Talanta 72 (2007) 289295

An automated FTIR method for the routineive

F Roa Mc onald

nadaaire, Qet, La

ober 2er 200

Abstract

An accurate primary Fourier transform infrared (FTIR) method for the determination of moisture in mineral and ester based lubricants hasbeen developed based on the extraction of moisture into dry acetonitrile. FTIR evaluation of acetonitrile extracts from new and used lubricantsas well as common lubricant additives and contaminants which might co-extract indicated that phenolic constituents interfered significantly withmoisture measurements. By measuring moisture at 3676 cm1 on the shoulder of the asymmetric OH stretching band, spectral interferences fromextracted phto dry acetoLubricant saseparate thefor extractioextracts, andregression osystem, thisfor the deter 2006 Else

Keywords: A

1. Introdu

The leveas water iscant contamand excessfied water ivariety of Amethods avand related(KF) metho(WAP), is

CorresponE-mail ad

0039-9140/$doi:10.1016/jenolic constituents were minimized. The spectra of calibration standards (02100 ppm), prepared by gravimetric addition of waternitrile, were recorded in a 1000-m CaF2 transmission flow cell and produced linear standard curves having an S.D. of 20 ppm.mple preparation involved the vigorous shaking (20 min) of a 1:1.5 (w/v) mixture of lubricant and dry acetonitrile, centrifugation tophases, acquisition of the FTIR spectrum of the upper acetonitrile layer, and subtraction of the spectrum of the dry acetonitrile usedn. A Continuous Oil Analyzer and Treatment (COAT) FTIR system was programmed to allow the automated analysis of acetonitrilethe methodology was validated by analyzing 58 new and used oils, independently analyzed by the Karl Fischer (KF) method. Linear

f FTIR versus KF results for these oils produced a linear plot with a between-method S.D. of 80 ppm. As implemented on the COATFTIR method is capable of analyzing 72 acetonitrile extracts/h and provides a high-speed alternative to the KF titrimetric proceduresmination of water in lubricants.vier B.V. All rights reserved.

cetonitrile; FTIR spectroscopy; Lubricants; Moisture content; Karl Fischer

ction

l of moisture in lubricants is an essential parameter,one of the most widespread and destructive lubri-inants. Water contamination can lead to corrosion

ive wear of machinery, and the presence of emulsi-n a lubricant can impair its effectiveness. Among themerican Society for Testing and Materials (ASTM)

ailable for the determination of moisture in lubricantsproducts, the ASTM D6304 coulometric Karl Fischerd [1], performed with a water evaporator accessory

considered the gold standard. The WAP is used to

ding author. Tel.: +1 514 398 8618; fax: +1 514 398 7977.dress: frederik.vandevoort@mcgill.ca (F.R. van de Voort).

rapidly heat the oil sample under vacuum to vaporize water in thesample, which is flushed with dry nitrogen into the KF reagentsolution. This procedure minimizes matrix effects commonlyassociated with lubricant additives, which can undergo aldolcondensation and redox side reactions that can affect KF resultsif oil is directly added to the KF reagent. Thus, ASTM D6304with a WAP is considered the most comprehensive and accurateprocedure available for the analysis of moisture in lubricantsand petroleum products, including lubricant additives, base oils,fully formulated lubricants and automatic transmission fluids,amongst others. Most oil analysis laboratories invest in a KarlFischer apparatus for the quantitative determination of mois-ture along with other equipment for conducting other pertinentanalyses such as particulate contamination, acid number, vis-cosity, ferrography, elemental analysis, oxidation and more, allat substantial cost in terms of equipment and consumables.

see front matter 2006 Elsevier B.V. All rights reserved..talanta.2006.10.042of moisture in lubricants: An alternatrederik R. van de Voort a,, Jacqueline Sedman a,Gill IR Group, Department of Food Science and Agricultural Chemistry, Macd

Sainte-Anne-de-Bellevue, Quebec, Cab Thermal-Lube Inc., 255 Avenue Labrosse, Pointe-Cl

c Naval Engineering Test Establishment, 9401 Wanklyn StreReceived 16 August 2006; received in revised form 16 Oct

