70 75 80
Linear (line)
Ψ
LineAll corrected dataD seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
Line slope (P0) = 267
J1 J2
J3 J4E2
E1
A1
A2
H4 H1
C
120601
H2H3I1 I2 I3 I4 I5
F1 F2 F3 F4
D1 D2 D3 D4 D5 D6
Squares raw data Triangles corrected data
Figure 3 Summary of experimental results
Figures 2 and 3 herein are graphical representations ofTable 31015840s raw and corrected data for Guiochonrsquos columnsplotted in the same reference frames as in Figures 1(a) and1(b) respectively
As shown in the plots of the thirty-one total columnsevaluated in this data base only six columns had reportedraw data values for 119875
0close to the normThese were columns
B2 C E
1 H
4 J
1and J
2 Interestingly of the six conforming
columns 3 of the columns were packed with fully porousparticles and 3 columns with partially porous particles Itcan be seen however that when properly interpreted andadjusted all of Guiochonrsquos experimental data falls fairlywithin the norms outlined in the reference chart in Table 1In order to explore the reasons why the raw data does notconform more closely to the norms outlined in Table 1 eachof Guiochonrsquos studies included in Table 3 is evaluated in turn
Finally Table 4 herein is a list of symbols and parameterdefinitions used throughout this paper as well as their unitsof measure
91 Guiochonrsquos 1998 Study In a 1998 publication by Kohand Guiochon entitled ldquoEffect of the Column Length on theCharacteristics of the Packed Bed and the Column Efficiencyin a Dynamic Axial Compression Columnrdquo [19] Guiochonand his coauthor report the results of a study they hadcarried out on column packing using a rather novel techniquewhich was a combination of slurry packing and mechanicalcompression
The study involved the packing of a single cylinder ofdiameter 5 cm to various column lengths ranging between 1and 25 cm approximately The particles were all silica basedand coated with a reverse stationary phase C
18 However two
different categories of particles were involved One categorycontained highly irregular particles which were about 130micrometers in average diameter In addition these particleshad a high specific pore volume (11mLg) and exhibitedsignificant particle breakage under the packing conditionsinvestigated in the study Conversely the other particlecategory involved very small particles (6 micrometers) whichwere spherical in shape had a low specific pore volume(03mLg) and exhibited a greater ability to withstandparticle breakage under the studyrsquos packing conditions
The permeability data for the irregular particles isneglected herein because according to the authors theparticles exhibited significant breakage during packing andwithout an accurate value for the characteristic dimensionof the particle the data cannot be used in the Kozeny-Blake model to determine the value of 119875
0 The spherical
particles on the other hand behaved successfully under thepacking conditions chosen and yielded values calculated bythe authors for 119875
0of 619 268 226 and 229 for the four
different lengths of columns studied The data for thesecolumns is reported in our Table 3 in the rows for ColumnsB1through B
4 The higher value of 619 (the shortest column)
was rejected by the authors because they felt that either theparticles had been somewhat broken during packing or thecolumn frits had been blocked with fines Accordingly thepermeability data for this column is likewise neglected herein
The remaining three columns however were not con-sistent with respect to the authorsrsquo calculated values for 119875
0
Since the particles were identical the range of values for 1198750
of 226 to 268 must be due to experimental error The rawdata for column B
2 which produced a 119875
0value of 268 is
all self-consistent For example it generates a value for 119875119905
of 148 which is a reasonable value based upon the knownlow particle porosity for these Nova Pak particles and acorresponding value for 120576
0of 04073 which is a reasonable
value for a well-packed column However the other twocolumns columns B
3and B
4 had the much lower value of
126 for 119875119905 Accordingly in Table 3 herein corrected values
for column external porosity for columns B3and B
4are
displayed These corrected values are only slightly differentthan the reported values are within the experimental errorof the measurement and therefore in combination with the
Journal of Materials 15
Table 4 Glossary of terms
Symbol Unit dimensions (cgs) Definition119860 cm2 Column cross-sectional area119863 cm Column diameter119889119901
cm Particle spherical diameter equivalentΔ119875 gcmminus1secminus2 Pressure drop along column119889119901119898
cm Average particle diameter (measured)1198960
None Guiochonrsquos residual reciprocal permeability coefficient119875119905
None Guiochonrsquos modified permeability coefficient in Kozeny-Blake equation119871 cm Column length1198750
None Permeability coefficient in Kozeny-Blake equation119876 cm3minminus1 (exception) Volumetric fluid flow rate1199050
sec Time for elution of totally permeating unretained solute119896 cm2 General permeability coefficient in Darcy equation119870
119865cm2 Halaszrsquos permeability coefficient in Darcy equation (1974 paper)
119870 cm2 Halaszrsquos permeability coefficient in Darcy equation (1972 paper)119870
119888None Guiochonrsquos permeability coefficient in the Kozeny-Blake equation (as modified by Carmen)
Greek1205760
None External column porosity120576119894
None Internal column porosity120576119905
None Total column porosity120576119901
None Particle porosityΦ None Giddingsrsquo ratio of external and total column porosity120601 None Flow resistance parameter based on mobile phase velocity1206010
None Flow resistance parameter based on superficial velocityΩ
119901None Particle sphericity factor
120578 gcmminus1secminus1 Fluid absolute viscosity120583119894
cmsecminus1 Average fluid interstitial linear velocity120583119904
cmsecminus1 Average fluid superficial linear velocity120583119905
cmsecminus1 Average mobile phase velocityΨ None Porosity dependence term in the Kozeny-Blake equation based on mobile phase velocityΨ
0None Porosity dependence term in the Kozeny-Blake equation based on superficial velocity
raw data for column B2 are supportive of the value of 267 for
1198750
92 Guiochonrsquos 1999 Study In a 1999 publication by Farkaset al entitled ldquoValidity of Darcyrsquos Law at Low Flow Rates inLiquid Chromatographyrdquo [20] Guiochon and his coauthorsreport the results of some very exactingmeasurements whichthey made to validate Darcyrsquos law at extremely low fluidflow rates Using a column containing particles packed toa measured external column porosity of 0399 and havinga measured total column porosity of 0593 (Figure 2 in thepaper) Guiochon et al measured the column pressure dropand fluid flow rate at flow rates ranging between 0015 and05mLminwith ethylene glycol as the fluidThe authors statethat the particles in the column are spherical and they appearto have gone to extraordinary lengths to obtain an accuratemeasure of the particle diameter and of course as a result ofmodern techniques of particle size classification the particlesize distribution is very narrow
In Figure 1 of the paper the authors show a plot ofcolumn measured pressure drop in bar versus ldquofluid linearvelocityrdquo in cmsec This plot represents their measurementstaken in the empty column The only velocity that exists in
an empty column is the superficial velocity and thus theabscissa values in the plot in Figure 1 designated as ldquofluidlinear velocityrdquo represent superficial linear velocity
In Figure 2 in their paper however the authors showanother plot of measured pressure drop in bar also versusldquofluid linear velocityrdquo in cmsec The velocity on the abscissain this plot although also designated as ldquofluid linear velocityrdquodoes not equate (for permeability related purposes) to theldquofluid linear velocityrdquo in Figure 1 This is because in Figure 2the columnwas filled with porous particles and themeasuredvelocity was the mobile phase velocity Indeed the onlyvelocity used throughout the entire paper is the mobile phasevelocity (mobile phase velocity and superficial velocity areone and the same in an empty column)
In Table 1 of their paper the authors report that for thecolumn represented in Figure 2 the slope of a line achievedwhen the mobile phase velocityin units of cmsec is plottedversus pressure drop in units of bar is 1829 They also reportthat the flow resistance parameter 120601 for this column is 865In Table 3 herein it can be seen that the raw data for thiscolumn (Column C) generates an apparent value of 257 for1198750 This value is very close to Giddingsrsquo value of 267 Indeed
by referring to Figures 2 and 3 herein one can see that thereported raw data in this study without any modification
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
10 Journal of Materials
fraction of a well-packed column Guiochon confirms in hisreview paper that this is indeed what he assumes in his state-ment of the Kozeny-Blake equation [14 p 9] Accordinglyhis adoption of the value of 180 for 119875
0solely depends on the
validity of his assertion that the value of the permeabilitycoefficient is 1 times 10minus3
We already know from our discussion of Guiochonrsquostextbook teaching that when this value is associated with apermeability equation based on mobile phase velocity it iswrong As reflected in our Table 2 when Guiochonrsquos workedexample involving this figure is corrected to account for theembedded factor due to the irregularity of the particles whichhis worked example assumes the value of the permeabilitycoefficient generated by his equation is not 1 times 10minus3 it is 166times 10minus3
So where did this value of 1 times 10minus3 come from AlthoughGuiochon in his review paper cites Neuersquos handbook for thisvalue Neue himself offers absolutely no authority for it Onthe other hand Guiochon does provide one other citation inhis review paper for this value The citation is to an article byHalasz et al published in 1975 [16]
If one then goes to that article one sees that Halaszet al do indeed proffer the same equation as Neue (albeitin a somewhat different format) and yes it does containthe value of 1 times 10minus3 See [16 Equation (1)] Neverthelessthe authors fail to identify from whence they obtained thisvalue On the other hand they do make reference to apaper that Endele et al wrote a year earlier in 1974 withKlaus Unger entitled ldquoInfluence of the Particle Size (5ndash35 120583)of Spherical Silica on Column Efficiencies in High-PressureLiquid Chromatographyrdquo [17] It appears that this is wherethe body is buried for it is in this article that Halasz reportshaving actually derived a value of 1 times 10minus3 for the permeabilitycoefficient
In summarizing the results of their experiments whichincidentally are based on measurements of superficial notmobile phase velocity (see their Equation (7)) Halasz et alcalculate an average value for the permeability coefficient(which they notate as119870
119865) of 1198892
1199011080 Accordingly they state
the following ldquo[I]t is proposed to define an average particlesize 119889
119901 for columns packed with spherical silica using the
balanced density packing method by the equation 1198892119901
=
1198701198651000 = 1198651205781198711031199032120587Δ119875rdquo [17 p 382 their Equation (7)]To recapitulate it thus appears that (1) Halasz was the
source of the value of 1 times 10minus3 for the permeability coefficientwhich Guiochon adopts in his review paper (2)Halaszrsquo valuewas based upon measurements which involved sphericalparticles and (3)Halaszrsquo value was properly derived from anequation which utilized the superficial velocity as the fluidvelocity
However we still have a problem if we are going to acceptthe value of 1 times 10minus3 as a basis for calculating the constantin Kozeny-Blake we still need to know the value of Halaszrsquocolumnsrsquo external porosity For example in the absenceof measured values for the external porosity Guiochonrsquosassertion in his review paper that the value of the constantis 180 is without foundation
Unfortunately however Halasz bases his methodologyon ldquoassumedrdquo versus ldquomeasuredrdquo values for external porosityNevertheless he does provide some clues that may helpus peel back this onion First he says that his value forthe permeability coefficient applies to ldquocolumns using thebalanced density packing methodrdquo Secondly after statinghis proposed value for the coefficient we find the followingtelltale sentence ldquothe permeability of columns packed by theconventional ldquodryrdquo method are worse by a factor of about 2than those packed by the balanced density methodrdquo [17 p382] For this proposition Halasz cites yet another of his ownarticles this time a paper written by himself and M Naefein 1972 entitled ldquoInfluence of Column Parameters on PeakBroadening in High-Pressure Liquid Chromatographyrdquo [18]
When one follows this thread by going to Halaszrsquos 1972paper one finds a possible solution to the puzzle For this iswhere Halasz describes his calculations of the permeabilitycoefficient from experiments with a number of conventionaldry-packed columns containing spherical silica particlesNot surprisingly the resulting value for the permeabilitycoefficient for such columns was not 1 times 10minus3 Instead Halaszsays that his experiments with such columns resulted inthe following value for the permeability coefficient ldquo119870 sim
11988921199012000rdquo [18 p 78] In other words consistent with what
Halasz says in his 1974 article the value for the coefficient forthese dry-packed columns was indeed ldquoworse by a factor ofabout 2rdquo compared to its value for columns prepared with thebalanced density method
The obvious implication of this is that when Halasz gota value of 1 times 10minus3 for the permeability coefficient in 1974he must have been evaluating columns that were packed toan interstitial fraction considerably higher than the externalporosity of the columns that were the subject of his 1972experiments Our problem however is still not resolvedbecause Halasz did not accurately determine the externalporosity of the columns under study in either the 1972 or 1974columns
We surmise that in their 1972 paper Halasz and hiscoauthors made use of the teaching of Giddings outlinedin his 1965 textbook which at the time of their 1972 paperwas seven years old Here is what they say ldquoexperiencehas shown that in a regular packed column with the sievefractions usual for chromatography 120576
0is independent of the
particle size if the particles are more or less spherical In anacceptable approximation 120576
0is equal to 04 In those columns
packed with nonporous supports 119906V and 119906 are identicalUsing porous supports (ie silica gel alumina chromosorbPorasil and so forth) 119906 = 4119865(1205871198892
119888120576119879) where 120576
119879is the
total porosity and indicates the pore volume of the supportrdquoIt will be appreciated that 119906V stands for interstitial velocity(our 120583
119894) and 119906 stands for mobile phase velocity (our 120583
119905) The
ldquoexperiencerdquo which Halasz and his coauthors are referringto concerning ldquocolumns packed with nonporous supportsalumina and chromosorbrdquo is that recorded by Giddings inTable 53-1 in his 1965 text Continuing to establish theinterrelationships inherent in the various entities involvedin chromatographic columns with respect to both interstitialandmobile phase velocity the authors conclude ldquoformally 119906V
Journal of Materials 11
can be calculated for a porous support assuming 1205760is equal
to 04rdquo (emphasis supplied) The authors then announce themethodology underlying their frame of reference with regardto column permeability as follows ldquowe determined in all ofour columns the 119906119906V ratio to be 056 with a differentialrefractometer Consequently the total porosity 120576
119905was equal
to 0714rdquo It will be appreciated that their ratio of 119906119906V is oneand the same as GiddingsrsquoΦ parameter as utilized by him inhis 1965 textbook and as we define hereinabove
We can now identify the origin of the discrepancybetween Halasz and Giddings and by extension betweenHalasz and Guiochon (and Neue) For even though Halaszmakes use of the results in Giddingsrsquo Table 53-1 he failsto appreciate the teaching inherent therein regarding theimportance of an accurate value for column external porositywhen using porous particles In establishing his values forΦGiddings was careful to ground his values for the externalporosity of columns packed with porous particles in acomparison of their measured pressure drops with measuredpressure drops in columns packed with nonporous particleswhere the values for (external) porositywere accurate becausethey are equal to the columnsrsquo total porosity
In their 1972 paper Halasz et al used a refractometer todetect the inert solute 119899-pentane in the mobile phase of 119899-heptaneThey injected the 119899-pentane into the flowing mobilephase the flow rate of which they presumably measured witha volumetric cylinder and stop watch Since the exit timeof the 119899-pentane peak was identified by the refractometertheir measurement of holdup volume was accurate andaccordingly so was their value of 0714 for 120576
119905(there is no
uncertainty related to column total porosity when using inertsolutes)This in turn led to an accurate value for their119906 termmobile phase velocity However since they did not measurethe interstitial velocity 119906V but rather used an estimate basedupon the measured flow rate and an ldquoassumedrdquo value of 04for external porosity their calculatedestimated value for 119906Vwas arbitrary This in turn means that the value of 056 fortheirΦ ratio was likewise arbitrary
In Table 3 herein we show the results for column number1 in the 1972 study which we designate as Column A
1 Note
that the authorsrsquo raw values correspond to a value of 500 for1198750 a valuewhich is obviously far too high As is plainly evident
in our Figures 2 and 3 this column data set does not conformin any reasonable fashion to the norms in our referencecharts Accordingly the data is badly flawed In our correcteddata for Column A
1in Table 3 herein we adjust the column
external porosity to the lowest value permitted in our frame ofreference (038) This correction results in a revised value forGiddingsrsquoΦparameter of 0532 In additionwe are compelledto adjust the particle downward (to 11 micrometers approx)as is also dictated by our frame of reference We justify ourlow corrected value for external porosity and our downwardadjustment for particle size on the aggressive ldquodry-packingrdquoprocess described by the authors ldquo[T] the best results wereachieved with the following method the column (id =2mm) 25 cm long was packed with 25 equal portions of theldquobrushrdquo After adding each portion the column was vibratedand afterward tapped on the floor and also vigorously tappedwith a rod from the siderdquo Based upon a column packing
experience of the principal author herein spanningmore than30 years we believe that these packing conditions generateda packed column with a low external porosity and also one inwhich the particles experienced significant breakage
The main differences between the columns in the 1972and 1974 papers concerned the particle sizes and themethodsused to pack the columns The particle diameters in the 1972study fell in a range of about 15ndash200 micrometers while thosein the 1974 study were much smaller being in the range of 5ndash40 micrometers approximately Moreover while the columnsin the 1972 study were dry-packed the columns in the 1974study were slurry-packed (what Halasz calls ldquothe balanceddensity packing methodrdquo) The authors of the 1974 paperrecite that their permeability results shown in their Table 4correspond to values for 120601
0of 1080 and values for 120601 of 910
In Table 3 herein we show the results for the 1974 study as anaverage value in our column designated as A
2 Note that the
raw data reported by the authors which again has the built-inassumption that the column external porosity was 04 placesthe value of 119875
0at 192 In our corrected values for Column A
2
we show that a value of 267 for 1198750places the column external
porosity at a value of 043 The ratio of the value for 120601 of 910when compared to the value for 120601 in the 1972 study of 2007is about 22 (2007910 = 22) which is in good agreementwith the authorsrsquo statement that the permeability of the 1972columns was worse by a factor of about 2 Again based onthe experience of the principal author herein in the packingof hundreds of thousands of commercially sold slurry packedcolumns we conclude that the balanced density techniquedescribed by the authors in the 1974 paper is fully capable ofproducing columns with this high and external porosity
