I '

Itemixing of Aqueous Polymer Two-Phase Systems?'in Low Gravity —i

S. BAMBERGER '

Department of NeurologyOregon Health Sciences UniversityPortland, Oregon 97201

J. M. VAN ALSTINE

Space Science Laboratory'"'NASA/Marshall Space Flight

Center, Alabama 35812

J. M. HARRIS

J. K. BAIRD

Department of ChemistryUniversity of AlabamaHuntsville, Alabama 35812

J. BOYCE

Dept. of PathologyUniversity of British ColumbiaVancouver, Canada V6T 1VJ5

R. S. SNYDER

Space Science LaboratoryNASA/Marshal l Space- F l igh t

Center , Alabama 35812

D. E. BROOKS

Departments of Pathologyand Chemistry

Univers i ty of Br i t i sh Co lumbiaVancouver, Canada VST 1W5

andDepartment of NeurologyOregon Health Sciences UniversityPortland, Oregon 97201

( K A S A - X f i - 1 0 1 1 4 1 ) D E H I X I K G Cf 2QOECDSIC1YKEE THC-PEfl£E £I£lf.£S. IK LCi G B f l V I l X(SASA) 26 £ CSCL 22A

OnclasG3/29

https://ntrs.nasa.gov/search.jsp?R=19880015425 2020-04-04T16:11:12+00:00Z

Abstract

When polymers such as dextran and poly(ethylene glycol) aremixed in aqueous solution biphasic systems often form. On Earththe emulsion formed by mixing the phases rapidly demixes becauseof phase density differences. Biological materials can bepuri f ied by selective partitioning between the phases. In thecase of cells and other particulates the efficiency of theseseparations appears to be somewhat compromised by the demixingprocess. To modify this process and to evaluate the potential oftwo-phase partitioning in space, experiments on the e f fec t s ofgravity on phase emulsion demixing were undertaken. The behaviorof phase systems with essentially identical phase densit ies wasstudied at one-g and during low-g parabolic aircraf t maneuvers .The results indicate that demixing can occur rather rapidly inspace, although more slowly than on Earth. We have examined thedemixing process from a theoretical s tandpoint by applying thetheory of Ostwald ripening. ' This theory predicts demixing ratesmany orders of magni tude lower than observed. Other possibledemixing mechanisms are considered.

INTRODUCTION

When pairs of polymers are dissolved in aqueous solution two

immiscible liquid phases often form. A common such pair is

dextrane (a polymer of glucose un i t s ) and polyCethlyene glycol)

( P E G ) , which form a PEG-rich phase floating on top of dextran-

rich phase, both phases containing primarily water ( i .e . , 90%

w / w ) . An important use of these phase systems is in p u r i f y i n g

biological materials by par t i t ioning between one or both phases

and the phase interface (1-6) .

On Ear th , an emulsion formed by m i x i n g the polymer phases

rapidly demixes because of d i f f e r e n c e s in phase dens i ty ( 1 , 2 ) .

The demixing process seems to occur in two steps: in the f irs t

step, droplets coalesce to form mill imeter-sized phase regions,

which in the second step, undergo rapid, gravi ty- induced .•jy

convective streaming to produce two cleanly separated phas? i fe .

Recent work on the mechanism of cell par t i t ion indicates tha t the

rapid f lu id flows that accompany phase demixing greatly reduce

cell separation e f f i c i ency ( 2 , 7 , 8 ) .

A major goal of our research is to carry out two-phase

par t i t ion experiments in the low gravi ty (g) environment of

space. Low-g is expected to present the advantages of reduced

cell sedimenta t ion and reduced rate of phase emulsion d e n i x i n g ,

thus providing an o p p o r t u n i t y to carry out un icue cell

p u r i f i c a t i o n s (7-11).

This manuscr ip t examines phase demixing when g r a v i t a t i o n a l

e f f e c t s are .small. Five mechanisms which can po ten t i a l ly

contr ibute oo demixing of the aqueous polymer two-phase syst'vns

in low-g are: Ostwald ripening (12-15), wall wetting (16-17),

nucleation ( . -j, spinodal decomposition ( ; )f and coalescence

(18-21). The Ostwald ripening mechanism, which has been used

successfully to explain the growth of large crystals by transport

of material from small crystals, is considered in detail in the

present manuscript. A heuristic der ivat ion of Ostwald r ipening

theory is given, and the resulting equations are used to predict

the rate at which large phase droplets will grow at the expense

of smaller droplets.