Available online 29 Novembquantitative determinationto Karl Fischer titrationbert Cocciardi b, Steve Juneau cCampus of McGill University, 21,111 Lakeshore Road,H9X 3V9uebec, Canada H9R 1A3

Salle, Quebec, Canada H8R 1Z2006; accepted 19 October 20066

290 F.R. van de Voort et al. / Talanta 72 (2007) 289295

Among the instruments often found in commercial oilanalysis laboratories is a Fourier transform infrared (FTIR) spec-trometer. Tmonitoringparameterssoot, glycotion and watherein). Fa semiquanprimary rereported inthe case oflished critedetecting wit does notof water prused to mocontainingthe other hthe spectraeliminatedsame formcontent mausing gravthis methodapplicablebecause ofappropriatebands of wbonding, wThese comtitative FTImethod.

An altercome theseof water wcoupled wicontributiople was deabsorptionaddition ofmoisture lehydraulic, aues for gascalcium careaction ismore H2O.lubricants cas well asthe acetonebonyl absoapproach foils, as theled to the dthe extracttrile followin the spec

oped using acetonitrile-water standards [5]. This paper describesthe adaptation of this spectroscopic approach to the analysis of

d usd as

erim

ater

itiveercianc. (ere bonlys, faes, s

nd amific, N3 Ats ais. Fo, cobtainils wnt (Nlandenter). Fras se

re co

stru

ailss sofB BmetIBM

y Wm f-Clailsomet

singthe

laceys. Tl intoCrysss stt at 3addif 1.0

naly

sam

ed onhis instrumentation is used for lubricant condition, which involves the trending of a wide range of, including oxidation, nitration, sulfate by-products,l contamination, diesel fuel dilution, antiwear deple-ter contamination (see Ref. [2] and references cited

TIR condition monitoring by direct trending [3] istitative procedure, in that no calibration against a

ference method is performed. Instead, results areterms of peak areas or heights (or baseline tilt in

soot) and interpreted in relation to empirically estab-ria. Whereas this methodology provides a means ofater contamination in in-service lubricants over time,allow for quantitative determination of the amountesent owing to overlap of the IR absorption bandsnitor water contamination with those of other OH-constituents (often present in lubricating oils). Onand, in the differential version of this methodology,l contributions of these constituents are, in principle,by subtracting out the spectrum of a new oil of the

ulation as the samples being analyzed, and the watery be estimated from a calibration equation obtainedimetrically prepared standards. However, the use ofology as the basis for the development of a generallyquantitative method is highly problematic not onlythe limitations associated with the need to have thereference oil at hand but also because the absorptionater are subject to matrix effects such as hydrogenhich differential spectroscopy cannot compensate for.plications have stymied the development of a quan-R procedure that would be competitive with the KF

native approach developed in our laboratory to over-limitations made use of the stoichiometric reactionith dimethoxypropane (DMP) to produce acetone,th differential spectroscopy to eliminate the spectralns of the oil [4]. The moisture content of the sam-termined by measuring the intensity of the (C O)of acetone, calibrations being derived by standardwater to a dry oil. This method accurately detects

vels as low as 50 ppm and is workable for compressor,nd gear oils. However, it gives erroneously high val-

oline and diesel engine oils that are over-based withrbonate, because the acid catalyst required for theneutralized by the base, thereby producing CO2 andThis method is also problematic for mineral-basedontaining substantial levels of ester-based additives

ester-based lubricants, as the (C O) absorption ofproduced by the reaction is masked by the ester car-

rption. This problem also prevents the FTIR-DMProm being used for the analysis of moisture in edibley are predominantly triacylglycerols. This limitationevelopment of an alternative FTIR method based onion of moisture from edible oils using dry acetoni-ed by measurement of the water absorption bands

trum and quantification based on a calibration devel-

new an

metho

2. Exp

2.1. M

Addcomm

Lube Ioils wcomm

glycoladditivnolic aScientmeshwith iingresenginewere o

used olishme(Cleveport C(JOAPples wmoistu

2.2. In

Detand itan ABspectroby anprietarPlatforPointepler (Gspectroproces

Forwere ppler tracontrotionalstainlewas se

by co-gain o

2.3. A

Forweighed lubricants and its implementation as an automatedwell as its validation against the KF method.