In summary we conclude that for purposes of theKozeny-Blake equation (a) the external porosity of theHalasz 1972 columns was 038 (b) the external porosity of thecolumns which he tested in 1974 was 043 and (c) Halaszrsquos1972 and 1974 studies when properly analyzed corroborateGiddingsrsquo value for 119875
0of 267 not Carmanrsquos value of 180
Although the fault for all of the misdirection about the valueof the permeability coefficient being 1 times 10minus3 therefore restsprimarily on the shoulders of Halasz one has to question whyGuiochon (andNeue) so readily adopted itWhatevermay bethe reason Guiochonrsquos support for this proposition has beenand continues to be a cause for much of the conventionalacceptance of Carmanrsquos (erroneous) figure of 180 as thecharacteristic value of the coefficient in the Kozeny-Blakeequation
9 Guiochonrsquos Experimental Data
In addition to the data from Halasz 1972 and 1974 studiesTable 3 herein contains data from a representative sampleof Guiochonrsquos personal experimental permeability investi-gations over approximately the last decade as reported inthe numerous trade journals particularly the Journal ofChromatography A In each case the table has a single rowdedicated to Guiochonrsquos reported raw data and a single rowdedicated to corrected values for each column based on thereference chart in Table 1 herein
12 Journal of MaterialsTa
ble3Summaryexperim
entald
ata
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Naefe
andHalasz
columns
1198601(19
72)
Colum
nA1rawdata
07140
0560
040
000314
0523
782007
2811
4016
563
498119864minus04
500
357
908119864minus10
648119864minus10
2500
020
100
000135
0001350
00038
100
053
132
074
Colum
nA1corrected
07140
0532
03800
0334
0539
7813
3518
705002
701
749119864minus04
267
191
908119864minus10
648119864minus10
2500
020
100
000110
000110
100038
100
053
139
074
Endelean
dHalasz
columns
1198602(19
74)
Colum
nA2rawdata
08426
0475
040
00044
30738
11910
1080
4740
563
110119864minus03
192
162
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
033
016
Colum
nA2corrected
08426
0511
04309
0412
0723
11910
1080
3411
405
110119864minus03
267
225
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
031
016
Guiochon
columns
1198612ndash1198614(19
98)
Colum
nB 2
rawdata
05639
0722
040
730157
0264
22771
1367
2931
520
130119864minus03
263
148
467119864minus10
263119864minus10
429
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 3
rawdata
05582
0704
03932
0165
0272
44764
1369
3382
606
131119864minus03
226
126
471119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 4
rawdata
05509
0710
03909
0160
0263
72784
1422
3422
621
128119864minus03
229
126
459119864minus10
253119864minus10
1328
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 3
corrected
05653
0723
040
860157
0265
44774
1369
2899
513
129119864minus03
267
151
465119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 4
corrected
05651
0723
040
830157
0265
72776
1374
2907
514
129119864minus03
267
151
464119864minus10
262119864minus10
1375
500
100
000
060
000
0600
00165
98008
020
015
Guiochon
columnC
(1999)
Colum
nCrawdata
05930
0673
03990
0194
0323
183
865
1459
3372
569
116119864minus03
257
152
109119864minus09
645119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Colum
nCcorrected
05930
0673
03990
0194
0323
190
900
1518
3372
569
111119864minus03
267
158
105119864minus09
620119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Guiochon
columns
1198631ndash1198636(2006)
Colum
nD
1rawdata
07830
0456
03570
0426
0663
86694
886
7115
909
144119864minus03
975
76334119864minus10
261119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
013
Colum
nD
2rawdata
07230
0491
03550
0368
0571
88656
907
6726
930
152119864minus03
975
70353119864minus10
255119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
014
Colum
nD
3rawdata
06850
0506
03466
0338
0518
97685
1000
7025
1026
146119864minus03
975
67338119864minus10
231119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD
4rawdata
06560
0521
03415
0315
0478
103
696
1062
7143
1089
144119864minus03
975
64332119864minus10
218119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD5rawdata
06190
0545
03371
0282
0425
109
692
1118
7100
1147
144119864minus03
975
60334119864minus10
207119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
016
Colum
nD
6rawdata
05760
0587
03383
0238
0359
107
635
1103
6516
1131
157119864minus03
975
56364119864minus10
210119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
017
Colum
nD
1corrected
07830
0542
04243
0359
0623
86907
1158
3397
434
110119864minus03
267
209
334119864minus10
261119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
013
Colum
nD
2corrected
07230
0584
04221
0301
0521
88857
1186
3212
444
117119864minus03
267
193
353119864minus10
255119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
014
Colum
nD
3corrected
06850
0603
04129
0272
0463
97896
1307
3354
490
112119864minus03
267
183
338119864minus10
231119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
015
Colum
nD
4corrected
06560
0621
040
730249
0420
103
911
1388
3411
520
110119864minus03
267
175
332119864minus10
218119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
015
Colum
nD
5corrected
06190
0650
04025
0217
0362
109
905
1462
3390
548
110119864minus03
267
165
334119864minus10
207119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
016
Colum
nD
6corrected
05760
0701
04038
0172
0289
107
831
1442
3111
540
120119864minus03
267
154
364119864minus10
210119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
017
Guiochon
columns
1198641-1198642
(2007)
Colum
nE 1
rawdata
064
500594
03830
0262
0425
356
1119
1735
4371
678
894119864minus04
256
165
804119864minus11
519119864minus11
1500
046
100
000
030
000
0300
000
41300
030
079
047
Colum
nE 2
rawdata
05400
0800
04320
0108
0190
470
1000
1853
2161
400
100119864minus03
463
250
729119864minus11
393119864minus11
1500
046
100
000
027
000
0270
000
41300
030
070
056
Colum
nE 1
corrected
064
500594
040
000245
040
8356
969
1502
3628
563
103119864minus03
267
172
804119864minus11
519119864minus11
1500
046
100
000
028
000
0279
000
41300
030
075
047
Colum
nE 2
corrected
05400
0800
04320
0108
0190
470
577
1068
2161
400
173119864minus03
267
144
729119864minus11
393119864minus11
1500
046
088
000
023
000
0205
000
41300
030
070
056
Guiochon
columns
1198651ndash1198654
(2008)
Colum
nF 1
rawdata
06350
0586
03720
0263
0419
197
725
1142
4865
766
138119864minus03
149
95399119864minus11
253119864minus11
300
021
100
000
017
000
0170
00035
100
048
129
076
Colum
nF 2
rawdata
064
200581
03730
0269
0429
361
807
1258
4863
758
124119864minus03
166
107
358119864minus11
230119864minus11
500
021
100
000
017
000
0170
00035
100
048
129
075
Colum
nF 3
rawdata
064
100588
03770
0264
0424
670
748
1166
4643
724
134119864minus03
161
103
387119864minus11
248119864minus11
1000
021
100
000
017
000
0170
00035
100
048
128
075
Colum
nF 4
rawdata
06390
0595
03800
0259
0418
1014
752
1177
4476
701
133119864minus03
168
107
384119864minus11
246119864minus11
1500
021
100
000
017
000
0170
00035
100
048
127
075
Colum
nF 1
corrected
06350
0630
040
000235
0392
197
954
1502
3572
563
105119864minus03
267
170
399119864minus11
253119864minus11
300
021
100
000
019
000
0195
00035
100
048
120
076
Colum
nF 2
corrected
064
200623
040
000242
0403
361
964
1502
3611
563
104119864minus03
267
171
358119864minus11
230119864minus11
500
021
100
000
019
000
0186
00035
100
048
120
075
Colum
nF 3
corrected
064
100624
040
000241
0402
670
963
1502
360
6563
104119864minus03
267
171
386119864minus11
248119864minus11
1000
021
100
000
019
000
0193
00035
100
048
120
075
Colum
nF 4
corrected
06390
0626
040
000239
0398
1014
960
1502
3594
563
104119864minus03
267
171
384119864minus11
246119864minus11
1500
021
100
000
019
000
0192
00035
100
048
120
075
Journal of Materials 13
Table3Con
tinued
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Guiochon
columns
1198671-1198672serie
s(2010)
Colum
nH
1rawdata
05320
0744
03960
0136
0225
120
563
1057
3125
587
178119864minus03
180
96122119864minus10
652119864minus11
1500
046
100
000
026
000
0262
00104
050
005
013
009
Colum
nH
2rawdata
05420
0686
03720
0170
0271
61747
1379
4152
766
134119864minus03
180
98823119864minus11
446119864minus11
1000
046
100
000
025
000
0248
00054
050
005
013
009
Colum
nH
1corrected
05320
0752
040
000132
0220
120
799
1502
2993
563
125119864minus03
267
142
122119864minus10
652119864minus11
1500
046
100
000
031
000
0313
00104
050
005
013
009
Colum
nH
2corrected
05420
0738
040
000142
0237
61814
1502
3049
563
123119864minus03
267
145
823119864minus11
446119864minus11
1000
046
100
000
026
000
0259
00054
050
005
013
009
Guiochon
columns
1198673-1198674serie
s(2010)
Colum
nH
3rawdata
06270
060
903820
0245
0396
463
700
1117
4296
685
143119864minus03
163
102
447119864minus11
280119864minus11
1500
021
100
000
018
000
0177
00036
050
024
063
038
Colum
nH
4rawdata
05170
0791
040
900108
0183
433
616
1191
2639
511
162119864minus03
233
121
580119864minus11
300119864minus11
1500
021
100
000
019
000
0189
00036
050
024
059
047
Colum
nH
3corrected
06270
0638
040
000227
0378
463
942
1502
3527
563
106119864minus03
267
167
447119864minus11
280119864minus11
1500
021
100
000
021
000
0205
00036
050
024
060
038
Colum
nH
4corrected
05170
0791
040
900108
0183
433
699
1353
2639
511
143119864minus03
265
137
580119864minus11
300119864minus11
1500
021
100
000
020
000
0201
00036
050
024
059
047
Guiochon
columns
1198681-1198686
(2010)
Colum
nI 1rawdata
06580
0562
03700
0288
0457
488
741
1127
5156
784
135119864minus03
144
95390119864minus11
256119864minus11
500
021
100
000
017
000
0170
00104
050
024
065
037
Colum
nI 2rawdata
06550
0576
03770
0278
044
6873
661
1009
4745
724
151119864minus03
139
91437119864minus11
286119864minus11
1000
021
100
000
017
000
0170
00104
050
024
064
037
Colum
nI 3rawdata
06550
0588
03850
0270
0439
1209
610
931
4341
663
164119864minus03
141
92474119864minus11
310119864minus11
1500
021
100
000
017
000
0170
00104
050
024
062
037
Colum
nI 4rawdata
06430
0572
03680
0275
0435
260
787
1225
5153
801
127119864minus03
153
98367119864minus11
236119864minus11
500
030
100
000
017
000
0170
00104
050
012
032
018
Colum
nI 5rawdata
06540
0583
03810
0273
044
144
3683
1044
4531
693
146119864minus03
151
99423119864minus11
277119864minus11
1000
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 6rawdata
06550
0585
03830
0272
044
164
6665
1016
4438
678
150119864minus03
150
98434119864minus11
285119864minus11
1500
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 1corrected
06580
060
8040
000258
0430
488
988
1502
3701
563
101119864minus03
267
176
390119864minus11
256119864minus11
500
021
100
000
020
000
0196
00104
050
024
060
037
Colum
nI 2corrected
06550
0611
040
000255
0425
873
984
1502
3684
563
102119864minus03
267
175
437119864minus11
286119864minus11
1000
021
100
000
021
000
0207
00104
050
024
060
037
Colum
nI 3corrected
06550
0611
040
000255
0425
1209
984
1502
3684
563
102119864minus03
267
175
474119864minus11
310119864minus11
1500
021
100
000
022
000
0216
00104
050
024
060
037
Colum
nI 4corrected
06430
0622
040
000243
040
5260
966
1502
3617
563
104119864minus03
267
172
367119864minus11
236119864minus11
500
030
100
000
019
000
0188
00104
050
012
029
018
Colum
nI 5corrected
06540
0612
040
000254
0423
443
982
1502
3679
563
102119864minus03
267
175
423119864minus11
277119864minus11
1000
030
100
000
020
000
0204
00104
050
012
029
018
Colum
nI 6corrected
06550
0611
040
000255
0425
646
984
1502
3684
563
102119864minus03
267
175
434119864minus11
285119864minus11
1500
030
100
000
021
000
0207
00104
050
012
029
018
Guiochon
columns
1198691-1198692
(2011)
Colum
nJ 1rawdata90
∘A(120578
=00054
P)04980
0803
040
000098
0163
65654
1313
2801
563
153119864minus03
234
116126119864minus10
627119864minus11
1500
046
100
000
029
000
0287
00054
050
005
013
010
Colum
nJ 2rawdata90
∘A(120578
=00104
P)04980
0803
040
000098
0163
124
651
1307
2801
563
154119864minus03
232
116127119864minus10
630119864minus11
1500
046
100
000
029
000
0287
00104
050
005
013
010
Colum
nJ 1rawdata160
∘A(120578
=00054
P)05630
0714
04020
0161
0269
71896
1592
3099
550
112119864minus03
289
163
102119864minus10
573119864minus11
1500
046
100
000
030
000
0302
00054
050
005
012
009
Colum
nJ 2rawdata160
∘A(120578
=00104
P)05630
0714
04020
0161
0269
129
847
1504
3099
550
118119864minus03
273
154
108119864minus10
606119864minus11
1500
046
100
000
030
000
0302
00104
050
005
012
009
Colum
nJ 1corrected
04980
0803
040
000098
0163
65748
1502
2801
563
134119864minus03
267
133
126119864minus10
627119864minus11
1500
046
100
000
031
000
0307
00054
050
005
013
010
Colum
nJ 2corrected
04980
0803
040
000098
0163
124
748
1502
2801
563
134119864minus03
267
133
127119864minus10
630119864minus11
1500
046
100
000
031
000
0308
00104
050
005
013
010
Colum
nJ 1corrected
05630
0714
04020
0161
0269
71827
1470
3099
550
121119864minus03
267
150
102119864minus10
573119864minus11
1500
046
100
000
029
000
0290
00054
050
005
012
009
Colum
nJ 2corrected
05630
0714
04020
0161
0269
129
827
1470
3099
550
121119864minus03
267
150
108119864minus10
606119864minus11
1500
046
100
000
030
000
0299
00104
050
005
012
009
14 Journal of Materials
050
100150200250300350400
045 050 055 060 065 070 075 080 085
LineAll corrected data
Linear (all corrected data)Squares raw data Triangles corrected data
Pt
120576t
D seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
J1 J2 H4
H1H2
D6
B2 B3 B4
J3 J4
E2
C
H3
D5 D4
E1
D2
A1
D1
A2
y = 267xLine slope = 267
Figure 2 Summary of experimental results
0
500
1000
1500
2000
2500
20 25 30 35 40 45 50 55 60 65 70 75 80
Linear (line)
Ψ
LineAll corrected dataD seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
Line slope (P0) = 267
J1 J2
J3 J4E2
E1
A1
A2
H4 H1
C
120601
H2H3I1 I2 I3 I4 I5
F1 F2 F3 F4
D1 D2 D3 D4 D5 D6
Squares raw data Triangles corrected data
Figure 3 Summary of experimental results
Figures 2 and 3 herein are graphical representations ofTable 31015840s raw and corrected data for Guiochonrsquos columnsplotted in the same reference frames as in Figures 1(a) and1(b) respectively
As shown in the plots of the thirty-one total columnsevaluated in this data base only six columns had reportedraw data values for 119875
0close to the normThese were columns
B2 C E
1 H
4 J
1and J
2 Interestingly of the six conforming
columns 3 of the columns were packed with fully porousparticles and 3 columns with partially porous particles Itcan be seen however that when properly interpreted andadjusted all of Guiochonrsquos experimental data falls fairlywithin the norms outlined in the reference chart in Table 1In order to explore the reasons why the raw data does notconform more closely to the norms outlined in Table 1 eachof Guiochonrsquos studies included in Table 3 is evaluated in turn
Finally Table 4 herein is a list of symbols and parameterdefinitions used throughout this paper as well as their unitsof measure
91 Guiochonrsquos 1998 Study In a 1998 publication by Kohand Guiochon entitled ldquoEffect of the Column Length on theCharacteristics of the Packed Bed and the Column Efficiencyin a Dynamic Axial Compression Columnrdquo [19] Guiochonand his coauthor report the results of a study they hadcarried out on column packing using a rather novel techniquewhich was a combination of slurry packing and mechanicalcompression
The study involved the packing of a single cylinder ofdiameter 5 cm to various column lengths ranging between 1and 25 cm approximately The particles were all silica basedand coated with a reverse stationary phase C
18 However two
different categories of particles were involved One categorycontained highly irregular particles which were about 130micrometers in average diameter In addition these particleshad a high specific pore volume (11mLg) and exhibitedsignificant particle breakage under the packing conditionsinvestigated in the study Conversely the other particlecategory involved very small particles (6 micrometers) whichwere spherical in shape had a low specific pore volume(03mLg) and exhibited a greater ability to withstandparticle breakage under the studyrsquos packing conditions
The permeability data for the irregular particles isneglected herein because according to the authors theparticles exhibited significant breakage during packing andwithout an accurate value for the characteristic dimensionof the particle the data cannot be used in the Kozeny-Blake model to determine the value of 119875
0 The spherical
particles on the other hand behaved successfully under thepacking conditions chosen and yielded values calculated bythe authors for 119875
0of 619 268 226 and 229 for the four
different lengths of columns studied The data for thesecolumns is reported in our Table 3 in the rows for ColumnsB1through B
4 The higher value of 619 (the shortest column)
was rejected by the authors because they felt that either theparticles had been somewhat broken during packing or thecolumn frits had been blocked with fines Accordingly thepermeability data for this column is likewise neglected herein
The remaining three columns however were not con-sistent with respect to the authorsrsquo calculated values for 119875
0
Since the particles were identical the range of values for 1198750
of 226 to 268 must be due to experimental error The rawdata for column B
2 which produced a 119875
0value of 268 is
all self-consistent For example it generates a value for 119875119905
of 148 which is a reasonable value based upon the knownlow particle porosity for these Nova Pak particles and acorresponding value for 120576
0of 04073 which is a reasonable
value for a well-packed column However the other twocolumns columns B
3and B
4 had the much lower value of
126 for 119875119905 Accordingly in Table 3 herein corrected values
for column external porosity for columns B3and B
4are
displayed These corrected values are only slightly differentthan the reported values are within the experimental errorof the measurement and therefore in combination with the
Journal of Materials 15
Table 4 Glossary of terms
Symbol Unit dimensions (cgs) Definition119860 cm2 Column cross-sectional area119863 cm Column diameter119889119901
cm Particle spherical diameter equivalentΔ119875 gcmminus1secminus2 Pressure drop along column119889119901119898
cm Average particle diameter (measured)1198960
None Guiochonrsquos residual reciprocal permeability coefficient119875119905
None Guiochonrsquos modified permeability coefficient in Kozeny-Blake equation119871 cm Column length1198750
None Permeability coefficient in Kozeny-Blake equation119876 cm3minminus1 (exception) Volumetric fluid flow rate1199050
sec Time for elution of totally permeating unretained solute119896 cm2 General permeability coefficient in Darcy equation119870
119865cm2 Halaszrsquos permeability coefficient in Darcy equation (1974 paper)
119870 cm2 Halaszrsquos permeability coefficient in Darcy equation (1972 paper)119870
119888None Guiochonrsquos permeability coefficient in the Kozeny-Blake equation (as modified by Carmen)
Greek1205760
None External column porosity120576119894
None Internal column porosity120576119905
None Total column porosity120576119901
None Particle porosityΦ None Giddingsrsquo ratio of external and total column porosity120601 None Flow resistance parameter based on mobile phase velocity1206010
None Flow resistance parameter based on superficial velocityΩ
119901None Particle sphericity factor
120578 gcmminus1secminus1 Fluid absolute viscosity120583119894
cmsecminus1 Average fluid interstitial linear velocity120583119904
cmsecminus1 Average fluid superficial linear velocity120583119905
cmsecminus1 Average mobile phase velocityΨ None Porosity dependence term in the Kozeny-Blake equation based on mobile phase velocityΨ
0None Porosity dependence term in the Kozeny-Blake equation based on superficial velocity
raw data for column B2 are supportive of the value of 267 for
1198750
92 Guiochonrsquos 1999 Study In a 1999 publication by Farkaset al entitled ldquoValidity of Darcyrsquos Law at Low Flow Rates inLiquid Chromatographyrdquo [20] Guiochon and his coauthorsreport the results of some very exactingmeasurements whichthey made to validate Darcyrsquos law at extremely low fluidflow rates Using a column containing particles packed toa measured external column porosity of 0399 and havinga measured total column porosity of 0593 (Figure 2 in thepaper) Guiochon et al measured the column pressure dropand fluid flow rate at flow rates ranging between 0015 and05mLminwith ethylene glycol as the fluidThe authors statethat the particles in the column are spherical and they appearto have gone to extraordinary lengths to obtain an accuratemeasure of the particle diameter and of course as a result ofmodern techniques of particle size classification the particlesize distribution is very narrow