To test the predictions from Ostwald r ipening theory we have

measured the rate of phase demixing at one-g of isopycnic phase

systems ( i . e . , systems with i n s i g n i f i c a n t phase densi ty

d i f f e r e n c e s ) , and we have also measured the rate of phase

demixing of nonisopycnic phase systems during low-g parabolic

maneuvers in a NASA KC-135 a i r c r a f t . r

MATERIALS AND METHODS

Two-Phase Systems

The preparation and physicochemical character izat ion of

aqueous polymer two-phase systems has been described previously

(3-11). Isopycnic dextran-ficoll two-phase systems were prepared

following the description of their existence by Albertsson ( 1 ) .

Small amounts of PEG were added to the systems to achieve equal

phase density at lower total polymer concent ra t ion . All two-

phase systems were prepared by mixing appropriate we igh t s of the

fol lowing aqueous stock solutions: 30% ( w / w ) dextran (D)

T500 (M" = 461 ,700 , M = 181,700, Pharmacia Fine Chemica l s , lotw n

3447), 30% (w/w) DT40 (M = 40,000, M = 33,000, lot Fl-18974),w n

30% ficoll (F) 400 (M = 400,000, lot 7867), 30% PEG 8000 (M =w w

6650, Union Carbide, lot 3529-9104), 0.60 M sodium cloridr^ 0.24r

•A sodium phosphate b u f f e r pH 7.5. All reagents we .TO ^Cn gr'ade or

better from various sources.

Dextran concentrations were determined polar imetr ical ly and

PEG concentrations were determined by gravimetric and r e f r a c t i v e

index measurements ( 3 , 1 1 ) . Once compounded the systems were

fi l tered ( 0 . 4 5 yin f i l te r , Sybron Malge, I nc . ) and .allowed to

equil ibrate for 24 hours at the operating t empera tu re .

Phase systems were physica l ly chacac ta r ized (Table T) by

measu r ing : (a ) b u l k phase electrostat ic po ten t i a l s (dex. t ran- r i i :h

phase negat ive) generated by the asymmetr ic p a r t i t i o n of inns

between the phases ( 3 , 1 1 , 2 4 ) ; (b) in terfacia l tensions, us ing i:he

rotating drop method ( 2 2 ) ; (c) phase diagram tie-line lengths

(11) ; (d) phase dens i t ies and (e) phase viscosities.

(11,22,24). . In order to obtain these- measurements isopycnic

systems were al-tered by increasing or decreasing the PEG

concentration 0 .3% w/w from the isopycnic values, in order to

float one phase on top of the other, and interpolating values for

the isopycnic compositions. The data given in Table T represent

the averages of at least ten independent determinat ions .

Phase systems are designated, according to convention

(3 ,11) , as ( a , b , c ) d where a, b, and c refer to the % ( w / w )

concentration of dextran T500, PEG and f icol l , respect ively, in

the system. Systems containing dextran T40 are designated w i t h

an asterisk: e.g., ( a * , b , c ) d . Parameter d refers to b u f f e r

composition: I = 109 mM Na2HPO4 , 35 mM N a H 2 P O 4 / pH 7 . 2 ; V = 150

mft NaCl, 7.3 mM Na2HP04 , 2.3 mM NaH 2 PO 4 , pH 7.2; and VI = 85 mM

Na2HPO4 , 25 mM NaH2PO4 , pH 7.5.

In order to better visualize phase demix ing , the PFO-|8icoll-

rich phases were dyed by including (0 .05 mg/ml) Trypan m.uc-1

(Aldr ich) in the systems.

Unit-g and Low—g Emulsion Deraixing Experiments

One-g and low-g emulsion demixing exper iments were

under taken using an apparatus (F igu re 1) consist ing of a N i k o n F3

camera wi th Kodacolor ASA. 400 E i lm and a 55 mm, 1.28 N i k o n

Macrolens anchored to a p la t form holding a removable phonc-'no:!'! I •'

which was backli t by a f luorescent l igh t .box . The clear

plexiglass module contained four chambers 30 mm h igh , 8 mm v/ ide

and 6 mm thick ( l igh t path axis) and a digi ta l clock. Rach

chamber contained a 5 mm diameter stainless steel ball to

facilitiate complete emulsification of the phases when the module

was shaken manually. Usually the air-liquid interface aids in

emulsification upon shaking, but in low-g and control demixing

chambers air spaces were eliminated.