ental

ials

-free (unformulated) mineral and ester base oils andl additive packages were obtained from Thermal-Pointe-Claire, PQ, Canada). Ester and mineral baselended with additive packages and/or constituentsfound in new or used oils, including esters, polyalkyl-tty acid rust inhibitors, phosphate-based antiwearulfurized olefin extreme pressure additives, and phe-

ine antioxidants. Reagent-grade acetonitrile (Fisherepean, ON, Canada) was dried and kept dry over 48

molecular sieves and dispensed using a re-pipette,r inlet protected by desiccant to prevent moisturermulated commercial lubricants, including diesel

mpressor, hydraulic, gear and other oils (Table 1),ed from a variety of industrial suppliers. Samples ofere provided by the Naval Engineering Test Estab-ETE; Montreal, PQ, Canada), Insight Services Inc.

, OH), and the U.S. Armed Forces Technical Sup-(TSC) of the tri-service Joint Oil Analysis Program

om among these new and used oils, a subset of sam-lected to represent a variety of oil types and differentntents and submitted to NETE for KF analysis.

mentation

of the COAT FTIR Oil Analyzer (shown in Fig. 1)tware have been described elsewhere [6]. Briefly,omem (Quebec City, PQ, Canada) WorkIR FTIRer equipped with a DTGS detector and controlled-compatible Pentium 150-MHz PC running pro-

indows-based UMPIRE PROTM (Universal Methodor InfraRed Evaluation) software (Thermal-Lube,ire, PQ, Canada) is connected to a Gilson autosam-n Inc., Middleton, WI). The software controls theer, autosampler, and pump and handles spectral dataautomatically.moisture analysis procedure, acetonitrile extracts

d in septum-capped vials and loaded into autosam-hey were then sequentially aspirated under computera 1000-m CaF2 transmission flow cell (Interna-

tal Laboratories, Garfield, NJ), using a sharpenedeel needle to penetrate the sealed vials. Pump time0 s/sample. The spectra of the samples were collectedtion of eight scans at a resolution of 8 cm1 with a. Weak Beer-Norton apodization was employed.

tical protocol

ple preparation, 10 g of sample (0.0001 g) wasan analytical balance into a tared 50-mL clinical cen-

F.R. van de Voort et al. / Talanta 72 (2007) 289295 291

Table 1Partial listing of commercial lubricants and the functional categories assigned to them by the manufacturer

Lubricant Category

Rimula X SAE 40 Diesel engine oilMil-L-9000 Diesel engine oilOMD 113 Diesel engine oilMobil Delvac 1240 Diesel engine oilMobil DTE BB Bearing oilRP ThermalGlyde 220 Gear oilMobil DTE 832 Compressor oilHydrex AW 32 Bearing oilMobil Rarus 1024 Compressor oilTerrestic 220 Anti-friction bearing oilTerrestic 68 Turbine oilMobil Gear SHC 460 Gear oilMobil Gear SHC 460 Gear oilMobil Delo 400 Diesel engine oilMobil DTE 26 Hydraulic oilMobil DTE Heavy Medium Turbine oil

trifuge tube, and 15 mL of dry acetonitrile was dispensed into thetube using a re-pipette. The samples were mixed vigorously for20 min using a multi-sample, reciprocal action shaker (EberbachCorporation, Ann Arbor, MI), and then the oil and acetoni-trile phases were separated by centrifugation (Eppendorf CanadaLtd., Mississauga, ON) for 10 min at 10,000 rcf (relative cen-trifugal force). The upper acetonitrile layer was transferred intoa 20-mL autosampler vial, which was then capped with a sep-tum cap having a Mylar liner. The vials were loaded into theautosampler tray from position #3 onward, with the first two

Fig. 1. COATconsists of aautosampler t

vials in thto extract tthe secondbeam specan open-beof each runabsorbancetracted fromdifferentialof ester-bastrile extract

f thered

m1of t.