In Figure 1 of the paper the authors show a plot ofcolumn measured pressure drop in bar versus ldquofluid linearvelocityrdquo in cmsec This plot represents their measurementstaken in the empty column The only velocity that exists in
an empty column is the superficial velocity and thus theabscissa values in the plot in Figure 1 designated as ldquofluidlinear velocityrdquo represent superficial linear velocity
In Figure 2 in their paper however the authors showanother plot of measured pressure drop in bar also versusldquofluid linear velocityrdquo in cmsec The velocity on the abscissain this plot although also designated as ldquofluid linear velocityrdquodoes not equate (for permeability related purposes) to theldquofluid linear velocityrdquo in Figure 1 This is because in Figure 2the columnwas filled with porous particles and themeasuredvelocity was the mobile phase velocity Indeed the onlyvelocity used throughout the entire paper is the mobile phasevelocity (mobile phase velocity and superficial velocity areone and the same in an empty column)
In Table 1 of their paper the authors report that for thecolumn represented in Figure 2 the slope of a line achievedwhen the mobile phase velocityin units of cmsec is plottedversus pressure drop in units of bar is 1829 They also reportthat the flow resistance parameter 120601 for this column is 865In Table 3 herein it can be seen that the raw data for thiscolumn (Column C) generates an apparent value of 257 for1198750 This value is very close to Giddingsrsquo value of 267 Indeed
by referring to Figures 2 and 3 herein one can see that thereported raw data in this study without any modification
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
Journal of Materials 11
can be calculated for a porous support assuming 1205760is equal
to 04rdquo (emphasis supplied) The authors then announce themethodology underlying their frame of reference with regardto column permeability as follows ldquowe determined in all ofour columns the 119906119906V ratio to be 056 with a differentialrefractometer Consequently the total porosity 120576
119905was equal
to 0714rdquo It will be appreciated that their ratio of 119906119906V is oneand the same as GiddingsrsquoΦ parameter as utilized by him inhis 1965 textbook and as we define hereinabove
We can now identify the origin of the discrepancybetween Halasz and Giddings and by extension betweenHalasz and Guiochon (and Neue) For even though Halaszmakes use of the results in Giddingsrsquo Table 53-1 he failsto appreciate the teaching inherent therein regarding theimportance of an accurate value for column external porositywhen using porous particles In establishing his values forΦGiddings was careful to ground his values for the externalporosity of columns packed with porous particles in acomparison of their measured pressure drops with measuredpressure drops in columns packed with nonporous particleswhere the values for (external) porositywere accurate becausethey are equal to the columnsrsquo total porosity
In their 1972 paper Halasz et al used a refractometer todetect the inert solute 119899-pentane in the mobile phase of 119899-heptaneThey injected the 119899-pentane into the flowing mobilephase the flow rate of which they presumably measured witha volumetric cylinder and stop watch Since the exit timeof the 119899-pentane peak was identified by the refractometertheir measurement of holdup volume was accurate andaccordingly so was their value of 0714 for 120576
119905(there is no
uncertainty related to column total porosity when using inertsolutes)This in turn led to an accurate value for their119906 termmobile phase velocity However since they did not measurethe interstitial velocity 119906V but rather used an estimate basedupon the measured flow rate and an ldquoassumedrdquo value of 04for external porosity their calculatedestimated value for 119906Vwas arbitrary This in turn means that the value of 056 fortheirΦ ratio was likewise arbitrary
In Table 3 herein we show the results for column number1 in the 1972 study which we designate as Column A
1 Note
that the authorsrsquo raw values correspond to a value of 500 for1198750 a valuewhich is obviously far too high As is plainly evident
in our Figures 2 and 3 this column data set does not conformin any reasonable fashion to the norms in our referencecharts Accordingly the data is badly flawed In our correcteddata for Column A
1in Table 3 herein we adjust the column
external porosity to the lowest value permitted in our frame ofreference (038) This correction results in a revised value forGiddingsrsquoΦparameter of 0532 In additionwe are compelledto adjust the particle downward (to 11 micrometers approx)as is also dictated by our frame of reference We justify ourlow corrected value for external porosity and our downwardadjustment for particle size on the aggressive ldquodry-packingrdquoprocess described by the authors ldquo[T] the best results wereachieved with the following method the column (id =2mm) 25 cm long was packed with 25 equal portions of theldquobrushrdquo After adding each portion the column was vibratedand afterward tapped on the floor and also vigorously tappedwith a rod from the siderdquo Based upon a column packing
experience of the principal author herein spanningmore than30 years we believe that these packing conditions generateda packed column with a low external porosity and also one inwhich the particles experienced significant breakage
The main differences between the columns in the 1972and 1974 papers concerned the particle sizes and themethodsused to pack the columns The particle diameters in the 1972study fell in a range of about 15ndash200 micrometers while thosein the 1974 study were much smaller being in the range of 5ndash40 micrometers approximately Moreover while the columnsin the 1972 study were dry-packed the columns in the 1974study were slurry-packed (what Halasz calls ldquothe balanceddensity packing methodrdquo) The authors of the 1974 paperrecite that their permeability results shown in their Table 4correspond to values for 120601
0of 1080 and values for 120601 of 910
In Table 3 herein we show the results for the 1974 study as anaverage value in our column designated as A
2 Note that the
raw data reported by the authors which again has the built-inassumption that the column external porosity was 04 placesthe value of 119875
0at 192 In our corrected values for Column A
2
we show that a value of 267 for 1198750places the column external
porosity at a value of 043 The ratio of the value for 120601 of 910when compared to the value for 120601 in the 1972 study of 2007is about 22 (2007910 = 22) which is in good agreementwith the authorsrsquo statement that the permeability of the 1972columns was worse by a factor of about 2 Again based onthe experience of the principal author herein in the packingof hundreds of thousands of commercially sold slurry packedcolumns we conclude that the balanced density techniquedescribed by the authors in the 1974 paper is fully capable ofproducing columns with this high and external porosity
In summary we conclude that for purposes of theKozeny-Blake equation (a) the external porosity of theHalasz 1972 columns was 038 (b) the external porosity of thecolumns which he tested in 1974 was 043 and (c) Halaszrsquos1972 and 1974 studies when properly analyzed corroborateGiddingsrsquo value for 119875
0of 267 not Carmanrsquos value of 180
Although the fault for all of the misdirection about the valueof the permeability coefficient being 1 times 10minus3 therefore restsprimarily on the shoulders of Halasz one has to question whyGuiochon (andNeue) so readily adopted itWhatevermay bethe reason Guiochonrsquos support for this proposition has beenand continues to be a cause for much of the conventionalacceptance of Carmanrsquos (erroneous) figure of 180 as thecharacteristic value of the coefficient in the Kozeny-Blakeequation
9 Guiochonrsquos Experimental Data
In addition to the data from Halasz 1972 and 1974 studiesTable 3 herein contains data from a representative sampleof Guiochonrsquos personal experimental permeability investi-gations over approximately the last decade as reported inthe numerous trade journals particularly the Journal ofChromatography A In each case the table has a single rowdedicated to Guiochonrsquos reported raw data and a single rowdedicated to corrected values for each column based on thereference chart in Table 1 herein
12 Journal of MaterialsTa
ble3Summaryexperim
entald
ata
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Naefe
andHalasz
columns
1198601(19
72)
Colum
nA1rawdata
07140
0560
040
000314
0523
782007
2811
4016
563
498119864minus04
500
357
908119864minus10
648119864minus10
2500
020
100
000135
0001350
00038
100
053
132
074
Colum
nA1corrected
07140
0532
03800
0334
0539
7813
3518
705002
701
749119864minus04
267
191
908119864minus10
648119864minus10
2500
020
100
000110
000110
100038
100
053
139
074
Endelean
dHalasz
columns
1198602(19
74)
Colum
nA2rawdata
08426
0475
040
00044
30738
11910
1080
4740
563
110119864minus03
192
162
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
033
016
Colum
nA2corrected
08426
0511
04309
0412
0723
11910
1080
3411
405
110119864minus03
267
225
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
031
016
Guiochon
columns
1198612ndash1198614(19
98)
Colum
nB 2
rawdata
05639
0722
040
730157
0264
22771
1367
2931
520
130119864minus03
263
148
467119864minus10
263119864minus10
429
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 3
rawdata
05582
0704
03932
0165
0272
44764
1369
3382
606
131119864minus03
226
126
471119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 4
rawdata
05509
0710
03909
0160
0263
72784
1422
3422
621
128119864minus03
229
126
459119864minus10
253119864minus10
1328
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 3
corrected
05653
0723
040
860157
0265
44774
1369
2899
513
129119864minus03
267
151
465119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 4
corrected
05651
0723
040
830157
0265
72776
1374
2907
514
129119864minus03
267
151
464119864minus10
262119864minus10
1375
500
100
000
060
000
0600
00165
98008
020
015
Guiochon
columnC
(1999)
Colum
nCrawdata
05930
0673
03990
0194
0323
183
865
1459
3372
569
116119864minus03
257
152
109119864minus09
645119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Colum
nCcorrected
05930
0673
03990
0194
0323
190
900
1518
3372
569
111119864minus03
267
158
105119864minus09
620119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Guiochon
columns
1198631ndash1198636(2006)
Colum
nD
1rawdata
07830
0456
03570
0426
0663
86694
886
7115
909
144119864minus03
975
76334119864minus10
261119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
013
Colum
nD
2rawdata
07230
0491
03550
0368
0571
88656
907
6726
930
152119864minus03
975
70353119864minus10
255119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
014
Colum
nD
3rawdata
06850
0506
03466
0338
0518
97685
1000
7025
1026
146119864minus03
975
67338119864minus10
231119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD
4rawdata
06560
0521
03415
0315
0478
103
696
1062
7143
1089
144119864minus03
975
64332119864minus10
218119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD5rawdata
06190
0545
03371
0282
0425
109
692
1118
7100
1147
144119864minus03
975
60334119864minus10
207119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
016
Colum
nD
6rawdata
05760
0587
03383
0238
0359
107
635
1103
6516
1131
157119864minus03
975
56364119864minus10
210119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
017
Colum
nD
1corrected
07830
0542
04243
0359
0623
86907
1158
3397
434
110119864minus03
267
209
334119864minus10
261119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
013
Colum
nD
2corrected
07230
0584
04221
0301
0521
88857
1186
3212
444
117119864minus03
267
193
353119864minus10
255119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
014
Colum
nD
3corrected
06850
0603
04129
0272
0463
97896
1307
3354
490
112119864minus03
267
183
338119864minus10
231119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
015
Colum
nD
4corrected
06560
0621
040
730249
0420
103
911
1388
3411
520
110119864minus03
267
175
332119864minus10
218119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
015
Colum
nD
5corrected
06190
0650
04025
0217
0362
109
905
1462
3390
548
110119864minus03
267
165
334119864minus10
207119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
016
Colum
nD
6corrected
05760
0701
04038
0172
0289
107
831
1442
3111
540
120119864minus03
267
154
364119864minus10
210119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
017
Guiochon
columns
1198641-1198642
(2007)
Colum
nE 1
rawdata
064
500594
03830
0262
0425
356
1119
1735
4371
678
894119864minus04
256
165
804119864minus11
519119864minus11
1500
046
100
000
030
000
0300
000
41300
030
079
047
Colum
nE 2
rawdata
05400
0800
04320
0108
0190
470
1000
1853
2161
400
100119864minus03
463
250
729119864minus11
393119864minus11
1500
046
100
000
027
000
0270
000
41300
030
070
056
Colum
nE 1
corrected
064
500594
040
000245
040
8356
969
1502
3628
563
103119864minus03
267
172
804119864minus11
519119864minus11
1500
046
100
000
028
000
0279
000
41300
030
075
047
Colum
nE 2
corrected
05400
0800
04320
0108
0190
470
577
1068
2161
400
173119864minus03
267
144
729119864minus11
393119864minus11
1500
046
088
000
023
000
0205
000
41300
030
070
056
Guiochon
columns
1198651ndash1198654
(2008)
Colum
nF 1
rawdata
06350
0586
03720
0263
0419
197
725
1142
4865
766
138119864minus03
149
95399119864minus11
253119864minus11
300
021
100
000
017
000
0170
00035
100
048
129
076
Colum
nF 2
rawdata
064
200581
03730
0269
0429
361
807
1258
4863
758
124119864minus03
166
107
358119864minus11
230119864minus11
500
021
100
000
017
000
0170
00035
100
048
129
075
Colum
nF 3
rawdata
064
100588
03770
0264
0424
670
748
1166
4643
724
134119864minus03
161
103
387119864minus11
248119864minus11
1000
021
100
000
017
000
0170
00035
100
048
128
075
Colum
nF 4
rawdata
06390
0595
03800
0259
0418
1014
752
1177
4476
701
133119864minus03
168
107
384119864minus11
246119864minus11
1500
021
100
000
017
000
0170
00035
100
048
127
075
Colum
nF 1
corrected
06350
0630
040
000235
0392
197
954
1502
3572
563
105119864minus03
267
170
399119864minus11
253119864minus11
300
021
100
000
019
000
0195
00035
100
048
120
076
Colum
nF 2
corrected
064
200623
040
000242
0403
361
964
1502
3611
563
104119864minus03
267
171
358119864minus11
230119864minus11
500
021
100
000
019
000
0186
00035
100
048
120
075
Colum
nF 3
corrected
064
100624
040
000241
0402
670
963
1502
360
6563
104119864minus03
267
171
386119864minus11
248119864minus11
1000
021
100
000
019
000
0193
00035
100
048
120
075
Colum
nF 4
corrected
06390
0626
040
000239
0398
1014
960
1502
3594
563
104119864minus03
267
171
384119864minus11
246119864minus11
1500
021
100
000
019
000
0192
00035
100
048
120
075
Journal of Materials 13
Table3Con
tinued
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Guiochon
columns
1198671-1198672serie
s(2010)
Colum
nH
1rawdata
05320
0744
03960
0136
0225
120
563
1057
3125
587
178119864minus03
180
96122119864minus10
652119864minus11
1500
046
100
000
026
000
0262
00104
050
005
013
009
Colum
nH
2rawdata
05420
0686
03720
0170
0271
61747
1379
4152
766
134119864minus03
180
98823119864minus11
446119864minus11
1000
046
100
000
025
000
0248
00054
050
005
013
009
Colum
nH
1corrected
05320
0752
040
000132
0220
120
799
1502
2993
563
125119864minus03
267
142
122119864minus10
652119864minus11
1500
046
100
000
031
000
0313
00104
050
005
013
009
Colum
nH
2corrected
05420
0738
040
000142
0237
61814
1502
3049
563
123119864minus03
267
145
823119864minus11
446119864minus11
1000
046
100
000
026
000
0259
00054
050
005
013
009
Guiochon
columns
1198673-1198674serie
s(2010)
Colum
nH
3rawdata
06270
060
903820
0245
0396
463
700
1117
4296
685
143119864minus03
163
102
447119864minus11
280119864minus11
1500
021
100
000
018
000
0177
00036
050
024
063
038
Colum
nH
4rawdata
05170
0791
040
900108
0183
433
616
1191
2639
511
162119864minus03
233
121
580119864minus11
300119864minus11
1500
021
100
000
019
000
0189
00036
050
024
059
047
Colum
nH
3corrected
06270
0638
040
000227
0378
463
942
1502
3527
563
106119864minus03
267
167
447119864minus11
280119864minus11
1500
021
100
000
021
000
0205
00036
050
024
060
038
Colum
nH
4corrected
05170
0791
040
900108
0183
433
699
1353
2639
511
143119864minus03
265
137
580119864minus11
300119864minus11
1500
021
100
000
020
000
0201
00036
050
024
059
047
Guiochon
columns
1198681-1198686
(2010)
Colum
nI 1rawdata
06580
0562
03700
0288
0457
488
741
1127
5156
784
135119864minus03
144
95390119864minus11
256119864minus11
500
021
100
000
017
000
0170
00104
050
024
065
037
Colum
nI 2rawdata
06550
0576
03770
0278
044
6873
661
1009
4745
724
151119864minus03
139
91437119864minus11
286119864minus11
1000
021
100
000
017
000
0170
00104
050
024
064
037
Colum
nI 3rawdata
06550
0588
03850
0270
0439
1209
610
931
4341
663
164119864minus03
141
92474119864minus11
310119864minus11
1500
021
100
000
017
000
0170
00104
050
024
062
037
Colum
nI 4rawdata
06430
0572
03680
0275
0435
260
787
1225
5153
801
127119864minus03
153
98367119864minus11
236119864minus11
500
030
100
000
017
000
0170
00104
050
012
032
018
Colum
nI 5rawdata
06540
0583
03810
0273
044
144
3683
1044
4531
693
146119864minus03
151
99423119864minus11
277119864minus11
1000
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 6rawdata
06550
0585
03830
0272
044
164
6665
1016
4438
678
150119864minus03
150
98434119864minus11
285119864minus11
1500
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 1corrected
06580
060
8040
000258
0430
488
988
1502
3701
563
101119864minus03
267
176
390119864minus11
256119864minus11
500
021
100
000
020
000
0196
00104
050
024
060
037
Colum
nI 2corrected
06550
0611
040
000255
0425
873
984
1502
3684
563
102119864minus03
267
175
437119864minus11
286119864minus11
1000
021
100
000
021
000
0207
00104
050
024
060
037
Colum
nI 3corrected
06550
0611
040
000255
0425
1209
984
1502
3684
563
102119864minus03
267
175
474119864minus11
310119864minus11
1500
021
100
000
022
000
0216
00104
050
024
060
037
Colum
nI 4corrected
06430
0622
040
000243
040
5260
966
1502
3617
563
104119864minus03
267
172
367119864minus11
236119864minus11
500
030
100
000
019
000
0188
00104
050
012
029
018
Colum
nI 5corrected
06540
0612
040
000254
0423
443
982
1502
3679
563
102119864minus03
267
175
423119864minus11
277119864minus11
1000
030
100
000
020
000
0204
00104
050
012
029
018
Colum
nI 6corrected
06550
0611
040
000255
0425
646
984
1502
3684
563
102119864minus03
267
175
434119864minus11
285119864minus11
1500
030
100
000
021
000
0207
00104
050
012
029
018
Guiochon
columns
1198691-1198692
(2011)
Colum
nJ 1rawdata90
∘A(120578
=00054
P)04980
0803
040
000098
0163
65654
1313
2801
563
153119864minus03
234
116126119864minus10
627119864minus11
1500
046
100
000
029
000
0287
00054
050
005
013
010
Colum
nJ 2rawdata90
∘A(120578
=00104
P)04980
0803
040
000098
0163
124
651
1307
2801
563
154119864minus03
232
116127119864minus10
630119864minus11
1500
046
100
000
029
000
0287
00104
050
005
013
010
Colum
nJ 1rawdata160
∘A(120578
=00054
P)05630
0714
04020
0161
0269
71896
1592
3099
550
112119864minus03
289
163
102119864minus10
573119864minus11
1500
046
100
000
030
000
0302
00054
050
005
012
009
Colum
nJ 2rawdata160
∘A(120578
=00104
P)05630
0714
04020
0161
0269
129
847
1504
3099
550
118119864minus03
273
154
108119864minus10
606119864minus11
1500
046
100
000
030
000
0302
00104
050
005
012
009
Colum
nJ 1corrected
04980
0803
040
000098
0163
65748
1502
2801
563
134119864minus03
267
133
126119864minus10
627119864minus11
1500
046
100
000
031
000
0307
00054
050
005
013
010
Colum
nJ 2corrected
04980
0803
040
000098
0163
124
748
1502
2801
563
134119864minus03
267
133
127119864minus10
630119864minus11
1500
046
100
000
031
000
0308
00104
050
005
013
010
Colum
nJ 1corrected
05630
0714
04020
0161
0269
71827
1470
3099
550
121119864minus03
267
150
102119864minus10
573119864minus11
1500
046
100
000
029
000
0290
00054
050
005
012
009
Colum
nJ 2corrected
05630
0714
04020
0161
0269
129
827
1470
3099
550
121119864minus03
267
150
108119864minus10
606119864minus11
1500
046
100
000
030
000
0299
00104
050
005
012
009
14 Journal of Materials
050
100150200250300350400
045 050 055 060 065 070 075 080 085
LineAll corrected data
Linear (all corrected data)Squares raw data Triangles corrected data
Pt
120576t
D seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
J1 J2 H4
H1H2
D6
B2 B3 B4
J3 J4
E2
C
H3
D5 D4
E1
D2
A1
D1
A2
y = 267xLine slope = 267
Figure 2 Summary of experimental results
0
500
1000
1500
2000
2500
20 25 30 35 40 45 50 55 60 65 70 75 80
Linear (line)
Ψ
LineAll corrected dataD seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
Line slope (P0) = 267
J1 J2
J3 J4E2
E1
A1
A2
H4 H1
C