An experiment involved mixing the phases by quickly

inverting the module 20 times, starting the t imer, localizing the

module on the experimental p la t form and taking 10 pictures at l-

to 2-second intervals (Figures 2 - 4 ) .

Low gravity experiments were performed aboard NASA KC-135

aircraft . The a i rcraf t flew a series of parabolas in which the

experiments and personnel on board experienced up to 30 seconds

of low-g (Figure 1 ) ( 8 , 2 5 ) . The acceleration experienced by tho

parti t ion module was monitored by three accelerometers (Sunds t r ad

Data Controls, Model 300A1) mounted to the experimental p la t fo rm

on orthogonal axes, one of which was parallel to the samp, ijj}

vertical axis and two of which were parallel to the a i rc ra ' f t s

major axes. For a typical maneuver the low-g accleration on all

axes averaged less than 10~2 g.

Demixing experiments with isopycnic systems at one-g

(Figures 3 , 4 ) were conducted under condit ions similar to those

described above using plexiglass chambers 50 mm h igh , 10 mm wii.lo,

and 10, 7.5, 5, or 2.5 mm thick. Mix ing was accomplished by

.manual ly s t i r r ing the phase system wi th a plastic rod.

Quantitation and Demixing Process

Slides taken of demixing phase systems at defined time

intervals were projected, the outlines of connected domains

traced on paper, and the area of each domain estimated using an

IBM-PC computer equipped wi th a Microsoft Mouse mechanically

fixed to a single or ientat ion. A program was developed which

used the screen cursor movement , created by tracing the domain

outlines wi th the Mouse, to d ig i t i ze and store the pictures of

the areas as polygons. The domain area was calculated by adding

the areas of the trapezoids formed from the i nd iv idua l sides of

the polygon and the project ion of each side onto the hor izonta l

axis.

An example of the result of each stage is shown in Figure 5

where a dis t r ibut ion of the drop areas measured is also j>

provided. The mean and standard dev i a t i on of each such ^

histrogram was calculated, the radius of a circle of area equal

to the mean area determined, and the results .expressed as a

func t i on of time on a log-log plot. These plots were then

subjected to linear regression analysis (Table I I ) ; the line of

best f i t is illustrated ( F i g . 6) .

RESULTS AND DISCUSSION

Ostwald Ripening

When two immiscible l iquids of d i f f e r en t density are

emulsif ied, a mixture of droplets forms. In the presence of a

gravitational f ield, these droplets will agglomerate by

sedimentation and the emulsion will demix. Tf, however, the two

liquids are of identical density, or if there is no sensible

gravitat ional f ield act ing, sedimentat ion plays no cole. Ra the r ,

the two emulsified liquid phases might be expected to demix in

order to min imize the surface area of contact between them and to

preferential ly localize one phase next to the container wall

( 1 6 ) . In the following discussion we treat the dynamics of th i s

separation process as an example of Ostwald ripening. This

concept has been successfully used to explain how large crystals

%)grow at the expense of 3ma.ll crystals ( 1 5 ) . ;,7a have .-leriried the

equations required to apply this same concept to the transport of

materials from microscopic regions to large droplets of one

polymer-enriched phase suspended in the other.

Consider a chemical substance A, which is partitioned

between a submicroscopic droplet of radius R and an exterior

phase. The concentration of A just outside the droplet is

C A ( R ) . In the l imi t that the two phases are separated by a f l a t

s u r f a c e ( i n f i n i t e radius of c u r v a t u r e ) , the equ i l i b r ium

concent ra t ion is C ^ ( c o ) . The two concentrat ions are related by

the Gibbs-Kelvin equat ion:

C A ( R ) = C A ( - ) exp (2 0 v A /Rk t ) (1)