120601
H2H3I1 I2 I3 I4 I5
F1 F2 F3 F4
D1 D2 D3 D4 D5 D6
Squares raw data Triangles corrected data
Figure 3 Summary of experimental results
Figures 2 and 3 herein are graphical representations ofTable 31015840s raw and corrected data for Guiochonrsquos columnsplotted in the same reference frames as in Figures 1(a) and1(b) respectively
As shown in the plots of the thirty-one total columnsevaluated in this data base only six columns had reportedraw data values for 119875
0close to the normThese were columns
B2 C E
1 H
4 J
1and J
2 Interestingly of the six conforming
columns 3 of the columns were packed with fully porousparticles and 3 columns with partially porous particles Itcan be seen however that when properly interpreted andadjusted all of Guiochonrsquos experimental data falls fairlywithin the norms outlined in the reference chart in Table 1In order to explore the reasons why the raw data does notconform more closely to the norms outlined in Table 1 eachof Guiochonrsquos studies included in Table 3 is evaluated in turn
Finally Table 4 herein is a list of symbols and parameterdefinitions used throughout this paper as well as their unitsof measure
91 Guiochonrsquos 1998 Study In a 1998 publication by Kohand Guiochon entitled ldquoEffect of the Column Length on theCharacteristics of the Packed Bed and the Column Efficiencyin a Dynamic Axial Compression Columnrdquo [19] Guiochonand his coauthor report the results of a study they hadcarried out on column packing using a rather novel techniquewhich was a combination of slurry packing and mechanicalcompression
The study involved the packing of a single cylinder ofdiameter 5 cm to various column lengths ranging between 1and 25 cm approximately The particles were all silica basedand coated with a reverse stationary phase C
18 However two
different categories of particles were involved One categorycontained highly irregular particles which were about 130micrometers in average diameter In addition these particleshad a high specific pore volume (11mLg) and exhibitedsignificant particle breakage under the packing conditionsinvestigated in the study Conversely the other particlecategory involved very small particles (6 micrometers) whichwere spherical in shape had a low specific pore volume(03mLg) and exhibited a greater ability to withstandparticle breakage under the studyrsquos packing conditions
The permeability data for the irregular particles isneglected herein because according to the authors theparticles exhibited significant breakage during packing andwithout an accurate value for the characteristic dimensionof the particle the data cannot be used in the Kozeny-Blake model to determine the value of 119875
0 The spherical
particles on the other hand behaved successfully under thepacking conditions chosen and yielded values calculated bythe authors for 119875
0of 619 268 226 and 229 for the four
different lengths of columns studied The data for thesecolumns is reported in our Table 3 in the rows for ColumnsB1through B
4 The higher value of 619 (the shortest column)
was rejected by the authors because they felt that either theparticles had been somewhat broken during packing or thecolumn frits had been blocked with fines Accordingly thepermeability data for this column is likewise neglected herein
The remaining three columns however were not con-sistent with respect to the authorsrsquo calculated values for 119875
0
Since the particles were identical the range of values for 1198750
of 226 to 268 must be due to experimental error The rawdata for column B
2 which produced a 119875
0value of 268 is
all self-consistent For example it generates a value for 119875119905
of 148 which is a reasonable value based upon the knownlow particle porosity for these Nova Pak particles and acorresponding value for 120576
0of 04073 which is a reasonable
value for a well-packed column However the other twocolumns columns B
3and B
4 had the much lower value of
126 for 119875119905 Accordingly in Table 3 herein corrected values
for column external porosity for columns B3and B
4are
displayed These corrected values are only slightly differentthan the reported values are within the experimental errorof the measurement and therefore in combination with the
Journal of Materials 15
Table 4 Glossary of terms
Symbol Unit dimensions (cgs) Definition119860 cm2 Column cross-sectional area119863 cm Column diameter119889119901
cm Particle spherical diameter equivalentΔ119875 gcmminus1secminus2 Pressure drop along column119889119901119898
cm Average particle diameter (measured)1198960
None Guiochonrsquos residual reciprocal permeability coefficient119875119905
None Guiochonrsquos modified permeability coefficient in Kozeny-Blake equation119871 cm Column length1198750
None Permeability coefficient in Kozeny-Blake equation119876 cm3minminus1 (exception) Volumetric fluid flow rate1199050
sec Time for elution of totally permeating unretained solute119896 cm2 General permeability coefficient in Darcy equation119870
119865cm2 Halaszrsquos permeability coefficient in Darcy equation (1974 paper)
119870 cm2 Halaszrsquos permeability coefficient in Darcy equation (1972 paper)119870
119888None Guiochonrsquos permeability coefficient in the Kozeny-Blake equation (as modified by Carmen)
Greek1205760
None External column porosity120576119894
None Internal column porosity120576119905
None Total column porosity120576119901
None Particle porosityΦ None Giddingsrsquo ratio of external and total column porosity120601 None Flow resistance parameter based on mobile phase velocity1206010
None Flow resistance parameter based on superficial velocityΩ
119901None Particle sphericity factor
120578 gcmminus1secminus1 Fluid absolute viscosity120583119894
cmsecminus1 Average fluid interstitial linear velocity120583119904
cmsecminus1 Average fluid superficial linear velocity120583119905
cmsecminus1 Average mobile phase velocityΨ None Porosity dependence term in the Kozeny-Blake equation based on mobile phase velocityΨ
0None Porosity dependence term in the Kozeny-Blake equation based on superficial velocity
raw data for column B2 are supportive of the value of 267 for
1198750
92 Guiochonrsquos 1999 Study In a 1999 publication by Farkaset al entitled ldquoValidity of Darcyrsquos Law at Low Flow Rates inLiquid Chromatographyrdquo [20] Guiochon and his coauthorsreport the results of some very exactingmeasurements whichthey made to validate Darcyrsquos law at extremely low fluidflow rates Using a column containing particles packed toa measured external column porosity of 0399 and havinga measured total column porosity of 0593 (Figure 2 in thepaper) Guiochon et al measured the column pressure dropand fluid flow rate at flow rates ranging between 0015 and05mLminwith ethylene glycol as the fluidThe authors statethat the particles in the column are spherical and they appearto have gone to extraordinary lengths to obtain an accuratemeasure of the particle diameter and of course as a result ofmodern techniques of particle size classification the particlesize distribution is very narrow
In Figure 1 of the paper the authors show a plot ofcolumn measured pressure drop in bar versus ldquofluid linearvelocityrdquo in cmsec This plot represents their measurementstaken in the empty column The only velocity that exists in
an empty column is the superficial velocity and thus theabscissa values in the plot in Figure 1 designated as ldquofluidlinear velocityrdquo represent superficial linear velocity
In Figure 2 in their paper however the authors showanother plot of measured pressure drop in bar also versusldquofluid linear velocityrdquo in cmsec The velocity on the abscissain this plot although also designated as ldquofluid linear velocityrdquodoes not equate (for permeability related purposes) to theldquofluid linear velocityrdquo in Figure 1 This is because in Figure 2the columnwas filled with porous particles and themeasuredvelocity was the mobile phase velocity Indeed the onlyvelocity used throughout the entire paper is the mobile phasevelocity (mobile phase velocity and superficial velocity areone and the same in an empty column)
In Table 1 of their paper the authors report that for thecolumn represented in Figure 2 the slope of a line achievedwhen the mobile phase velocityin units of cmsec is plottedversus pressure drop in units of bar is 1829 They also reportthat the flow resistance parameter 120601 for this column is 865In Table 3 herein it can be seen that the raw data for thiscolumn (Column C) generates an apparent value of 257 for1198750 This value is very close to Giddingsrsquo value of 267 Indeed
by referring to Figures 2 and 3 herein one can see that thereported raw data in this study without any modification
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
12 Journal of MaterialsTa
ble3Summaryexperim
entald
ata
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Naefe
andHalasz
columns
1198601(19
72)
Colum
nA1rawdata
07140
0560
040
000314
0523
782007
2811
4016
563
498119864minus04
500
357
908119864minus10
648119864minus10
2500
020
100
000135
0001350
00038
100
053
132
074
Colum
nA1corrected
07140
0532
03800
0334
0539
7813
3518
705002
701
749119864minus04
267
191
908119864minus10
648119864minus10
2500
020
100
000110
000110
100038
100
053
139
074
Endelean
dHalasz
columns
1198602(19
74)
Colum
nA2rawdata
08426
0475
040
00044
30738
11910
1080
4740
563
110119864minus03
192
162
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
033
016
Colum
nA2corrected
08426
0511
04309
0412
0723
11910
1080
3411
405
110119864minus03
267
225
435119864minus10
366119864minus10
3000
040
100
000
063
000
0629
00010
100
013
031
016
Guiochon
columns
1198612ndash1198614(19
98)
Colum
nB 2
rawdata
05639
0722
040
730157
0264
22771
1367
2931
520
130119864minus03
263
148
467119864minus10
263119864minus10
429
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 3
rawdata
05582
0704
03932
0165
0272
44764
1369
3382
606
131119864minus03
226
126
471119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 4
rawdata
05509
0710
03909
0160
0263
72784
1422
3422
621
128119864minus03
229
126
459119864minus10
253119864minus10
1328
500
100
000
060
000
0600
00165
98008
021
015
Colum
nB 3
corrected
05653
0723
040
860157
0265
44774
1369
2899
513
129119864minus03
267
151
465119864minus10
263119864minus10
838
500
100
000
060
000
0600
00165
98008
020
015
Colum
nB 4
corrected
05651
0723
040
830157
0265
72776
1374
2907
514
129119864minus03
267
151
464119864minus10
262119864minus10
1375
500
100
000
060
000
0600
00165
98008
020
015
Guiochon
columnC
(1999)
Colum
nCrawdata
05930
0673
03990
0194
0323
183
865
1459
3372
569
116119864minus03
257
152
109119864minus09
645119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Colum
nCcorrected
05930
0673
03990
0194
0323
190
900
1518
3372
569
111119864minus03
267
158
105119864minus09
620119864minus10
1000
046
100
000
097
000
0970
01990
059
006
015
010
Guiochon
columns
1198631ndash1198636(2006)
Colum
nD
1rawdata
07830
0456
03570
0426
0663
86694
886
7115
909
144119864minus03
975
76334119864minus10
261119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
013
Colum
nD
2rawdata
07230
0491
03550
0368
0571
88656
907
6726
930
152119864minus03
975
70353119864minus10
255119864minus10
1499
046
100
000
048
000
0481
00150
100
010
028
014
Colum
nD
3rawdata
06850
0506
03466
0338
0518
97685
1000
7025
1026
146119864minus03
975
67338119864minus10
231119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD
4rawdata
06560
0521
03415
0315
0478
103
696
1062
7143
1089
144119864minus03
975
64332119864minus10
218119864minus10
1499
046
100
000
048
000
0481
00150
100
010
029
015
Colum
nD5rawdata
06190
0545
03371
0282
0425
109
692
1118
7100
1147
144119864minus03
975
60334119864minus10
207119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
016
Colum
nD
6rawdata
05760
0587
03383
0238
0359
107
635
1103
6516
1131
157119864minus03
975
56364119864minus10
210119864minus10
1499
046
100
000
048
000
0481
00150
100
010
030
017
Colum
nD
1corrected
07830
0542
04243
0359
0623
86907
1158
3397
434
110119864minus03
267
209
334119864minus10
261119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
013
Colum
nD
2corrected
07230
0584
04221
0301
0521
88857
1186
3212
444
117119864minus03
267
193
353119864minus10
255119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
014
Colum
nD
3corrected
06850
0603
04129
0272
0463
97896
1307
3354
490
112119864minus03
267
183
338119864minus10
231119864minus10
1499
046
100
000
055
000
0550
00150
100
010
024
015
Colum
nD
4corrected
06560
0621
040
730249
0420
103
911
1388
3411
520
110119864minus03
267
175
332119864minus10
218119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
015
Colum
nD
5corrected
06190
0650
04025
0217
0362
109
905
1462
3390
548
110119864minus03
267
165
334119864minus10
207119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
016
Colum
nD
6corrected
05760
0701
04038
0172
0289
107
831
1442
3111
540
120119864minus03
267
154
364119864minus10
210119864minus10
1499
046
100
000
055
000
0550
00150
100
010
025
017
Guiochon
columns
1198641-1198642
(2007)
Colum
nE 1
rawdata
064
500594
03830
0262
0425
356
1119
1735
4371
678
894119864minus04
256
165
804119864minus11
519119864minus11
1500
046
100
000
030
000
0300
000
41300
030
079
047
Colum
nE 2
rawdata
05400
0800
04320
0108
0190
470
1000
1853
2161
400
100119864minus03
463
250
729119864minus11
393119864minus11
1500
046
100
000
027
000
0270
000
41300
030
070
056
Colum
nE 1
corrected
064
500594
040
000245
040
8356
969
1502
3628
563
103119864minus03
267
172
804119864minus11
519119864minus11
1500
046
100
000
028
000
0279
000
41300
030
075
047
Colum
nE 2
corrected
05400
0800
04320
0108
0190
470
577
1068
2161
400
173119864minus03
267
144
729119864minus11
393119864minus11
1500
046
088
000
023
000
0205
000
41300
030
070
056
Guiochon
columns
1198651ndash1198654
(2008)
Colum
nF 1
rawdata
06350
0586
03720
0263
0419
197
725
1142
4865
766
138119864minus03
149
95399119864minus11
253119864minus11
300
021
100
000
017
000
0170
00035
100
048
129
076
Colum
nF 2
rawdata
064
200581
03730
0269
0429
361
807
1258
4863
758
124119864minus03
166
107
358119864minus11
230119864minus11
500
021
100
000
017
000
0170
00035
100
048
129
075
Colum
nF 3
rawdata
064
100588
03770
0264
0424
670
748
1166
4643
724
134119864minus03
161
103
387119864minus11
248119864minus11
1000
021
100
000
017
000
0170
00035
100
048
128
075
Colum
nF 4
rawdata
06390
0595
03800
0259
0418
1014
752
1177
4476
701
133119864minus03
168
107
384119864minus11
246119864minus11
1500
021
100
000
017
000
0170
00035
100
048
127
075
Colum
nF 1
corrected
06350
0630
040
000235
0392
197
954
1502
3572
563
105119864minus03
267
170
399119864minus11
253119864minus11
300
021
100
000
019
000
0195
00035
100
048
120
076
Colum
nF 2
corrected
064
200623
040
000242
0403
361
964
1502
3611
563
104119864minus03
267
171
358119864minus11
230119864minus11
500
021
100
000
019
000
0186
00035
100
048
120
075
Colum
nF 3
corrected
064
100624
040
000241
0402
670
963
1502
360
6563
104119864minus03
267
171
386119864minus11
248119864minus11
1000
021
100
000
019
000
0193
00035
100
048
120
075
Colum
nF 4
corrected
06390
0626
040
000239
0398
1014
960
1502
3594
563
104119864minus03
267
171
384119864minus11
246119864minus11
1500
021
100
000
019
000
0192
00035
100
048
120
075
Journal of Materials 13
Table3Con
tinued
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Guiochon
columns
1198671-1198672serie
s(2010)
Colum
nH
1rawdata
05320
0744
03960
0136
0225
120
563
1057
3125
587
178119864minus03
180
96122119864minus10
652119864minus11
1500
046
100
000
026
000
0262
00104
050
005
013
009
Colum
nH
2rawdata
05420
0686
03720
0170
0271
61747
1379
4152
766
134119864minus03
180
98823119864minus11
446119864minus11
1000
046
100
000
025
000
0248
00054
050
005
013
009
Colum
nH
1corrected
05320
0752
040
000132
0220
120
799
1502
2993
563
125119864minus03
267
142
122119864minus10
652119864minus11
1500
046
100
000
031
000
0313
00104
050
005
013
009
Colum
nH
2corrected
05420
0738
040
000142
0237
61814
1502
3049
563
123119864minus03
267
145
823119864minus11
446119864minus11
1000
046
100
000
026
000
0259
00054
050
005
013
009
Guiochon
columns
1198673-1198674serie
s(2010)
Colum
nH
3rawdata
06270
060
903820
0245
0396
463
700
1117
4296
685
143119864minus03
163
102
447119864minus11
280119864minus11
1500
021
100
000
018
000
0177
00036
050
024
063
038
Colum
nH
4rawdata
05170
0791
040
900108
0183
433
616
1191
2639
511
162119864minus03
233
121
580119864minus11
300119864minus11
1500
021
100
000
019
000
0189
00036
050
024
059
047
Colum
nH
3corrected
06270
0638
040
000227
0378
463
942
1502
3527
563
106119864minus03
267
167
447119864minus11
280119864minus11
1500
021
100
000
021
000
0205
00036
050
024
060
038
Colum
nH
4corrected
05170
0791
040
900108
0183
433
699
1353
2639
511
143119864minus03
265
137
580119864minus11
300119864minus11
1500
021
100
000
020
000
0201
00036
050
024
059
047
Guiochon
columns
1198681-1198686
(2010)
Colum
nI 1rawdata
06580
0562
03700
0288
0457
488
741
1127
5156
784
135119864minus03
144
95390119864minus11
256119864minus11
500
021
100
000
017
000
0170
00104
050
024
065
037
Colum
nI 2rawdata
06550
0576
03770
0278
044
6873
661
1009
4745
724
151119864minus03
139
91437119864minus11
286119864minus11
1000
021
100
000
017
000
0170
00104
050
024
064
037
Colum
nI 3rawdata
06550
0588
03850
0270
0439
1209
610
931
4341
663
164119864minus03
141
92474119864minus11
310119864minus11
1500
021
100
000
017
000
0170
00104
050
024
062
037
Colum
nI 4rawdata
06430
0572
03680
0275
0435
260
787
1225
5153
801
127119864minus03
153
98367119864minus11
236119864minus11
500
030
100
000
017
000
0170
00104
050
012
032
018
Colum
nI 5rawdata
06540
0583
03810
0273
044
144
3683
1044
4531
693
146119864minus03
151
99423119864minus11
277119864minus11
1000
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 6rawdata
06550
0585
03830
0272
044
164
6665
1016
4438
678
150119864minus03
150
98434119864minus11
285119864minus11
1500
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 1corrected
06580
060
8040
000258
0430
488
988
1502
3701
563
101119864minus03
267
176
390119864minus11
256119864minus11
500
021
100
000
020
000
0196
00104
050
024
060
037
Colum
nI 2corrected
06550
0611
040
000255
0425
873
984
1502
3684
563
102119864minus03
267
175
437119864minus11
286119864minus11
1000
021
100
000
021
000
0207
00104
050
024
060
037
Colum
nI 3corrected
06550
0611
040
000255
0425
1209
984
1502
3684
563
102119864minus03
267
175
474119864minus11
310119864minus11
1500
021
100
000
022
000
0216
00104
050
024
060
037
Colum
nI 4corrected
06430
0622
040
000243
040
5260
966
1502
3617
563
104119864minus03
267
172
367119864minus11
236119864minus11
500
030
100
000
019
000
0188
00104
050
012
029
018
Colum
nI 5corrected
06540
0612
040
000254
0423
443
982
1502
3679
563
102119864minus03
267
175
423119864minus11
277119864minus11
1000
030
100
000
020
000
0204
00104
050
012
029
018
Colum
nI 6corrected
06550
0611
040
000255
0425
646
984
1502
3684
563
102119864minus03
267
175
434119864minus11
285119864minus11
1500
030
100
000
021
000
0207
00104
050
012
029
018
Guiochon
columns
1198691-1198692
(2011)
Colum
nJ 1rawdata90
∘A(120578
=00054
P)04980
0803
040
000098
0163
65654
1313
2801
563
153119864minus03
234
116126119864minus10
627119864minus11
1500
046
100
000
029
000
0287
00054
050
005
013
010
Colum
nJ 2rawdata90
∘A(120578
=00104
P)04980
0803
040
000098
0163
124
651
1307
2801
563
154119864minus03
232
116127119864minus10
630119864minus11
1500
046
100
000
029
000
0287
00104
050
005
013
010
Colum
nJ 1rawdata160
∘A(120578
=00054
P)05630
0714
04020
0161
0269
71896
1592
3099
550
112119864minus03
289
163
102119864minus10
573119864minus11
1500
046
100
000
030
000
0302
00054
050
005
012
009
Colum
nJ 2rawdata160
∘A(120578
=00104
P)05630
0714
04020
0161
0269
129
847
1504
3099
550
118119864minus03
273
154
108119864minus10
606119864minus11
1500
046
100
000
030
000
0302
00104
050
005
012
009
Colum
nJ 1corrected
04980
0803
040
000098
0163
65748
1502
2801
563
134119864minus03
267
133
126119864minus10
627119864minus11
1500
046
100
000
031
000
0307
00054
050
005
013
010
Colum
nJ 2corrected
04980
0803
040
000098
0163
124
748
1502
2801
563
134119864minus03
267
133
127119864minus10
630119864minus11
1500
046
100
000
031
000
0308
00104
050
005
013
010
Colum
nJ 1corrected
05630
0714
04020
0161
0269
71827
1470
3099
550
121119864minus03
267
150
102119864minus10
573119864minus11
1500
046
100
000
029
000
0290
00054
050
005
012
009
Colum
nJ 2corrected
05630
0714
04020
0161
0269
129
827
1470
3099
550
121119864minus03
267
150
108119864minus10
606119864minus11
1500
046
100
000
030
000
0299
00104
050
005
012
009
14 Journal of Materials
050
100150200250300350400
045 050 055 060 065 070 075 080 085
LineAll corrected data
Linear (all corrected data)Squares raw data Triangles corrected data
Pt
120576t
D seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
J1 J2 H4
H1H2
D6
B2 B3 B4