csMS

a is the interfacial tension between the phases, VA is the umo

lar volume of A, k is Bol tzmann 's constant, and T is the Din

te temperature. C A (« ) is the thermodynamic concentration cnn

n the exterior phase given by the appropriate phase Rm

m. Equation 1 shows that the concentration of A in the uqh

or phase outside a droplet of radius R is always greater

: A ( « ) . . t-.he

Jhen two phases, initially having a flat surface of contact 2

»n them, are shaken, droplets wi th a d is t r ibut ion of sizes the

Immediately after mixing C A ( R ) = CA^ ^or a^ R < OT/ but "* "-(1

Dwn in Eq. 1, at equilibrium C A ( R ) > C A ( » ) for all R < », so

1 d i f f u s e from the droplets to the exterior phase. That is

y, these regions begin to shr ink. This causes the i\

ntration of A in the exterior phase to rise. At some po in t , v|t

oncentrat ion of A reaches the value C ^ ( R n ) which is the ) lot

ntrat ion of A in equil ibrium with a droplet of radius Rn ,

R^ is the radius of the largest droplet in the d i s t r ibu t ion ?n

n represents a distribution of smaller droplets. When the

ntra t ion of A in the exterior phase around the small droplet

dius Rn exceeds C A (R T n ) the largest droplet begins to grow by

tion of A. Droplets of radi i smaller than Rjn con t inue to T

k. As the time t approaches », droplets wi th radii s-.na.ller

R,^ d isappear , leaving behind the largest region or d rop l e t , r

radius has grown to the point that it approaches fn

ibr ium w i t h bhe f lat sur face concentration C A ( ° » ) . tho

Sn

of A. This may be because the concentration of B is so much

higher or because the diffusion coefficient of B is much greater

than that of A. In either case, the physical result is that when

a molecule of A reaches the surface of the droplet by diffusion

through the exterior phase, there is already present at the

surface enough B so as to maintain the correct equilibrium

composition of the droplet upon incorporation of an additional

molecule of A. If the mole fractions of A and B in the droplet

are XA and XQ, respectively, then upon incorporation of XA moles

of A, XQ moles of 8 are also incorporated. If the molecular

volumes are VA and VQ, respectively, the average -.nolecular volume

in the droplet is V where V = XAVA "*" XBVB* The v°lume °f tne

droplet is proportional to Rm , so that the number of molecules

1 -"-1

in the droplet is R^ /V. If the rate of incorporation of;|

molecules into the droplet by d i f f u s i o n is ^R^C^R^) '- C A ( » ) ) ,

then the rate growth of the number of molecules in the droplet is

given by

l /V(dRm3 /d t ) - D A R m ( C A ( R m ) ~ C A < - » ( 2 )

w h i c h is an equation of motion for P^ taken as a func t ion of

time. To integrate Eq. 2 we f i r s t evaluate Eq, 1 at R = R^ and

expand the exponent ia l Cor the case (2av , /R kT) « 1. The cosul t• A ftl

is

20v./R k T ) ( 3 )rt HI

°RECEDING PAGE BLANK NOT FILMED

10

Ignoring the factor of 2 in Eq. 3, we subsitute this result Into

the right hand side of Eq. 2 to obtain

l /V(dRm3 /dt)

( 4 )

By cancelling common factors, this equation can be rewri t ten

dR 3/dt = D . C . ( » ) 0 V v . / k T (5)m A A A

Since R^t ) at large t is much larger than R ^ O ) , we, may taken

R m ( 0 ) = 0 and integrate Eq. 5 directly to obtain

R m3 ( t ) « D , C . ( « ) C T V v a / k T . t (6)

Hi A rV f\

^which shows that Rm (t) increases l inearly with the t ime t.

In the rigorous derivation which this derivation replaces,

R m ( t ) is replaced by R ( t ) , the radius of the average droplet in

the distribution. The maximum radius, R^t ) , is just 50% larqor,

a difference which is unimportant in a heuristic derivation.

Also, in the rigorous der ivat ion, the right hand side of Eq. f»

appears mult ipled by a factor of B/9, which is again of the or-ler

of uni ty ( 1 5 ) .

Our results may be summarized by saying that as the smal l .

phase regions dissolve, we are left wi th a single droplet .

According to Eq. 6, the radius, R ^ t ) , of this droplet grows

asymptotically with time as t1/3, while the volume of the drople t

grows linearly with t.

. ' • ORIGINAL PAGE IS . =' :

. OE POOR QUALITH - . . ... "' :

can apply Equation 6 to the specific case of demixing in

ius-polymer phase system, ident ifying dextran as compound

1G as compound B. Dextran of M 185,000 has a molar

if 0.611 cm /gm (2) hence its molecular volume can be

:ed to be 1.83xlO~19 cm3/molecule. Similarly, PEG

550 has a molar volume of 0.333 cm /gm (2) and a molecular ycol)

O 1 "^)f 9.20x10 cm /molecule. We have chosen a typical s

Astern (1) in which the composition of the exterior phase illy

5 dextran (mole f ract ion xdex = 1 .13xlO~ 6 ) , 5 . 7 % PEG (xp E G lel the

L 0 ~ 4 ) , and 9 3 . 2 5 % water (x = 1, V = 2 .99x lO~ 2 3 cm 3 /gm). ; * f t e r

rage molecular volume in the droplet, V, is calculated to L x i n g

cm . For this system we have measured a as 0 .02

D (dex) = DA = 2xlO~ 7 cm2/sec. The concentration of ad i ly

in the exterior phase, Ca ( » ) , is given by XA/V

0 molecules/cm . 'ijf

serting these parameters into Equat ion 6 gives ncn

he

R m3 ( t ) - = (2xlO"2° cm3/sec) t (7) he

ients

g this equat ion we then calculate that 10 weeks will be

3 to grow a droplet one micron in -diameter ( s t a r t ing at

For a droplet to grow to ten microns is calculated to 5

9 years.

conclusion, if Ostwald r ipening is to provide the i.i ad. um

t mechanism by which the two aqueous-polymer phases demix

absence of dens i ty d i f f e r e n c e s , this demixing can be Q C u V -Y

d to take years. re th;in

12

90% into the ficoll-rich phase in this system. Within 3 seconds

following mixing at unit gravity the emulsion demixos into

dispered visible domains of each phase. By 12 seconds,

sedimentation begins and convective streams are present by 25

seconds. This streaming occurs for approximately 30 seconds

during which time clear regions of each phase appear at the top

and bottom of the chamber. As demixing progresses, the middle

emulsion region slowly contracts to form a planar in ter face . In

keeping with sedimentation being the dominant demix ing mechanism,

shortening chamber height decreased the time required for

demixing, while decreasing chamber thickness exh ib i t ed no

effect. Undyed and Trypan-blue dyed systems exibited identicnl

behavior. Demixing of this non-isopycnic system at low-g in KC-

135 aircraft mimicked the behavior of isopycnic systems at u n i t

0gravi ty , even to the extent of early preferent ia l locall iz$>t ion

of PEG-ficoll-rich phase at the container walls. Also, d e m i x i n g

of the isopycnic systems was unchanged at unit-g and low-g.

Figures 5 and 6^ and Table II give the quant i ta t ive results

obtained by the imaging analysis process described in Mater ia ls

and Methods. Note that phase streaming prevents use of th is

procedure in analyzing the demixing of non-isopycnic systems in

unit-g. The data Cit a log-log plot, thus implying a power l.iw

relationship between domain area a n d - t i m e . The log-log plot WMS

less obviously linear in the case of t h e . ( 7 * , 5 , 0 ) V sys tem.

Presumably, this ref lects the limited time scope of the KC-13!5

experiments.

PRECEDING PAGE BLANK NOT

14

The preceding experiments show that aqueous polymer two-

phase systems, which are characterized by moderately high

viscosities and low interfacial tensions, will demix, following

mixing by agitation, in the absence of appreciable gravi ta t ional

influences. The mechanisms by which this demixing occurs are not

known. Calculations based on our approximate treatment of

Ostwald ripening strongly suggest that this mechanism is not of

importance in the present case. Not only does the est imate of

the years required to produce a 10 micron domain wi ldly d i f f e r

from that observed, but measured power law c o e f f i c i e n t s (Table

II) d i f f e r s igni f icant ly from the t ' dependence predic ted. It

does not seem l ikely that either nucleation or spinodal

decomposition are also involved, since the mix ing /demix ing

process takes place at constant temperature and, presumably,

constant phase composition. We have no evidence that ':h s? n i x i n g

process perturbs the composition of the phases from thei r

equil ibrium values. Hence, there seems to be no reason why these

phase separation mechanisms should be relevant .

Coalescence of phase drops in the concentrated emulsions

cer ta inly occurs, and the disposition of the separated phases

appears to be governed by preferent ia l wet t ing of the chamber

walls by the f icol l -PEG-rich phases. If coalescence does provo

to be the dominant mechan i sm, the present data provides a '/>.-. sis

foe testing theories descr ibing the processes involved .

Consequent ly , we suggest that the coalescence process is d r i v e n

by f lu id motion associated with PEG-rich regions or droplets

displacing dextran-rich phase from the wall. "~ - - v.

15

tults are somewhat sirailiar to those of others

KJ the structures of immiscible metal alloys and

2! systems solidified under microgravity condi t ions

). On the other hand, it is interesting to note that

to ( 3 9 ) found that fluorocarbon (Krytox) oils and

d stable (10 hr) emulsions at low-g, on board

986.

ng in

imary, our results suggest that emulsions of the type

re will demix by a d i f f e r e n t mechanism and at a slower

tf-g than on Earth. The f inal disposition of desnixod

3 to some extent the rate of demix ing, can be

by manipu la t ion of chamber geometry and surface

havior. Application of polymer phase systems to spnce

ing is in progress.