J3 J4
E2
C
H3
D5 D4
E1
D2
A1
D1
A2
y = 267xLine slope = 267
Figure 2 Summary of experimental results
0
500
1000
1500
2000
2500
20 25 30 35 40 45 50 55 60 65 70 75 80
Linear (line)
Ψ
LineAll corrected dataD seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
Line slope (P0) = 267
J1 J2
J3 J4E2
E1
A1
A2
H4 H1
C
120601
H2H3I1 I2 I3 I4 I5
F1 F2 F3 F4
D1 D2 D3 D4 D5 D6
Squares raw data Triangles corrected data
Figure 3 Summary of experimental results
Figures 2 and 3 herein are graphical representations ofTable 31015840s raw and corrected data for Guiochonrsquos columnsplotted in the same reference frames as in Figures 1(a) and1(b) respectively
As shown in the plots of the thirty-one total columnsevaluated in this data base only six columns had reportedraw data values for 119875
0close to the normThese were columns
B2 C E
1 H
4 J
1and J
2 Interestingly of the six conforming
columns 3 of the columns were packed with fully porousparticles and 3 columns with partially porous particles Itcan be seen however that when properly interpreted andadjusted all of Guiochonrsquos experimental data falls fairlywithin the norms outlined in the reference chart in Table 1In order to explore the reasons why the raw data does notconform more closely to the norms outlined in Table 1 eachof Guiochonrsquos studies included in Table 3 is evaluated in turn
Finally Table 4 herein is a list of symbols and parameterdefinitions used throughout this paper as well as their unitsof measure
91 Guiochonrsquos 1998 Study In a 1998 publication by Kohand Guiochon entitled ldquoEffect of the Column Length on theCharacteristics of the Packed Bed and the Column Efficiencyin a Dynamic Axial Compression Columnrdquo [19] Guiochonand his coauthor report the results of a study they hadcarried out on column packing using a rather novel techniquewhich was a combination of slurry packing and mechanicalcompression
The study involved the packing of a single cylinder ofdiameter 5 cm to various column lengths ranging between 1and 25 cm approximately The particles were all silica basedand coated with a reverse stationary phase C
18 However two
different categories of particles were involved One categorycontained highly irregular particles which were about 130micrometers in average diameter In addition these particleshad a high specific pore volume (11mLg) and exhibitedsignificant particle breakage under the packing conditionsinvestigated in the study Conversely the other particlecategory involved very small particles (6 micrometers) whichwere spherical in shape had a low specific pore volume(03mLg) and exhibited a greater ability to withstandparticle breakage under the studyrsquos packing conditions
The permeability data for the irregular particles isneglected herein because according to the authors theparticles exhibited significant breakage during packing andwithout an accurate value for the characteristic dimensionof the particle the data cannot be used in the Kozeny-Blake model to determine the value of 119875
0 The spherical
particles on the other hand behaved successfully under thepacking conditions chosen and yielded values calculated bythe authors for 119875
0of 619 268 226 and 229 for the four
different lengths of columns studied The data for thesecolumns is reported in our Table 3 in the rows for ColumnsB1through B
4 The higher value of 619 (the shortest column)
was rejected by the authors because they felt that either theparticles had been somewhat broken during packing or thecolumn frits had been blocked with fines Accordingly thepermeability data for this column is likewise neglected herein
The remaining three columns however were not con-sistent with respect to the authorsrsquo calculated values for 119875
0
Since the particles were identical the range of values for 1198750
of 226 to 268 must be due to experimental error The rawdata for column B
2 which produced a 119875
0value of 268 is
all self-consistent For example it generates a value for 119875119905
of 148 which is a reasonable value based upon the knownlow particle porosity for these Nova Pak particles and acorresponding value for 120576
0of 04073 which is a reasonable
value for a well-packed column However the other twocolumns columns B
3and B
4 had the much lower value of
126 for 119875119905 Accordingly in Table 3 herein corrected values
for column external porosity for columns B3and B
4are
displayed These corrected values are only slightly differentthan the reported values are within the experimental errorof the measurement and therefore in combination with the
Journal of Materials 15
Table 4 Glossary of terms
Symbol Unit dimensions (cgs) Definition119860 cm2 Column cross-sectional area119863 cm Column diameter119889119901
cm Particle spherical diameter equivalentΔ119875 gcmminus1secminus2 Pressure drop along column119889119901119898
cm Average particle diameter (measured)1198960
None Guiochonrsquos residual reciprocal permeability coefficient119875119905
None Guiochonrsquos modified permeability coefficient in Kozeny-Blake equation119871 cm Column length1198750
None Permeability coefficient in Kozeny-Blake equation119876 cm3minminus1 (exception) Volumetric fluid flow rate1199050
sec Time for elution of totally permeating unretained solute119896 cm2 General permeability coefficient in Darcy equation119870
119865cm2 Halaszrsquos permeability coefficient in Darcy equation (1974 paper)
119870 cm2 Halaszrsquos permeability coefficient in Darcy equation (1972 paper)119870
119888None Guiochonrsquos permeability coefficient in the Kozeny-Blake equation (as modified by Carmen)
Greek1205760
None External column porosity120576119894
None Internal column porosity120576119905
None Total column porosity120576119901
None Particle porosityΦ None Giddingsrsquo ratio of external and total column porosity120601 None Flow resistance parameter based on mobile phase velocity1206010
None Flow resistance parameter based on superficial velocityΩ
119901None Particle sphericity factor
120578 gcmminus1secminus1 Fluid absolute viscosity120583119894
cmsecminus1 Average fluid interstitial linear velocity120583119904
cmsecminus1 Average fluid superficial linear velocity120583119905
cmsecminus1 Average mobile phase velocityΨ None Porosity dependence term in the Kozeny-Blake equation based on mobile phase velocityΨ
0None Porosity dependence term in the Kozeny-Blake equation based on superficial velocity
raw data for column B2 are supportive of the value of 267 for
1198750
92 Guiochonrsquos 1999 Study In a 1999 publication by Farkaset al entitled ldquoValidity of Darcyrsquos Law at Low Flow Rates inLiquid Chromatographyrdquo [20] Guiochon and his coauthorsreport the results of some very exactingmeasurements whichthey made to validate Darcyrsquos law at extremely low fluidflow rates Using a column containing particles packed toa measured external column porosity of 0399 and havinga measured total column porosity of 0593 (Figure 2 in thepaper) Guiochon et al measured the column pressure dropand fluid flow rate at flow rates ranging between 0015 and05mLminwith ethylene glycol as the fluidThe authors statethat the particles in the column are spherical and they appearto have gone to extraordinary lengths to obtain an accuratemeasure of the particle diameter and of course as a result ofmodern techniques of particle size classification the particlesize distribution is very narrow
In Figure 1 of the paper the authors show a plot ofcolumn measured pressure drop in bar versus ldquofluid linearvelocityrdquo in cmsec This plot represents their measurementstaken in the empty column The only velocity that exists in
an empty column is the superficial velocity and thus theabscissa values in the plot in Figure 1 designated as ldquofluidlinear velocityrdquo represent superficial linear velocity
In Figure 2 in their paper however the authors showanother plot of measured pressure drop in bar also versusldquofluid linear velocityrdquo in cmsec The velocity on the abscissain this plot although also designated as ldquofluid linear velocityrdquodoes not equate (for permeability related purposes) to theldquofluid linear velocityrdquo in Figure 1 This is because in Figure 2the columnwas filled with porous particles and themeasuredvelocity was the mobile phase velocity Indeed the onlyvelocity used throughout the entire paper is the mobile phasevelocity (mobile phase velocity and superficial velocity areone and the same in an empty column)
In Table 1 of their paper the authors report that for thecolumn represented in Figure 2 the slope of a line achievedwhen the mobile phase velocityin units of cmsec is plottedversus pressure drop in units of bar is 1829 They also reportthat the flow resistance parameter 120601 for this column is 865In Table 3 herein it can be seen that the raw data for thiscolumn (Column C) generates an apparent value of 257 for1198750 This value is very close to Giddingsrsquo value of 267 Indeed
by referring to Figures 2 and 3 herein one can see that thereported raw data in this study without any modification
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
Journal of Materials 13
Table3Con
tinued
119870119865
Particlecolum
ndescrip
tion
120576119905
Φ1205760
120576119894
120576119901
Δ119875
1206011206010
ΨΨ0
1198960
1198750
119875119905
119870119896
119871119863
Ω119901
119889119901119898
119889119901
120578119876
120583119904
120583119894
120583119905
Units
Non
eNon
eNon
eNon
eNon
ebar
Non
eNon
eNon
eNon
eNon
eNon
eNon
ecm
2cm
2cm
cmNon
ecm
cmgcmminus1secminus1cm
3 minminus1cm
secminus1cm
secminus1cm
secminus1
Guiochon
columns
1198671-1198672serie
s(2010)
Colum
nH
1rawdata
05320
0744
03960
0136
0225
120
563
1057
3125
587
178119864minus03
180
96122119864minus10
652119864minus11
1500
046
100
000
026
000
0262
00104
050
005
013
009
Colum
nH
2rawdata
05420
0686
03720
0170
0271
61747
1379
4152
766
134119864minus03
180
98823119864minus11
446119864minus11
1000
046
100
000
025
000
0248
00054
050
005
013
009
Colum
nH
1corrected
05320
0752
040
000132
0220
120
799
1502
2993
563
125119864minus03
267
142
122119864minus10
652119864minus11
1500
046
100
000
031
000
0313
00104
050
005
013
009
Colum
nH
2corrected
05420
0738
040
000142
0237
61814
1502
3049
563
123119864minus03
267
145
823119864minus11
446119864minus11
1000
046
100
000
026
000
0259
00054
050
005
013
009
Guiochon
columns
1198673-1198674serie
s(2010)
Colum
nH
3rawdata
06270
060
903820
0245
0396
463
700
1117
4296
685
143119864minus03
163
102
447119864minus11
280119864minus11
1500
021
100
000
018
000
0177
00036
050
024
063
038
Colum
nH
4rawdata
05170
0791
040
900108
0183
433
616
1191
2639
511
162119864minus03
233
121
580119864minus11
300119864minus11
1500
021
100
000
019
000
0189
00036
050
024
059
047
Colum
nH
3corrected
06270
0638
040
000227
0378
463
942
1502
3527
563
106119864minus03
267
167
447119864minus11
280119864minus11
1500
021
100
000
021
000
0205
00036
050
024
060
038
Colum
nH
4corrected
05170
0791
040
900108
0183
433
699
1353
2639
511
143119864minus03
265
137
580119864minus11
300119864minus11
1500
021
100
000
020
000
0201
00036
050
024
059
047
Guiochon
columns
1198681-1198686
(2010)
Colum
nI 1rawdata
06580
0562
03700
0288
0457
488
741
1127
5156
784
135119864minus03
144
95390119864minus11
256119864minus11
500
021
100
000
017
000
0170
00104
050
024
065
037
Colum
nI 2rawdata
06550
0576
03770
0278
044
6873
661
1009
4745
724
151119864minus03
139
91437119864minus11
286119864minus11
1000
021
100
000
017
000
0170
00104
050
024
064
037
Colum
nI 3rawdata
06550
0588
03850
0270
0439
1209
610
931
4341
663
164119864minus03
141
92474119864minus11
310119864minus11
1500
021
100
000
017
000
0170
00104
050
024
062
037
Colum
nI 4rawdata
06430
0572
03680
0275
0435
260
787
1225
5153
801
127119864minus03
153
98367119864minus11
236119864minus11
500
030
100
000
017
000
0170
00104
050
012
032
018
Colum
nI 5rawdata
06540
0583
03810
0273
044
144
3683
1044
4531
693
146119864minus03
151
99423119864minus11
277119864minus11
1000
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 6rawdata
06550
0585
03830
0272
044
164
6665
1016
4438
678
150119864minus03
150
98434119864minus11
285119864minus11
1500
030
100
000
017
000
0170
00104
050
012
031
018
Colum
nI 1corrected
06580
060
8040
000258
0430
488
988
1502
3701
563
101119864minus03
267
176
390119864minus11
256119864minus11
500
021
100
000
020
000
0196
00104
050
024
060
037
Colum
nI 2corrected
06550
0611
040
000255
0425
873
984
1502
3684
563
102119864minus03
267
175
437119864minus11
286119864minus11
1000
021
100
000
021
000
0207
00104
050
024
060
037
Colum
nI 3corrected
06550
0611
040
000255
0425
1209
984
1502
3684
563
102119864minus03
267
175
474119864minus11
310119864minus11
1500
021
100
000
022
000
0216
00104
050
024
060
037
Colum
nI 4corrected
06430
0622
040
000243
040
5260
966
1502
3617
563
104119864minus03
267
172
367119864minus11
236119864minus11
500
030
100
000
019
000
0188
00104
050
012
029
018
Colum
nI 5corrected
06540
0612
040
000254
0423
443
982
1502
3679
563
102119864minus03
267
175
423119864minus11
277119864minus11
1000
030
100
000
020
000
0204
00104
050
012
029
018
Colum
nI 6corrected
06550
0611
040
000255
0425
646
984
1502
3684
563
102119864minus03
267
175
434119864minus11
285119864minus11
1500
030
100
000
021
000
0207
00104
050
012
029
018
Guiochon
columns
1198691-1198692
(2011)
Colum
nJ 1rawdata90
∘A(120578
=00054
P)04980
0803
040
000098
0163
65654
1313
2801
563
153119864minus03
234
116126119864minus10
627119864minus11
1500
046
100
000
029
000
0287
00054
050
005
013
010
Colum
nJ 2rawdata90
∘A(120578
=00104
P)04980
0803
040
000098
0163
124
651
1307
2801
563
154119864minus03
232
116127119864minus10
630119864minus11
1500
046
100
000
029
000
0287
00104
050
005
013
010
Colum
nJ 1rawdata160
∘A(120578
=00054
P)05630
0714
04020
0161
0269
71896
1592
3099
550
112119864minus03
289
163
102119864minus10
573119864minus11
1500
046
100
000
030
000
0302
00054
050
005
012
009
Colum
nJ 2rawdata160
∘A(120578
=00104
P)05630
0714
04020
0161
0269
129
847
1504
3099
550
118119864minus03
273
154
108119864minus10
606119864minus11
1500
046
100
000
030
000
0302
00104
050
005
012
009
Colum
nJ 1corrected
04980
0803
040
000098
0163
65748
1502
2801
563
134119864minus03
267
133
126119864minus10
627119864minus11
1500
046
100
000
031
000
0307
00054
050
005
013
010
Colum
nJ 2corrected
04980
0803
040
000098
0163
124
748
1502
2801
563
134119864minus03
267
133
127119864minus10
630119864minus11
1500
046
100
000
031
000
0308
00104
050
005
013
010
Colum
nJ 1corrected
05630
0714
04020
0161
0269
71827
1470
3099
550
121119864minus03
267
150
102119864minus10
573119864minus11
1500
046
100
000
029
000
0290
00054
050
005
012
009
Colum
nJ 2corrected
05630
0714
04020
0161
0269
129
827
1470
3099
550
121119864minus03
267
150
108119864minus10
606119864minus11
1500
046
100
000
030
000
0299
00104
050
005
012
009
14 Journal of Materials
050
100150200250300350400
045 050 055 060 065 070 075 080 085
LineAll corrected data
Linear (all corrected data)Squares raw data Triangles corrected data
Pt
120576t
D seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
J1 J2 H4
H1H2
D6
B2 B3 B4
J3 J4
E2
C
H3
D5 D4
E1
D2
A1
D1
A2
y = 267xLine slope = 267
Figure 2 Summary of experimental results
0
500
1000
1500
2000
2500
20 25 30 35 40 45 50 55 60 65 70 75 80
Linear (line)
Ψ
LineAll corrected dataD seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
Line slope (P0) = 267
J1 J2
J3 J4E2
E1
A1
A2
H4 H1
C
120601
H2H3I1 I2 I3 I4 I5
F1 F2 F3 F4
D1 D2 D3 D4 D5 D6
Squares raw data Triangles corrected data
Figure 3 Summary of experimental results
Figures 2 and 3 herein are graphical representations ofTable 31015840s raw and corrected data for Guiochonrsquos columnsplotted in the same reference frames as in Figures 1(a) and1(b) respectively
As shown in the plots of the thirty-one total columnsevaluated in this data base only six columns had reportedraw data values for 119875
0close to the normThese were columns
B2 C E
1 H
4 J
1and J
2 Interestingly of the six conforming
columns 3 of the columns were packed with fully porousparticles and 3 columns with partially porous particles Itcan be seen however that when properly interpreted andadjusted all of Guiochonrsquos experimental data falls fairlywithin the norms outlined in the reference chart in Table 1In order to explore the reasons why the raw data does notconform more closely to the norms outlined in Table 1 eachof Guiochonrsquos studies included in Table 3 is evaluated in turn
Finally Table 4 herein is a list of symbols and parameterdefinitions used throughout this paper as well as their unitsof measure
91 Guiochonrsquos 1998 Study In a 1998 publication by Kohand Guiochon entitled ldquoEffect of the Column Length on theCharacteristics of the Packed Bed and the Column Efficiencyin a Dynamic Axial Compression Columnrdquo [19] Guiochonand his coauthor report the results of a study they hadcarried out on column packing using a rather novel techniquewhich was a combination of slurry packing and mechanicalcompression
The study involved the packing of a single cylinder ofdiameter 5 cm to various column lengths ranging between 1and 25 cm approximately The particles were all silica basedand coated with a reverse stationary phase C
18 However two
different categories of particles were involved One categorycontained highly irregular particles which were about 130micrometers in average diameter In addition these particleshad a high specific pore volume (11mLg) and exhibitedsignificant particle breakage under the packing conditionsinvestigated in the study Conversely the other particlecategory involved very small particles (6 micrometers) whichwere spherical in shape had a low specific pore volume(03mLg) and exhibited a greater ability to withstandparticle breakage under the studyrsquos packing conditions
The permeability data for the irregular particles isneglected herein because according to the authors theparticles exhibited significant breakage during packing andwithout an accurate value for the characteristic dimensionof the particle the data cannot be used in the Kozeny-Blake model to determine the value of 119875
0 The spherical
particles on the other hand behaved successfully under thepacking conditions chosen and yielded values calculated bythe authors for 119875
0of 619 268 226 and 229 for the four
different lengths of columns studied The data for thesecolumns is reported in our Table 3 in the rows for ColumnsB1through B
4 The higher value of 619 (the shortest column)
was rejected by the authors because they felt that either theparticles had been somewhat broken during packing or thecolumn frits had been blocked with fines Accordingly thepermeability data for this column is likewise neglected herein
The remaining three columns however were not con-sistent with respect to the authorsrsquo calculated values for 119875
0
Since the particles were identical the range of values for 1198750
of 226 to 268 must be due to experimental error The rawdata for column B
2 which produced a 119875
0value of 268 is
all self-consistent For example it generates a value for 119875119905
of 148 which is a reasonable value based upon the knownlow particle porosity for these Nova Pak particles and acorresponding value for 120576
0of 04073 which is a reasonable
value for a well-packed column However the other twocolumns columns B
3and B
4 had the much lower value of
126 for 119875119905 Accordingly in Table 3 herein corrected values
for column external porosity for columns B3and B
4are
displayed These corrected values are only slightly differentthan the reported values are within the experimental errorof the measurement and therefore in combination with the
Journal of Materials 15
Table 4 Glossary of terms
Symbol Unit dimensions (cgs) Definition119860 cm2 Column cross-sectional area119863 cm Column diameter119889119901
cm Particle spherical diameter equivalentΔ119875 gcmminus1secminus2 Pressure drop along column119889119901119898
cm Average particle diameter (measured)1198960
None Guiochonrsquos residual reciprocal permeability coefficient119875119905
None Guiochonrsquos modified permeability coefficient in Kozeny-Blake equation119871 cm Column length1198750
None Permeability coefficient in Kozeny-Blake equation119876 cm3minminus1 (exception) Volumetric fluid flow rate1199050
sec Time for elution of totally permeating unretained solute119896 cm2 General permeability coefficient in Darcy equation119870
119865cm2 Halaszrsquos permeability coefficient in Darcy equation (1974 paper)
119870 cm2 Halaszrsquos permeability coefficient in Darcy equation (1972 paper)119870
119888None Guiochonrsquos permeability coefficient in the Kozeny-Blake equation (as modified by Carmen)
Greek1205760
None External column porosity120576119894
None Internal column porosity120576119905
None Total column porosity120576119901