CONCLUSIONS

mmary, our results show that emulsions of aqueous

o-phase systems will demix in the absence of

nally-driven convection. The mechanism for this

s unclear , but we propose that preferent ia l we t t inq of

ner walls provides f l u i d movement which leads to

e of droplets. Clearly, Ostwald r ipening for t'n^e

'coo .slow ;;o be an important demix ing m e c h a n i s m .

2959

al l

.an

DBIGINAE PAGE ISOF POOR QUALITY

16

"PAGE MISSING FROM AVAILABLE VERSION"

n

ORIGINAL PAGE ISOF POOR QUALITY

8. J. M. Van Alstine, J. M. Harris, R. S. Snyder, P. A.

Curreri, S. 8. Bamberger, and D. E. Brooks, Ibid, 309

(1986) .

9. D. E. Brooks and S. B. Bamberger, in Materials Processing in

the Reduced Gravi ty Environment of Space {G. Rindone, ed .} ,

Elsevier Amsterdam, 1982, pp. 233-240.

10. R. S. Snyder, in Mater ia ls Sciences in Space (B.

Feuerbacher, H. Hamacher, and R. J. N a u m a n n , e d s . ) ,

Springer-Verlag, New York, 1986, pp. 465-481.

11. S. Bamberger, D. E. Brooks, K. A. Sharp, J. M. Van Als t ine ,

and T. J. V7ebber, in reference 2, Ch. 3, pp. 85-13^1

12. W. Ostwald, Z. Phys. Chem., 34, 495 ( 1 9 0 0 ) .

13. I. M. L i f s h i t z and V. V. Slyozov, J. Phys. Chem. Solids, 19,

35 (1961) .

14. C. Wagner , Z. Electrochem. 65, 581 (1961) .

15. J. K. H a i r d , in Proceedings of the 5 ch European Symp. on

Materials Sci. Under Microgravi ty , European Space Agency,

SP-222, 319 ( 1 9 3 4 ) .

16. J. W. C a h n , J. Chem. Phys., 66, 3667 ( 1 9 7 7 ) .

17. J. W. Cahh, J. Chem. Phys., 42, 93 (1965).

18. J. D. Chen, J. Colloid Interf. Sci., 107f 209 (1985).

19. G . D . M . McKay and S. G. Mason, Can J. Chem. Eng. , 41, 203

(1963) .

20. D. Beysens, in Materials Sciences in Space (B. Feuerbacher,

H. Hamacher, and R. J. Naumann, eds.) , Spr inger-Verlag, New

York , 1986, pp. 191-224.

21. T. K. Kwei and T. T. t tfang, in Polymer Blends, Vol. 1, (0. R.

Paul and S. Neuman, E d s . ) , Academic Press, New York, pp.

141-184. 4f?

22. S. Bamberger, G. V. F. Seaman, K. A. Sharp, and D. E..

Brooks, J.. Colloid In ter f . Sci. 99, 194 ( 1 9 8 4 ) .

23. J. M. Harr is , M. Yalpani , E. C. Struck, M. G. Case, M. S.

Paley, J. M. Van Alstine, and D. E. Brooks, Polymer Sci . ,

Polymer Chem. Ed . , 22, 341 ( 1 9 8 4 ) .

24. D. E. PJ rooks, K. A. Sharp, S. B'a>nbeirger, C. H. T a m b l y n , <;.

V. F. Seannan , and H. Wa l t e r , J. Colloid I n t e r f . Sci. , 102 , 1

( 1 9 8 4 ) .

ORIGINAL PAG® jg2Q OF POOR QUALIQ3

25. R. E. Shurney, in The Marshall Space Flight Center KC-135

Zero Gravity Test Program (R. E. Shurney, ed . ) , NASA T M

82541, NASA-MSFC, Alabama 35812, 1983.

26. A. Menawat, J. Henry Jr., and R. Siriwardane, J. Collod

In t e r f . Sci., 101, 110 ( 1 9 8 4 ) .

27. S. Schurch, D. F. Gerson, and D. J. L. Mclver, Dioch im,

Biophys. Acta 640, 557 (1981) .

28. J. M. Harr i s and M. Yalpani , in re f . 2, pp. 539-626.

29. J. M. Harris, J. Macromol, Sci., Rev. Polym. Chem. Phys. ,

C25, 325 (1985) .