None Particle porosityΦ None Giddingsrsquo ratio of external and total column porosity120601 None Flow resistance parameter based on mobile phase velocity1206010
None Flow resistance parameter based on superficial velocityΩ
119901None Particle sphericity factor
120578 gcmminus1secminus1 Fluid absolute viscosity120583119894
cmsecminus1 Average fluid interstitial linear velocity120583119904
cmsecminus1 Average fluid superficial linear velocity120583119905
cmsecminus1 Average mobile phase velocityΨ None Porosity dependence term in the Kozeny-Blake equation based on mobile phase velocityΨ
0None Porosity dependence term in the Kozeny-Blake equation based on superficial velocity
raw data for column B2 are supportive of the value of 267 for
1198750
92 Guiochonrsquos 1999 Study In a 1999 publication by Farkaset al entitled ldquoValidity of Darcyrsquos Law at Low Flow Rates inLiquid Chromatographyrdquo [20] Guiochon and his coauthorsreport the results of some very exactingmeasurements whichthey made to validate Darcyrsquos law at extremely low fluidflow rates Using a column containing particles packed toa measured external column porosity of 0399 and havinga measured total column porosity of 0593 (Figure 2 in thepaper) Guiochon et al measured the column pressure dropand fluid flow rate at flow rates ranging between 0015 and05mLminwith ethylene glycol as the fluidThe authors statethat the particles in the column are spherical and they appearto have gone to extraordinary lengths to obtain an accuratemeasure of the particle diameter and of course as a result ofmodern techniques of particle size classification the particlesize distribution is very narrow
In Figure 1 of the paper the authors show a plot ofcolumn measured pressure drop in bar versus ldquofluid linearvelocityrdquo in cmsec This plot represents their measurementstaken in the empty column The only velocity that exists in
an empty column is the superficial velocity and thus theabscissa values in the plot in Figure 1 designated as ldquofluidlinear velocityrdquo represent superficial linear velocity
In Figure 2 in their paper however the authors showanother plot of measured pressure drop in bar also versusldquofluid linear velocityrdquo in cmsec The velocity on the abscissain this plot although also designated as ldquofluid linear velocityrdquodoes not equate (for permeability related purposes) to theldquofluid linear velocityrdquo in Figure 1 This is because in Figure 2the columnwas filled with porous particles and themeasuredvelocity was the mobile phase velocity Indeed the onlyvelocity used throughout the entire paper is the mobile phasevelocity (mobile phase velocity and superficial velocity areone and the same in an empty column)
In Table 1 of their paper the authors report that for thecolumn represented in Figure 2 the slope of a line achievedwhen the mobile phase velocityin units of cmsec is plottedversus pressure drop in units of bar is 1829 They also reportthat the flow resistance parameter 120601 for this column is 865In Table 3 herein it can be seen that the raw data for thiscolumn (Column C) generates an apparent value of 257 for1198750 This value is very close to Giddingsrsquo value of 267 Indeed
by referring to Figures 2 and 3 herein one can see that thereported raw data in this study without any modification
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
14 Journal of Materials
050
100150200250300350400
045 050 055 060 065 070 075 080 085
LineAll corrected data
Linear (all corrected data)Squares raw data Triangles corrected data
Pt
120576t
D seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
J1 J2 H4
H1H2
D6
B2 B3 B4
J3 J4
E2
C
H3
D5 D4
E1
D2
A1
D1
A2
y = 267xLine slope = 267
Figure 2 Summary of experimental results
0
500
1000
1500
2000
2500
20 25 30 35 40 45 50 55 60 65 70 75 80
Linear (line)
Ψ
LineAll corrected dataD seriesI seriesJ seriesF series
H seriesA seriesB seriesC seriesE series
Line slope (P0) = 267
J1 J2
J3 J4E2
E1
A1
A2
H4 H1
C
120601
H2H3I1 I2 I3 I4 I5
F1 F2 F3 F4
D1 D2 D3 D4 D5 D6
Squares raw data Triangles corrected data
Figure 3 Summary of experimental results
Figures 2 and 3 herein are graphical representations ofTable 31015840s raw and corrected data for Guiochonrsquos columnsplotted in the same reference frames as in Figures 1(a) and1(b) respectively
As shown in the plots of the thirty-one total columnsevaluated in this data base only six columns had reportedraw data values for 119875
0close to the normThese were columns
B2 C E
1 H
4 J
1and J
2 Interestingly of the six conforming
columns 3 of the columns were packed with fully porousparticles and 3 columns with partially porous particles Itcan be seen however that when properly interpreted andadjusted all of Guiochonrsquos experimental data falls fairlywithin the norms outlined in the reference chart in Table 1In order to explore the reasons why the raw data does notconform more closely to the norms outlined in Table 1 eachof Guiochonrsquos studies included in Table 3 is evaluated in turn
Finally Table 4 herein is a list of symbols and parameterdefinitions used throughout this paper as well as their unitsof measure
91 Guiochonrsquos 1998 Study In a 1998 publication by Kohand Guiochon entitled ldquoEffect of the Column Length on theCharacteristics of the Packed Bed and the Column Efficiencyin a Dynamic Axial Compression Columnrdquo [19] Guiochonand his coauthor report the results of a study they hadcarried out on column packing using a rather novel techniquewhich was a combination of slurry packing and mechanicalcompression
The study involved the packing of a single cylinder ofdiameter 5 cm to various column lengths ranging between 1and 25 cm approximately The particles were all silica basedand coated with a reverse stationary phase C
18 However two
different categories of particles were involved One categorycontained highly irregular particles which were about 130micrometers in average diameter In addition these particleshad a high specific pore volume (11mLg) and exhibitedsignificant particle breakage under the packing conditionsinvestigated in the study Conversely the other particlecategory involved very small particles (6 micrometers) whichwere spherical in shape had a low specific pore volume(03mLg) and exhibited a greater ability to withstandparticle breakage under the studyrsquos packing conditions
The permeability data for the irregular particles isneglected herein because according to the authors theparticles exhibited significant breakage during packing andwithout an accurate value for the characteristic dimensionof the particle the data cannot be used in the Kozeny-Blake model to determine the value of 119875
0 The spherical
particles on the other hand behaved successfully under thepacking conditions chosen and yielded values calculated bythe authors for 119875
0of 619 268 226 and 229 for the four
different lengths of columns studied The data for thesecolumns is reported in our Table 3 in the rows for ColumnsB1through B
4 The higher value of 619 (the shortest column)
was rejected by the authors because they felt that either theparticles had been somewhat broken during packing or thecolumn frits had been blocked with fines Accordingly thepermeability data for this column is likewise neglected herein
The remaining three columns however were not con-sistent with respect to the authorsrsquo calculated values for 119875
0
Since the particles were identical the range of values for 1198750
of 226 to 268 must be due to experimental error The rawdata for column B
2 which produced a 119875
0value of 268 is
all self-consistent For example it generates a value for 119875119905
of 148 which is a reasonable value based upon the knownlow particle porosity for these Nova Pak particles and acorresponding value for 120576
0of 04073 which is a reasonable
value for a well-packed column However the other twocolumns columns B
3and B
4 had the much lower value of
126 for 119875119905 Accordingly in Table 3 herein corrected values
for column external porosity for columns B3and B
4are
displayed These corrected values are only slightly differentthan the reported values are within the experimental errorof the measurement and therefore in combination with the
Journal of Materials 15
Table 4 Glossary of terms
Symbol Unit dimensions (cgs) Definition119860 cm2 Column cross-sectional area119863 cm Column diameter119889119901
cm Particle spherical diameter equivalentΔ119875 gcmminus1secminus2 Pressure drop along column119889119901119898
cm Average particle diameter (measured)1198960
None Guiochonrsquos residual reciprocal permeability coefficient119875119905
None Guiochonrsquos modified permeability coefficient in Kozeny-Blake equation119871 cm Column length1198750
None Permeability coefficient in Kozeny-Blake equation119876 cm3minminus1 (exception) Volumetric fluid flow rate1199050
sec Time for elution of totally permeating unretained solute119896 cm2 General permeability coefficient in Darcy equation119870
119865cm2 Halaszrsquos permeability coefficient in Darcy equation (1974 paper)
119870 cm2 Halaszrsquos permeability coefficient in Darcy equation (1972 paper)119870
119888None Guiochonrsquos permeability coefficient in the Kozeny-Blake equation (as modified by Carmen)
Greek1205760
None External column porosity120576119894
None Internal column porosity120576119905
None Total column porosity120576119901
None Particle porosityΦ None Giddingsrsquo ratio of external and total column porosity120601 None Flow resistance parameter based on mobile phase velocity1206010
None Flow resistance parameter based on superficial velocityΩ
119901None Particle sphericity factor
120578 gcmminus1secminus1 Fluid absolute viscosity120583119894
cmsecminus1 Average fluid interstitial linear velocity120583119904
cmsecminus1 Average fluid superficial linear velocity120583119905
cmsecminus1 Average mobile phase velocityΨ None Porosity dependence term in the Kozeny-Blake equation based on mobile phase velocityΨ
0None Porosity dependence term in the Kozeny-Blake equation based on superficial velocity
raw data for column B2 are supportive of the value of 267 for
1198750
92 Guiochonrsquos 1999 Study In a 1999 publication by Farkaset al entitled ldquoValidity of Darcyrsquos Law at Low Flow Rates inLiquid Chromatographyrdquo [20] Guiochon and his coauthorsreport the results of some very exactingmeasurements whichthey made to validate Darcyrsquos law at extremely low fluidflow rates Using a column containing particles packed toa measured external column porosity of 0399 and havinga measured total column porosity of 0593 (Figure 2 in thepaper) Guiochon et al measured the column pressure dropand fluid flow rate at flow rates ranging between 0015 and05mLminwith ethylene glycol as the fluidThe authors statethat the particles in the column are spherical and they appearto have gone to extraordinary lengths to obtain an accuratemeasure of the particle diameter and of course as a result ofmodern techniques of particle size classification the particlesize distribution is very narrow
In Figure 1 of the paper the authors show a plot ofcolumn measured pressure drop in bar versus ldquofluid linearvelocityrdquo in cmsec This plot represents their measurementstaken in the empty column The only velocity that exists in
an empty column is the superficial velocity and thus theabscissa values in the plot in Figure 1 designated as ldquofluidlinear velocityrdquo represent superficial linear velocity
In Figure 2 in their paper however the authors showanother plot of measured pressure drop in bar also versusldquofluid linear velocityrdquo in cmsec The velocity on the abscissain this plot although also designated as ldquofluid linear velocityrdquodoes not equate (for permeability related purposes) to theldquofluid linear velocityrdquo in Figure 1 This is because in Figure 2the columnwas filled with porous particles and themeasuredvelocity was the mobile phase velocity Indeed the onlyvelocity used throughout the entire paper is the mobile phasevelocity (mobile phase velocity and superficial velocity areone and the same in an empty column)
In Table 1 of their paper the authors report that for thecolumn represented in Figure 2 the slope of a line achievedwhen the mobile phase velocityin units of cmsec is plottedversus pressure drop in units of bar is 1829 They also reportthat the flow resistance parameter 120601 for this column is 865In Table 3 herein it can be seen that the raw data for thiscolumn (Column C) generates an apparent value of 257 for1198750 This value is very close to Giddingsrsquo value of 267 Indeed
by referring to Figures 2 and 3 herein one can see that thereported raw data in this study without any modification
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
Journal of Materials 15
Table 4 Glossary of terms
Symbol Unit dimensions (cgs) Definition119860 cm2 Column cross-sectional area119863 cm Column diameter119889119901
cm Particle spherical diameter equivalentΔ119875 gcmminus1secminus2 Pressure drop along column119889119901119898
cm Average particle diameter (measured)1198960
None Guiochonrsquos residual reciprocal permeability coefficient119875119905
None Guiochonrsquos modified permeability coefficient in Kozeny-Blake equation119871 cm Column length1198750
None Permeability coefficient in Kozeny-Blake equation119876 cm3minminus1 (exception) Volumetric fluid flow rate1199050
sec Time for elution of totally permeating unretained solute119896 cm2 General permeability coefficient in Darcy equation119870
119865cm2 Halaszrsquos permeability coefficient in Darcy equation (1974 paper)
119870 cm2 Halaszrsquos permeability coefficient in Darcy equation (1972 paper)119870
119888None Guiochonrsquos permeability coefficient in the Kozeny-Blake equation (as modified by Carmen)
Greek1205760
None External column porosity120576119894
None Internal column porosity120576119905
None Total column porosity120576119901
None Particle porosityΦ None Giddingsrsquo ratio of external and total column porosity120601 None Flow resistance parameter based on mobile phase velocity1206010
None Flow resistance parameter based on superficial velocityΩ
119901None Particle sphericity factor
120578 gcmminus1secminus1 Fluid absolute viscosity120583119894
cmsecminus1 Average fluid interstitial linear velocity120583119904
cmsecminus1 Average fluid superficial linear velocity120583119905
cmsecminus1 Average mobile phase velocityΨ None Porosity dependence term in the Kozeny-Blake equation based on mobile phase velocityΨ
0None Porosity dependence term in the Kozeny-Blake equation based on superficial velocity
raw data for column B2 are supportive of the value of 267 for
1198750
92 Guiochonrsquos 1999 Study In a 1999 publication by Farkaset al entitled ldquoValidity of Darcyrsquos Law at Low Flow Rates inLiquid Chromatographyrdquo [20] Guiochon and his coauthorsreport the results of some very exactingmeasurements whichthey made to validate Darcyrsquos law at extremely low fluidflow rates Using a column containing particles packed toa measured external column porosity of 0399 and havinga measured total column porosity of 0593 (Figure 2 in thepaper) Guiochon et al measured the column pressure dropand fluid flow rate at flow rates ranging between 0015 and05mLminwith ethylene glycol as the fluidThe authors statethat the particles in the column are spherical and they appearto have gone to extraordinary lengths to obtain an accuratemeasure of the particle diameter and of course as a result ofmodern techniques of particle size classification the particlesize distribution is very narrow
In Figure 1 of the paper the authors show a plot ofcolumn measured pressure drop in bar versus ldquofluid linearvelocityrdquo in cmsec This plot represents their measurementstaken in the empty column The only velocity that exists in
an empty column is the superficial velocity and thus theabscissa values in the plot in Figure 1 designated as ldquofluidlinear velocityrdquo represent superficial linear velocity
In Figure 2 in their paper however the authors showanother plot of measured pressure drop in bar also versusldquofluid linear velocityrdquo in cmsec The velocity on the abscissain this plot although also designated as ldquofluid linear velocityrdquodoes not equate (for permeability related purposes) to theldquofluid linear velocityrdquo in Figure 1 This is because in Figure 2the columnwas filled with porous particles and themeasuredvelocity was the mobile phase velocity Indeed the onlyvelocity used throughout the entire paper is the mobile phasevelocity (mobile phase velocity and superficial velocity areone and the same in an empty column)
In Table 1 of their paper the authors report that for thecolumn represented in Figure 2 the slope of a line achievedwhen the mobile phase velocityin units of cmsec is plottedversus pressure drop in units of bar is 1829 They also reportthat the flow resistance parameter 120601 for this column is 865In Table 3 herein it can be seen that the raw data for thiscolumn (Column C) generates an apparent value of 257 for1198750 This value is very close to Giddingsrsquo value of 267 Indeed
by referring to Figures 2 and 3 herein one can see that thereported raw data in this study without any modification
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
16 Journal of Materials
whatsoever essentially confirms our frame of reference in allrespects
Although unnecessary because the small discrepancy inthe values of 119875
0could result from experimental error in
almost any of the measurements we believe that even thatcan be explained Because of the extraordinarily sensitivenature of the porosity dependence term in the Kozeny-Blakeequation it is suggested herein that this small differencearises from on the one hand the authorsrsquo technique ofusing a mixture of methanol and water as the fluid fortheir determination of the value of 120576
119905and their use of
dichloromethane as the fluid in their determination of thevalue of 120576
0and on the other hand their use of ethylene
glycol as the fluid when measuring the column pressuregradient A small difference in the wetting characteristics ofthese three fluids could easily disrupt the balance betweenthe measured values for 120601 and Ψ and accordingly accountfor all the discrepancies Therefore in the next row of Table3 we make a correction for column pressure to account forthis discontinuity in the authorsrsquo experimental protocol Notethat the corrected value is an adjustment upward of about 4which is within the experimental error of the measurementFinally the corrected data for this column establishes a valuefor 119875
119905of 158 which is indicative of an internal porosity for
these particles that is known to be slightly higher than theNova Pak particles reported in Guiochonrsquos 1998 paper Thecare with which the authors made their measurements ofparticle size and flow rates coupled with the measured valueof approximately 04 for 120576
0 again a reasonable value for awell-
packed columnmakes this study of column permeability oneof the most credible and most important in the publishedliterature
93 Guiochonrsquos 2006 Study In a 2006 publication with Fab-rice Gritti entitled ldquoExperimental Evidence of the Influenceof the Surface Chemistry of the Packing Material on theColumn Pressure Drop in Reverse-phase Liquid Chromatog-raphyrdquo [21] Guiochon and his coauthor claim to makewhat one can only characterize as a startling revelationIn their application of what they call the ldquoKozeny-Carmanequationrdquo they assert that their data support a value of 975 forthe Kozeny-Carman ldquoconstantrdquo which they acknowledge isldquonearly one-half the value traditionally acceptedrdquo Moreoverthe authors go on to suggestmdashequally surprisinglymdashthat thenature of the chemical coating on the surface of the particlesadds another variable to the relationship between pressuregradient and fluid flow We show the data for these columnsas our series D in Table 3 herein One can immediately see inFigure 3 herein that the raw data reported for these columnsdoes not resemble that of packed columns in any shape orform In fact the data is so bizarre that it falls outside all theboundaries of our frame of reference
To begin with the authors used a novel procedure todetermine the external porosity of the columns which isbased upon the so-called Donnan Effect Unfortunatelythe authors did not include for comparison purposes adetermination of the external columnporosities by some con-ventional technique In view of this lack of cross-validation
one is automatically skeptical of the very low values whichthey derived for the external column porosities
The authors recite the following in their paper ldquothecolumns used in this work are the same as those the adsorp-tion properties of which were reported earlier (referenceincluded)rdquo The reference was to another paper publishedin the same year by the same authors [22] in which thetotal porosity 120576
119905 of these same columns was measured and
reported in Table 1 It ismystifying therefore that the authorsignored these measurements in their analysis of permeabilityreported in this paper In Table 3 herein the values fromthis earlier paper for total column porosity for each of thecolumns in the study are included As can be seen in the tablethe difference between the total column porosities reportedfor these columns in the earlier study and the externalcolumn porosities reported for them in the present studywould indicate internal column porosities which are far toolarge for these particles when compared to the low valuesof 119875
119905dictated by the measured pressure drops
Similarly
such internal column porosities would be at significant oddswith the specifications published by the manufacturer forthese particles The results of ISEC (inverse size exclusionchromatography) measurements on a standard 39 times 150mmSymmetry column were reported by the manufacturer as120576119894= 0265 whereas the measurements by the authors would
establish a range of values for the internal column porositiesfrom 024 to 043
The other thing to note about this study is that thecolumns that the authors used in the study were ldquoprototypecolumnsrdquo which had all been ldquopacked and generously offeredby themanufacturer (Waters MilfordMA USA)rdquo [21 p 193]and rather thanmeasuring the particle diameters themselvesthe authors merely accepted the nominal particle size givenby the manufacturer This is a significant deviation fromGuiochonrsquos standard operating procedure described in his1999 paper discussed above inwhichGuiochon and his coau-thors went to extraordinary lengths to measure accuratelythe particle size underscoring its importance as followsldquothe nominal particle sizes given by manufacturers of silicaadsorbents used in chromatography are often approximateaverages which cannot be used for accurate calculationsof column permeability For this reason we measured theparticle size distribution by laser-beam scattering usinga Mastersizer instrument (Malvern Instruments Southbor-ough MA USA)rdquo [20 p 41] (emphasis added) In defenseof their failure to do likewise in this 2006 paper Guiochonand Gritti state the following ldquosince we did not emptythe columns we had no access to the inner diameter Wemeasured the external diameter of the tube insteadrdquo [21 p196] Although the condition that they not open the columnswas presumably imposed by the manufacturer this does notobviate the need for the authors to have obtained someindependent verification of the particle size especially sinceit is such a sensitive parameter in the pressure dropfluid flowrelationship9
It can be seen from Table 3 that both the nominal particlesize reported by the manufacturer and the external porositiesmeasured by the authors were too low For example lookingat Figure 3 herein it is obvious that the raw data values for
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
Journal of Materials 17
Ψ are greater than the maximum in our reference charts forpacked columns Similarly Figure 2 herein illustrates thatthe values for 119875
119905are too low and do not reflect values for
columns packed with fully porous particles Both the particlesize and the external porosity are suspects Accordingly wededuce that the particle size and value for Φ need to beadjusted As for the particle size we adopt the value of 55micrometers a reasonable deviation from the nominal sizegiven by themanufacturerThen usingGiddingsrsquo value of 267for119875
0 we show the impact upon internal columnporosity due
to variations in the level of surface modified stationary phaseThe corrected data all makes sense The column internal
porosity is at its highest for Column D1which contained
underivatized silica particles This is corroborated by thevalue of 119875
119905 which is also at its highest for this column
On the other hand both these two parameters are at theirlowest values for Column D
6 which is consistent with the
highest coating of C18stationary phase and is themore typical
value for commercially available reverse phase C18columns
Additionally the corrected column porosities are consistentwith reported values for standard symmetry columns soldby Waters and are consistent with a professionally developedslurry column packing protocol Finally the range of cor-rected 120601 values is totally consistent with what one wouldexpect
Finally it should be noted that in the very next yearafter they published this paper these same authors publishedanother paper on column permeability which is examinedbelow in which they discontinue the use of the DonnanEffect to measure column external porosity and where theyrevert to using the ISEC technique10 In essence then evenGuiochon himself has effectively abandoned this procedurebut paradoxically he still clings to the erroneous conclusionsbased upon its use (see his comments in his 2010 paperreviewed below concerning the allegedly embedded uniden-tified variables in the value of119870
119888)
94 Guiochonrsquos 2007 Study In a 2007 publication withcoauthors Gritti et al entitled ldquoComparison Between theEfficiencies of Columns Packed with Fully and PartiallyPorous C
18-bonded SilicaMaterialsrdquo [23] (ColumnE series in
Table 3 herein) Guiochon discusses the pressure dropfluidflow relationship in terms of the ldquoKozeny-Carman equationrdquoIn this paper the authors compare the permeability oftwo different types of columns one set filled with fullyporous particles and the other set filled with pellicular orpartially porous particles the latter of which they claim arerepresentative of a new paradigm in columnparticle designaimed at themore challengingmodern-day chromatographicapplications where speed of analysis combined with highefficiency require the use of very high total column pressures
In Figure 4 of their paper the authors display two valuesfor 119870
119888 which they define as the ldquoKozeny-Carman constantrdquo
The value of 250 supposedly corresponds to the permeabilitydata set for the columns containing the partially porousparticles and the value of 165 supposedly corresponds to thepermeability data set for the columns containing the fullyporous particles Although the plot shows a total of 13 data
points the methodology used to derive the values for theconstant on the ordinate of the plot is blind to the readerIn other words the authors do not show any connectionbetween the measured column pressures and the ordinatevalues in their Figure 4 for the values of the ldquoconstantrdquoMoreover although the caption under their Figure 4 statesthat the fluid used was acetonitrilewaterTFA (604001vv) at 119879 = 295K [23 p 297] none of the pressure datadisplayed in their Figure 2 matches this combination of fluidtemperature and eluent compositions Accordingly itmust beconcluded that the measured column pressures upon whichthe calculated values for the constant in their Figure 4 arebased are omitted from the paper
The authors conclude that ldquothe Kozeny-Carman constantof the Halo column is approximately 250 while that of theother column is close to 170 typical of the result found forconventional beds of spherical porous silica particles (119870
119888asymp
180) The much larger value of 119870119888for the Halo material is
unexpectedrdquo [23 p 295]Using Figure 2(B) from the paper as a reference it is
apparent that the values of 165 and 250 for 119870119888as displayed
in their Figure 4 do not compute These values for theKozeny-Carman constant would produce values of 230 and254 bar respectively for the total column pressure at a fluidvolumetric flow rate of 3mLmin which corresponds to theauthorsrsquo value for 119906
119904(which they display in Figure 2(B) as
being 3mmsecminus1) Surprisingly these values are less thanthose plotted in Figure 2(B) Therefore the values of 165 and250 cannot be representative of the value of119875
0for this data set
of measured column pressures Moreover the mobile phasewhich underlies their values for 119870
119888was at a temperature of
22∘C (295K) which is even lower than that used in Figure2(B) that is 60∘C Accordingly the column pressures whichsupport the 119870
119888values in their Figure 4 must be significantly
greater than 300 bar at this flow rate (fluid viscosity increasesas temperature decreases)
Table 3 herein contains the raw data for both columnsreported in the paper and is displayed as columns E
1and
E2 Referring again to our Figures 2 and 3 it is clear that the
authors mistook 119875119905for 119875
0 Accordingly we show the authorsrsquo
values for 119870119888in Table 3 as being values for 119875
119905 not 119875
0 Thus
corrected the raw data for the fully porous particles generatea value of 165 for 119875
119905and an apparent value of 256 for 119875
0
However themeasured value for external porosity of 03830 isstill on the low end of what is typical for well-packed columnsusing slurry packing In addition this would dictate a value of0425 for particle porosity a value which does notmake sense(too high) given the high pressure underwhich these particlesmust be slurry packed in order to sustain their very highoperating pressures It was the same high particle porosityfor instance in Guiochonrsquos 1998 study reviewed above thatcaused the irregular particles in that study to crush underthe hydraulic pressure of the slurry packing process usedtherein Accordingly we show an adjusted value of 04 forexternal porosity and as dictated by the microscopy resultsdisplayed in the paper for these particles a slightly loweraverage particle size On the basis of these adjustments wederive a value for 119875
0of 267 and a value of 172 for 119875
119905 both of
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
18 Journal of Materials
which are just what our frame of reference would cause us toexpect
With respect to the partially porous Halo particles theauthors refer to them as being ldquorugoserdquo Indeed the electron-micrographs confirm that these particles have a morphologywhich in addition to being quasi-spherical is also roughand scabrous In other words they correspond to whatGuiochon describes in his textbook as irregular particlesAccordingly in order to apply the Kozeny-Blake equation(as modified by Carman) to this data set one must knowthe value for the spherical particle diameter equivalent ofthese particles As displayed in Table 3 the raw data forthe partially porous particles generates a value of 250 for119875119905and an apparent value of 463 for 119875
0 When this data is
corrected for particle sphericity according to the Kozeny-Blake equation (as modified by Carman) the result is a valueof 088 for Ω
119901and a corresponding value of 21 micrometers
for the spherical particle diameter equivalent The resultingcorrected value of 144 for119875
119905is consistent with the low particle
porosity which characterizes the Halo particles used in thisstudy
95 Guiochonrsquos 2008 Study In another paper with Grittipublished in 2008 and entitled ldquoUltra High Pressure LiquidChromatography Column Permeability and Changes of theEluent Propertiesrdquo [3] (Column F series in Table 3 herein)Guiochon reports a study carried out on four columns whichagain were ldquogenerously offered by themanufacturerrdquo [3 p 2]In this paper he and his coauthor draw the conclusion thatldquothe average Kozeny-Carman permeability constant for thecolumns was 144 plusmn 35rdquo [3 p 1]
In Table 1 of their paper Guiochon and Gritti provide alist of column variables some of which were ldquogiven by themanufacturerrdquo and some of which were ldquomeasuredrdquo in theauthorsrsquo labThe authors describe the particles in the columnsunder study as ldquofine particles average119889
119901asymp 17 120583m of bridged
ethylsiloxanesilica hybrid-C18 named BEH-C
18rdquo [3 p 1]
Among the variables which they identify as having been givenby the manufacturer is the 17 120583m value for the particle size
Figure 5 of their paper is a plot of measured pressuredrop in bar versus fluid flow rate in mLmin In our Table 3herein the raw data from the authorsrsquo Figure 5 for eachof the columns in the study is displayed in the rows forColumns F
1through F
4The computed values for119875
0are based
upon the value of the particle size given in Table 1 of thepaper and are the same as those reported by the authorsfor their columns D1 through D4 namely 149 166 161 and168 The corresponding values for 119875
119905in the raw data average
are approximately 100 a value representative of nonporousparticles However we know that the particles under studyare porous because the reported values for 120576
119905are greater
than those reported for 1205760 Moreover the corresponding
particle porosity for these columns would be greater than04 which is entirely unreasonable (again too large) forfully porous particles which have to be slurry packed andoperated at extremely high pressures (greater than 15000 psi)Accordingly as is plainly evident from Figures 2 and 3 hereinthere is something fundamentally wrong with this data set
Since the reported values for total columnporosity are notin question this means that the values for external porosityare again too low and since the average particle size was notmeasured by the authors we adjust for these two parametersAccordingly in Table 3 herein a correction is made tothenominal particle size given by the manufacturer and theexternal porosities are set to the reasonable value of 04 Usingthe resultant corrected value of about 19 micrometers forparticle size (slightly higher than the nominal value given bythe manufacturer) reasonable values for 120601 as well as 119875
119905are
achieved for all columns in the study
96 Guiochonrsquos 2010 Studies Next we consider three moreGuiochon studies of permeability which were all reported in2010 and all of which he again coauthored with Gritti et alThe first of these articles is entitled ldquoPerformance of ColumnsPacked With the New Shell Particles Kinetex-C
18rdquo [24] In
this study the authors report permeability results for partiallyporous particles produced by two different manufacturersThe raw data reported for the two columns are included inour Table 3 as ColumnsH
1andH
2 As reflected in our Table 3
for the raw data and as is plainly evident from Figure 2 and 2herein the resulting values for119875
119905of 96 and 98 are way too low
being more representative of nonporous particles somethingwhich the particles under study are not
The authors after having gone into great detail describingwhat measurements and precautions they used to derive alltheir experimental measurements again fail to measure theaverage particle size Instead they back-calculate the valuefor particle size based upon the Kozeny-Blake equation withan embedded value of 180 for 119875
011 Additionally the external
porosity values are once again on the low side Accordinglyin our Table 3 for comparison purposes we show correctedvalues for the particles which are slightly higher than thecalculated valuesWe point out that using a value of 267 for119875
0
and a set value of 04 for external porosity establishes valuesfor 119875
119905of 142 and 145 values which are very reasonable based
upon the authorsrsquo own statement of the low particle porosityof these partially porous particles
The second of Guiochonrsquos 2010 papers entitled ldquoMassTransfer Resistance in Narrow-bore Columns Packed with17 120583m Particles in Very High Pressure Liquid Chromatog-raphyrdquo [25] further exemplifies the authorsrsquo reliance uponmanufacturersrsquo data as opposed to measuring things In thispaper the authors report two columns having the narrowdiameter of 21mm One column contains fully porousparticles while the other contains partially porous particlesThe authorsrsquo data sets are presented in our Table 3 as ColumnsH
3andH
4 respectively As can be seen from the rawdata and
Figures 2 and 3 herein the results for Column H4(233 for 119875
0
and 121 for 119875119905) are already essentially in conformance with
our frame of reference Moreover when we apply a modestadjustment for particle size over that obtained by the authorrsquosback-calculation technique we get even closer to the norm265 for 119875
0and 137 for 119875
119905
The authors rightfully point out that their ISEC mea-surements result in higher external porosity values for thecolumn containing partially porous particles reporting the
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
Journal of Materials 19
typical value of 04090 for the columnHowever surprisinglythey stop short of the obvious conclusion that since thetechnique of ISEC suffers from the limitation of not being ableto precisely differentiate between the free space between theparticles and the free space within the particles it is logicalthat nonporous particles would result in even higher externalporosities for these slurry packed columns representingas they do (nonporous particles that is) the limit of thephenomenon Indeed the authors appear to be willing to goto great lengths to justify their selective conclusions avoidingthe obvious and more technically relevant explanation whichis that the technique which they are using to determine avalue for external porosity is flawed and results in too lowan external porosity for columns packed with fully porousparticles
Not surprisingly then when we turn to the authorsrsquo dataon the fully porous column in this study we note that theexternal porosity is again low On the other hand when weset the value for external porosity at themore reasonable levelof 04 and again make a modest adjustment for particle sizewe derive the value of 267 for119875
0and the right-in-the-ballpark
value of 167 for 119875119905
The third of Guiochonrsquos 2010 papers entitled ldquoPerfor-mance of New Prototype Packed Columns for Very HighPressure Liquid Chromatographyrdquo [26] once again involvesa failure by the authors to accurately assess the contributionof the size of the particles under study to overall pressuredrop The authors use a value of 17 micrometers for theparticles the value supplied by the manufacturer of theseporous particles
We have included the raw data reported in this paper inour Table 3 as Columns I
1ndashI
6 As was the case in Guiochonrsquos
other recent studies involving fully porous particles thecolumn external porosities appear to be understated Amongother things the reported combination of a high value forcolumn total porosity and a low value for column exter-nal porosity dictates high particle porosity However thesecolumns and particles are designed to be run at extremelyhigh pressures Accordingly the particle porosities need to below in order for the particles towithstand suchhigh pressuresOn the other hand Guiochon and company report values for1198750(which they again notate as 119870
119888) ranging from 139 to 153
This in turn generates values for 119875119905from 91 to 98 values
which again when viewed in Figures 2 and 3 herein areclearly too low because they are representative of nonporousparticles If nothing else then the various reported porositiesare self-contradictory
In our corrected values in Table 3 we have corrected forboth column external porosity and particle size It can be seenthat an average value of 175 is generated for 119875
119905 which fairly
reflects the porosity of these particles (as described by theauthors themselves in their Table 1)
97 Guiochonrsquos 2011 Study Finally we come to Guiochonrsquos2011 publication relevant to our topic yet another paperhe coauthored with Gritti This paper is entitled ldquoThe MassTransfer Kinetics in Columns Packed with Halo-ES ShellParticlesrdquo [27] This study involved two columns of different
particle porosities One column contained particles havinga particle porosity 120576
119901 of 016 while the other contained
particles of porosity 027 The authors measured the averageparticle size for each column using SEM which resulted inparticle sizes of 287 and 302 micrometers for the respectivecolumns The pressure drop for each column was measuredin two different mixtures of Acetonitrile Water therebygenerating two data points for each column Both columnsin this study generated typical values of 04 for externalporosity The results are presented in Figure 3 in their paperIn Table 3 herein the results of the raw measurements arepresented as our Columns J
1and J
2 The values of 119875
0(which
yet again Guiochon and Gritti notate in their paper as 119870119888)
corresponding to the lower particle porosity column (J1) are
234 and 232 while values for the higher particle porositycolumn (J
2) are 289 and 273 The average of these four
measured values is 257Recognizing finally that their data indicates that Car-