J-?

30. M. Yalpani and D. E. Brooks, .1. Polyn. Sci., Polra. Chem. Sci.

23, 1395 ( 1 9 8 5 ) .

31. B. J. Herren, S. G. Shafer , J. M. Van Als t ine , J. M. Har r i s ,

and R. S. Snyder, J. Colloid In t e r f . Sci., in press ( 1 9 8 6 ) .

32. M. von Sino.luchowski , Physik . Z, 17, 557 (1916) .

33. M. von rvaluehowski , Ib id , 17, 585 ( 1 9 1 5 ) .

34. M. von S:aoluchov.-ski , Z. Phys ik . Chem. , 92, 129 ( 1 9 1 7 ) .

ORIGBVAI; PAGE rs21 OE POOR QUALITY

35. J. K. Baird, Can. J. Chem., 55, 2133 (1977) .

36. D. 0. Frazier, B. R. Facemire, W. P. Kaukler , W. K.

Witherow, and U. Fanning, NASA T M 82579, NASA-MSFC,

Alabama 35812, 1984.

37. S. H. Gelles and A. J. Markwor th , Spar V Final Repor t , NASA

Expt. No. 74-30, NASA-MSFC, Alabama 35812, 1978.

38. C. Potard, Proceedings of Meeting of I n t ' l . Ast ronaut io . i l

F e d ' n . , Sept. 6, 1981.

39. L. L. Lacy and G. H. Otto, Proceedings of AIAA-IAGU

Conference on Scientific Expts. of Skylab, AIAA Paper 74-

1242, MSFC, Alabama 35812, 1974. }

'/

40. J. T. Davies and E. K. R Rideal, in In te r fac ia l P

Academic Press, New York, 1963, pp. 108-153.

22

^ c

<QCO0

4JCO

w0)U)IT!

JSO4

<wocoQ)

•^4-U

U0)aou0,

ft

(8O

1-4CO

JS£Xi

T3

CO

M-t

4J-i-l

CO

ae0u

•M

©r-tJD(0t-*

>, *0JJ Ll•r* JJ

10 — » XC «~4 fl)(U E QO v^

en0) -CO(Q Cuf^ |cu o

taa,

jar-4

03 - —

•-1 o e

°-° z"ii j • H^^ j— » -^i, c- ^

Q)

JJ *"CM

J3 5s1X5 cJj

CT 3 •*•"*cO 5

•*-4 ^X.a g 3

•f— 4CU •) rf*>

cn " ~-03jC .2

H

J2Q) *-m (flW *U

(13 *j- AJ —^04 ^ 1~" - ^^r-< O3 CU

CO

jQ JC>i CJ

-r-l J-l

CO 1

0 Qo0) —

-r* O, JS

> O 0-H

O -J0) 1fO tt,

jC 1cu arj

a,

03 -~C CuO —•- ' 3 ~JJ \. CJ)—I 3 CO

CO CU

a crP~-» ».

E QO ~CJ

00r- <Nm P«0 0

• »rH rH

r p>4oj r^0 0

r-4 r-l

o a\* *^O i"*i

SQ

rJ Z

0 Q0 ^

CN r. .

CN CTlr-l rH

\o ^o<T\ •

• O^T on

> r-l

— ,m.

CT>»

— r-o

om

m

*r- m- ~~r

•oC >r

CJ tO r-lCO CCU r-t 0

.. COo u •n- cu "Of-l M-l 0)

U4 C

C 3 "«03 CD Pu CJJ CUX • -U0) CO CU•C T3 "C* O

U JJ OO cu c

r: uo a0 T3 Sm cE-i us

,-— 10 CUG r-l C

•— ' 03 -r-l

•rH rH

c u <c03 Q) COU JJJj 03 rQ

•x s: cu0 u•c c cu

CO u-;

U -C 3O CU J2S -U>« O OrH C -UO 03Q. OJ SZ

jj Q.(U rH UJ O

jr 03 o —iJJ CO _C C

a. oCT ^ >»c <y co Q.

•H HJ -H O

C 'H CO•H 3 > -H

03 JQJJ <y CUC CU rH Vj

O -C "-1 CUO J-l JT 3

3•"O r-l

CU C CU >cn 03 JJ03 03 U•G *. J3 tyQ.O O.ij-1

o cn u-if ^r O 30 JC JC!'0 -— Qjcy cu cn

— e cUJ 3 -H

O 1— ' -r-< C

r-i "C 'r-l

to O O rg0 U CO JJe -H c3 y-i C OrH -HOO T3 03> C -U! tO

03 C =rH O 0)<3 O O JJ30 COCTO rH >,

CO CO > CO

«3

-H

9

Q.

cn

coJJc

-<H

£ucuJJcy•ccujjo

uoc(J)JJ

cuU-' >

O -iJJ

c — «03 '0CU O 'e a- 0)

cn corrj 03

o .cjc a.jj1) r-s: u

•r-l

C 1fO pu

1co ur-l CC03 (X

• H

-i rHQJ 03JJ -H

05 ^>sr c

cu<-* 4J

•^ oa•o"J <U£ OT••H fTJ'-1 jC^O *^r(/J?} *x

r;" . — 13

CO COfO

TJ 'ij)l XLi3 <*PCO O03 rH

CU™ V

J^

C10o

jj m0) rHcu <ocr ••Hcr- -tJ3 CCO <U

Jj

>>-oCD CUr*

^M-H

(*•

. -o•cr <uca co\^ j^cn curH >

\ CUm u

ij •cu crjj •JJ CD(JJ —*

•— tC7

c c-r-t *f-i

^cs oC/3 CU

£>• CJ

JO

< cu .> cy con c

O > —JJ (0

e wc•U — -J> «r. o

•H . e_iC-.'cn ^

CO •. <cu t^ >3 • 3rH 003 >> » J2

mO • Lijj in cu

~— >U 31) Li O

u-> O Oa1 «j cLi ra

CO >CO CJJJ 3 CO rH -H

o > r:03 3)!-J *» i-'J3 JJ 3

•H yjQ) CO 03Li O CU03 cj e3 COcr--- cucn > J2

..CO

0z

ORIGINAL1 PAGE ISPOOR QUALITY

Table II. Regression Parameters of Phase SystemDemixing Kinetics

System3 Power Law Parameters13

Isopycnic; Ig

( 5 . 5 , 0 . 7 , 9 . 5 ) VI -3.49 1.51 0 . 9 9 2

( 7 , 0 . 3 , 1 2 ) VI -2.13 1.21 0 .978

KC-135;<0 .01 g

( 7 * , 5 , 0 ) II -1.37 1.18 0.391

a Equal volumes of each phase. Phase system composition asnoted in Table I and in Materials and Methods.

log ( rad ius ; ram) = A + 3 log (t; sec); r = correlate .<?«coe f f i c i en t f

24

gends

KC-135 low-g phase demixing experiment.

Droplet growth with time in a ( 5 . 5 , 0 . 7 , 9 . 5 ) VI

isopycnic system containing 0.05 mg/ml Trypan Blue

dye. Chamber dimension; 50 mm x 10 mm x 2.5 nun

( t h i c k ) .

Demixing in a ( 7 , 0 . 3 , 1 2 ) VI isopycnic phase :;y.stem at

d i f f e r e n t times a f t e r mix ing . Chamber d imens ion ; 50

mm x 10 mm x 10, 7 .5, 5 or 2.5 mm.

Unit-g and low-g demixing of (7* , 5 , 0 ) V two-phase

system containing 0 .05 mg/ml Trypan Blue dye.

Quant i ta t ion of demixing in isopycnic system in

Figure 3.

(a) Detail 90 minutes af ter mixing

•\)(b) Tracing of dyed and undyed areas (\

(c) Dis t r ibu t ion of traced areas

Plot of log tine af ter shaking versus log ( e q u i v a l e n t

domain via me te r ) for ( 7 , 5 , 0 ) V system at low-g ( )

and isopycnic ( 5 . 5 , 0 . 7 , 9 . 5 ) .VI system at un i t -q

( 0 ) . Bars respresent the standard d e v i a t i o n for the

individual size d i s t r i b u t i o n . 'IStephan redo ploase)

DRIGSvTAE PAGE ISOF POOR QUALITY

25

ORIGINAL PAGE 13OF POOR QUALITY

inVWj'•\ i i

$11

1&!•?;'

IJ,:.•;#£/v;

r^'.•* '

2 min 5 min

I3 min

»;»;

^

^I-fl4N

7 min

^

Lyi?

!i i

i;l

A rain 10 ciin

r ?

^

20

10

O.2 O.4 0.6Drop Alt* (mm'l

O.B

pb

tob

Log RADIUS (mm)I I

pen

pb

CJin

o

o

Q

N)b

D

D

OKI5INAEOF POOR

n