manrsquos value of 180 for 1198750may be too low the authors proceed
to challenge their own data Singling out the highest value of1198750 289 the authors comment that such a high value could
be explained by a clogging of the column end frit Howeverthey also report that they adjusted for extra-column effectsin their reported pressure drop values Since this procedurepresumably required measurements in the empty columnsone would think that this would have provided the necessaryadjustments to eliminate the possibility of a clogging issueMoreover even if the highest value for119875
0represents a clogged
frit it is unlikely that the other value for the same columntaken in the other fluidmdashwhich at 273 was almost as highmdashalso represented a clogged frit And then of course there arethe two data points for the other column 234 and 232
Lacking any other explanation for these unexpectedly(from their point of view) high values of 119875
0 they jettison
their own measured values for particle size and instead optfor the manufacturersrsquo lower values of 27 micrometers Thislower particle size of course results in the lower calculatedvalues for 119875
0 As reflected in Figure 3 of their paper they
are now able to report values for 1198750of 231 and 218 for the
higher particle porosity column and 207 and 206 for the lowerparticle porosity column
As can be seen from Table 3 herein when permeabilitymeasurements are used to determine particle size a singlevalue for 119875
0of 267 for all four data points defines a particle
size range of 290ndash308 micrometers which is almost exactlywhat the authors actually measured by SEM Before herejected his own SEMmeasurements Guiochon obtained anaverage value of 257 for 119875
0 Suffice it to say that his value is
significantly closer to the expected (by us) value of 267 thanit is to 180 Moreover if one makes a reasonable allowance forexperimental error one could say that Guiochonrsquos 2011 studyessentially confirms that the value of 119875
0is 267 and accord-
ingly he confirms our permeability frame of reference12
10 Conclusions
Giddingsrsquo teaching of column permeability in columnspacked with fully porous particles is based upon a calibration
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
20 Journal of Materials
derived from nonporous particles which has been titrated forporous particles based upon independently derived data forparticle porosity via his Φ parameter whereas Guiochonrsquosteaching is not This is the fundamental reason for thediscrepancy between their respective teachings on columnpermeability Guiochonrsquos textbook teaching is based upon aterm derived from the chromatographic literaturemdashthe flowresistance parameter 120601mdashwhich has its roots in thermody-namic theory rather than hydrodynamic theory In additionthe fact that Guiochon based his worked example on particleshaving an irregular morphology (119896
0= 10 times 10minus3) without
specifying the degree of irregularity of the particles embed-ded in his hypothetical value for column pressure ratherthan basing it upon smooth spherical particles where thecharacteristic dimension of the particle is well-defined andcontains no ambiguity with regard to particle morphologyalmost makes it appear as if he was deliberately leavinghimself some wiggle room in his teaching
But there can be no wiggle room in the value of 1198750
To begin with as modified by Carman to accommodateparticles with an irregular shape the Kozeny-Blake equationspecifically accounts for particle morphology within therealm of streamline flow Moreover since the relationshipin the equation between pressure gradient and fluid velocitydepends to a first approximation on the fourth powerof the column external porosity the value of 119875
0must be
both accurate and precise This degree of accuracy andprecision however can only be realized experimentally whenall variables are accurately measured in which case thevalues for column pressure embedded in the flow resistanceparameter 120601 on the one hand and the value for column totalporosity embedded in the porosity dependence term Ψ onthe other hand are self-calibrating13
Moreover Guiochonrsquos occasional (albeit apparentlyunwitting) publication of the value of the modifiedpermeability coefficient 119875
119905 as if it were the value of 119875
0
has only exacerbated the confusion surrounding the valueof 119875
0 As has been demonstrated herein experimentally
derived values of 180 and 150 for 119875119905are not infrequent for
columns packed with porous particles and accordinglymay unfortunately be confused with the values of Carman(119875
0= 180) and Ergun (119875
0= 150) [28]14
As detailed above a recurrent theme in Guiochonrsquosexperimental studies of packed bed permeability over thelast decade or so has been that Carman was right 119875
0is a
constant and its value is 180 Accordingly when the resultsof his experiments have not supported this theme Guiochonand his coauthors have seemingly gone out of their wayto attempt to produce results that are consistent with hispredetermined worldview The devices by which this hasbeen pursued include (a) ignoringmeasured particle size datain favor of the particle manufacturerrsquos stated size (b) notmeasuring particle size at all and back-calculating particlesize from permeability measurements where a value for 119875
0
of 180 has been plugged into the Kozeny-Blake equation as agiven and (c) hypothesizing unrealistic explanations for thediscrepancy such as the existence of a ldquoclogged end fritrdquo
On the other hand facts being the stubborn thingsthat they are Guiochonrsquos efforts to make the results of his
experiments fall into line have not infrequently failed to meetwith success
However instead of considering the obvious possibilitythat Kozeny-Blake is valid and that 119875
0is indeed a constant
but that its value is not 180 Guiochon has hedged his betsby reverting back to the ambiguity (and safety) of DarcyismFor example even after Guiochon does a complete flip-flopin his review paper from his textbook teaching andmdashwithoutany explanation or analysismdashsigns on to the conventionalwisdom that the value of119875
0is 180 as if taking one step forward
and two steps backward he proceeds to announce ldquothere issome uncertainty on the value of the numerical coefficientbut since few authors report column permeability data asystematic study has not been made yetrdquo [14 p 9]
Likewise in the second of their 2010 papers which wereviewed above Guiochon and his frequent coauthor Grittimake this illuminating statement ldquothe external porosity 120576
119890
of packed beds is an important characteristic of HPLCcolumns because it affects the column permeability whichis essentially controlled by the average particle size 119889
119901 The
shape of the packed particles their size distribution andthe chemical nature of their external surface play also asignificant role in the column permeability but their impactis more difficult to assess Their effect is empirically lumpedinto the Kozeny-Carman constant119870
119888rdquo [21 p 1487] (emphasis
supplied) Suffice it to say that this assertion that 1198750is
nothing but a hodge-podge of unidentified variables is totallyinconsistent with the notion that it is a constant with a fixedvalue of 180
Ironically Guiochon has essentially ignored his ownstudy published in the 1999 paper with Farkas and Zhongwhich represents one of the most credible studies of columnpermeability available in the literature and which contradictsboth extremes of his pronouncements on column perme-ability (a) there is a constant and its value is 180 and(b) there is only a residual permeability coefficient whichis a variable number that one must plug in to balancethe two sides of the pressureflow equation Indeed it isour conclusion that Guiochon has actually ignored all ofhis experimental investigations which when appropriatelyunderstood uniformly corroborateGiddingsrsquo conclusion thatthe value of 119875
0is 267 a conclusion which was implicit in the
first edition of his textbook published in 1965 and becameexplicit in the second edition published in 1991 [29 p 65]
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper
Acknowledgment
Sincere gratitude is expressed to Tivadar Farkas for providingthe raw measurements underlying the study reported inGuiochonrsquos 1999 paper of which Farkas was a coauthor
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
Journal of Materials 21
Endnotes
1 Endnotes The terminology ldquoKozeny-Carman equationrdquois not favored by the current authors This is because itrepresents the Kozeny-Blake equation with a value for1198750of 180 which is attributed to Carman and which is
believed by us to be in error Accordingly the terminol-ogy ldquoKozeny-Blake equation (as modified by Carman)rdquois preferred and represents Carmanrsquos contribution to theequation to accommodate nonspherical particles whichis a legitimate modification
2 For an in-depth review of Giddingsrsquo development of thisparameter see [1] herein
3 If the column length is in cm the particle diameter in120583m the viscosity in cP and the pressure drop in psi ourequation becomes
Δ119875 (psi) =15119906 (cms) 120578 (cP) 119871 (cm)
1198960119889119901
2(120583m2)
(587c)
As an example of calculation assume the followingvalues linear velocity 010 cms viscosity 10 cP columnlength 15 cm particle size 10 120583m and permeability con-stant 1 times 10minus3 The pressure drop is
Δ119875 =15 [010 cms] [10 cP] [15 cm]
[10 times 10minus3] [102]= 225 psi (587d)
Note that we have corrected two obvious typographicalerrors in the worked example (1) the value of oneparameter used in (587d) compared to the statement ofthat value in Guiochonrsquos text (15 instead of 25 cm for thelength of the column) and (2) the value of the calculatedpressure drop (225 instead of 325 psi)
4 The data in Tables 1 and 2 both of which are referencecharts are based upon a column of the length (15 cm)packed with particles of the size (10 micrometers)with fluid of the viscosity (10 cP) and running at themobile phase velocity (10 cmsec) specified in Guio-chonrsquos worked example
5 This was the technique used by Giddings to establish thevalue of 267 for 119875
0which is implicit in Table 53-1 in his
1965 text [1] and [8 p 209]6 It also leads to an apparent value for Guiochonrsquos coeffi-
cient 119875119905of 178 Note the coincidental and unfortunately
confusing agreement between this value and Carmanrsquosmuch touted value of 180 for 119875
0[6]
7 It also generates a value of 148 for 119875119905 Again note the
confusing coincidence between this value and the valueof 150 which is also frequently reported in the literatureas being the value of 119875
0[10 11]
8 Neue uses the terminology of 1000 and 1 times 10minus3 inter-changeably apparently in his handbook This practiceis sometimes used throughout the literature as theldquoconstantrdquo in the Kozeny-Blake equationmay be thoughtof as either a reciprocal coefficient or a coefficientof direct proportionality depending on the authorrsquos
accepted convention For instance this is whywe refer toGuiochonrsquos use of his term 119896
0as a ldquoreciprocalrdquo coefficient
9 Additionally since the authors did not have firsthandknowledge of the value of either the particle size or thetube inner diameter their conclusion that ldquothe residualexplanation is that the interstitial velocity distributionbetween the packed particles depends on the chemicalnature of the external surface of these particlesrdquo isunjustified
10 On the other hand it is widely accepted that at leastwhen practiced with the currently available less thanadequate solute standards the ISEC technique does notprovide an accurate differentiation between pores withinand without the particle fraction of a chromatographiccolumn containing fully porous particles Beginningaround 2007 Guiochon compounded this problem bychanging his method of interpreting ISEC curves Theconventional method had been to use the intersectionof both branches of the ISEC curve [12] and [10 p 143]Guiochon is now extrapolating a single branch to zeroelution volume In addition he is now using ldquoholduprdquovolumes measured on a gravimetric basis to establishvalues for total column porosity Both of these practicesmay be contributing to even lower external porosityvalues for fully porous particles from those that wouldbe obtained by using traditional ISEC protocols
11 Even more surprising the authors reference Giddingsrsquo1965 text as authority for their chosen value of 180 for1198750 This is a clearly erroneous citation of Giddingsrsquo 1965
teaching on permeability While Giddings stated in his1965 text that themost commonly referenced value for119875
0
in theKozeny-Blake equationwas indeed 180 he rejectedthis value in favor of a much higher value for 119875
0 He did
this by using his 120601rsquo parameter in his own permeabilityequation thereby establishing that his measured valueswere in complete disagreement with Carmanrsquos valueMoreover he also stated that Carmanrsquos discrepancy hadalso been identified by Giddings [8 p 208]
12 Our discussion of Guiochonrsquos publications ends withthis 2011 paper The present authors have decided notto critique each and every Guiochon paper ever since2011mdashof which there have been severalmdashin which Guio-chon and his collaborators state a value for 119875
0(or as
he calls it 119870119888) However that is due to the following
decisions by the present authors (a) enough is enough(b) many of the assertions in those papers of valuesfor what should be measured parameters are apparentlyassumed not measured and are in general opaque toand indeterminable by the reader and (c) these paperslike their recent predecessors claim in the introductorynarrative that the value of 119875
0is 180 but then go on to
record supposedly different experimental values for 1198750
without ever reconciling the two13 On the other hand even carefully constructed and
controlled experiments can take us only so far We haveused 267 in this paper as the value of the constant inthe Kozeny-Blake equation That is the number that we
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
22 Journal of Materials
calculated from the results which Giddings obtainedfromhis experiments conducted in 1965 over forty yearsago [1] Giddings himself in the 1991 edition of histextbook rounded his results off at the figure of 270[29 p 65] We have no doubt that the exact value ofthe constant is somewhere between 265 and 270 andtherefore we feel very comfortable in using 267 whichis the average of those two values in our analysis ofGuiochonrsquos teaching and his recent experimental workHowever a definitive determination of the constantrsquosvalue will have to await its theoretical development fromfirst principles something which has never been done
14 It is also possible that this type of mistaken identity mayhave infected the work of some of the other contributorsto the hydrodynamic literature
References
[1] H M Quinn ldquoReconciliation of packed column permeabilitydatamdashpart 1 the teaching of Giddings revisitedrdquo Special Topicsamp Reviews in Porous Media vol 1 no 1 pp 79ndash86 2010
[2] H M Quinn and J J Takarewski ldquoHigh performance liq-uid chromatography method and apparatusrdquo US Patent no5772874 1998
[3] F Gritti and G Guiochon ldquoUltra high pressure liquid chro-matography Column permeability and changes of the eluentpropertiesrdquo Journal of Chromatography A vol 1187 no 1-2 pp165ndash179 2008
[4] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[5] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 1994
[6] P C Carman ldquoFluid flow through granular bedsrdquo Transactionsof the Institution of Chemical Engineers vol 15 pp 155ndash166 1937
[7] L R Snyder and J J Kirkland Introduction to Modern LiquidChromaTography John Wiley amp Sons 2nd edition 1979
[8] J C Giddings Dynamics of Chromatography Part I Principlesand Theory Marcel Dekker New York NY USA 1965
[9] G Guiochon S G Shirazi and A M Katti Fundamentals ofPreparative and Nonlinear Chromatography Academic PressBoston Mass USA 2nd edition 2006
[10] S Ergun ldquoFluid flow through packed columnsrdquo ChemicalEngineering Progress vol 48 pp 89ndash94 1952
[11] R B Bird W E Stewart and E N Lightfoot TransportPhenomena John Wiley amp Sons 2002
[12] H Guan and G Guiochon ldquoStudy of physico-chemical prop-erties of some packing materials I Measurements of theexternal porosity of packed columns by inverse size-exclusionchromatographyrdquo Journal of Chromatography A vol 731 no 1-2 pp 27ndash40 1996
[13] G Guiochon ldquoFlow of gases in porous media Problems raisedby the operation of gas chromatography columnsrdquo Chromato-graphic Reviews vol 8 pp 1ndash47 1966
[14] G Guiochon ldquoThe limits of the separation power of unidimen-sional column liquid chromatographyrdquo Journal of Chromatog-raphy A vol 1126 no 1-2 pp 6ndash49 2006
[15] U Neue HPLC Columns-Theory Technology and PracticeWiley-VCH 1997
[16] I Halasz R Endele and J Asshauer ldquoUltimate limits in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 112 no C pp 37ndash60 1975
[17] R Endele I Halz and K Unger ldquoInfluence of the particle size(5ndash35 120583m) of spherical silica on column efficiencies in high-pressure liquid chromatographyrdquo Journal of Chromatography Avol 99 pp 377ndash393 1974
[18] I Halasz and M Naefe ldquoInfluence of column parameterson peak broadening in high pressure liquid chromatographyrdquoAnalytical Chemistry vol 44 no 1 pp 76ndash84 1972
[19] J-H Koh and G Guiochon ldquoEffect of the column lengthon the characteristics of the packed bed and the columnefficiency in a dynamic axial compression columnrdquo Journal ofChromatography A vol 796 no 1 pp 41ndash57 1998
[20] T Farkas G Zhong and G Guiochon ldquoValidity of Darcyrsquoslaw at low flow-rates in liquid chromatographyrdquo Journal ofChromatography A vol 849 no 1 pp 35ndash43 1999
[21] F Gritti and G Guiochon ldquoExperimental evidence of theinfluence of the surface chemistry of the packingmaterial on thecolumn pressure drop in reverse-phase liquid chromatographyrdquoJournal of Chromatography A vol 1136 no 2 pp 192ndash201 2006
[22] F Gritti and G Guiochon ldquoAdsorptionmechanism in reversed-phase liquid chromatography Effect of the surface coverage of amonomeric C18-silica stationary phaserdquo Journal of Chromatog-raphy A vol 1115 no 1-2 pp 142ndash163 2006
[23] F Gritti A Cavazzini N Marchetti and G Guiochon ldquoCom-parison between the efficiencies of columns packed with fullyand partially porous C18-bonded silica materialsrdquo Journal ofChromatography A vol 1157 no 1-2 pp 289ndash303 2007
[24] F Gritti I Leonardis D Shock P Stevenson A Shalliker andG Guiochon ldquoPerformance of columns packed with the newshell particles Kinetex-C18rdquo Journal of Chromatography A vol1217 no 10 pp 1589ndash1603 2010
[25] F Gritti and G Guiochon ldquoMass transfer resistance in narrow-bore columns packedwith 17120583mparticles in very high pressureliquid chromatographyrdquo Journal of Chromatography A vol 1217no 31 pp 5069ndash5083 2010
[26] F Gritti and G Guiochon ldquoPerformance of new prototypepacked columns for very high pressure liquid chromatographyrdquoJournal of Chromatography A vol 1217 no 9 pp 1485ndash14952010
[27] F Gritti and G Guiochon ldquoThe mass transfer kinetics incolumns packed with Halo-ES shell particlesrdquo Journal of Chro-matography A vol 1218 no 7 pp 907ndash921 2011
[28] M Rhodes Introduction to Particle Technology John Wiley ampSons 1998
[29] J C Giddings Unified Separation Science John Wiley amp Sons1991
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials
Submit your manuscripts athttpwwwhindawicom
ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CorrosionInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Polymer ScienceInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CeramicsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CompositesJournal of
NanoparticlesJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
Biomaterials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NanoscienceJournal of
TextilesHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Journal of
NanotechnologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal of
CrystallographyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CoatingsJournal of
Advances in
Materials Science and EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Smart Materials Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
MetallurgyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
BioMed Research International
MaterialsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Nano
materials
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Journal ofNanomaterials