F an

KFas a

WAPratio o(measu2033 cheightextract

2.4. K

Theysis wwith aFTIR Oil Analyzer used for the analysis of moisture. The systemmicro gear pump (A), cell holder (B), FTIR spectrometer (C),ray (D) and robotic arm (E).

MO) KF revalidated uMilwaukeeing it to 22until the enthan 5g/mto 160 C wto the samefrom the KFTIR mois

2.5. Calibr

Seven wby gravimecover a ranLubricant Category

Mobil Vacuoline 133 Bearing oilMobil Vacuoline 525 Bearing oilMobil Gear 632 Gear oilMobil Rarus 424J Compressor oilTellus T15 Hydraulic oilTellus 68 Hydraulic oilOmala 220(P) Gear oilTurbo T46 Turbine oilTurbo P100 Turbine oilTonna T220 Slide-way oilRimula X 15w/40 Diesel engine oilMorlina 10 Bearing oilRando HD-46(P) Hydraulic oilMeropa 320P Gear oilCosmolubric HF-130 Polyol ester oilHydro FE-56 Polyol ester oil

e tray being reserved for the dry acetonitrile usedhe samples, the first serving as a rinse solvent andto collect the spectrum of the solvent. The single-

tra of the solvent and samples were ratioed againstam background spectrum collected at the beginningto give their respective absorbance spectra, and thespectrum of the dry acetonitrile solvent was sub-the absorbance spectrum of each sample to give its

spectrum. To compensate for the partial miscibilityed oils with acetonitrile, the spectra of their acetoni-s were multiplied (prior to spectral subtraction) by theheight of the acetonitrile (C N) band at 2068 cm1relative to a two-point baseline between 2125 and) in the spectrum of the dry acetonitrile solvent to thehis band in the spectrum of the acetonitrile sample

alysis

apparatus used by the NETE laboratory for oil anal-Mettler Toledo DL32 (Columbus, OH) equippedand used Hydranal (SigmaAldrich Inc., St. Louis,agents for moisture quantification. The system wassing KF moisture standards (SigmaAldrich Inc.,, WI) by loading 0.1 g of each standard and heat-0 C, flushing with dry nitrogen for 60 s and titratingd point was reached, corresponding to a drift of lessin. Lubricant analysis used 1.0 g of sample heatedith a nitrogen flush time of 210 s and then titratedendpoint as the moisture standards. Data obtained

F analyses were used to evaluate the efficacy of theture method.

ation and validation

ater/acetonitrile calibration standards were preparedtrically adding distilled water to dry acetonitrile toge of 3002100 ppm moisture. The FTIR spectrum

292 F.R. van de Voort et al. / Talanta 72 (2007) 289295

of dry acetonitrile was subtracted from the spectra of the stan-dards in a 1:1 ratio to remove the spectral contributions ofacetonitrileamount of3676 cm1differentialdict the moextractionof the metdifferent ty(2001000ing the FTgravimetricreproducib12 new antents conseby calculatAccuracywhich camples and sasamples wtory. The aregressionKF values(SEP).

3. Results

3.1. Gener

Acetoniwater fromspectroscopto be largelubricantsthem. Morin the portpathlengthssis of low lFig. 2 preseof calibratitrile from wout. Owingtonitrile rathese specmetric OHthe single blinearly toto each ot2100 ppm)

In additcapable of ehols, phenobut not ionwide varietundefineda number o

ifferential spectra of water in acetonitrile over a concentration range of0 ppm, illustrating the two main water absorption bands, the symmetricmetric OH stretching bands at 3630 and 3540 cm1, respectively.

le. Such potential sources of spectral interferences werened b

var

ts thed ahe spuenthougaxim

ntialrencdgedectruwith

ed aw/wels aricanof O

lowe es

ifferential spectra of acetonitrile extracts from a wide variety of com-new and used lubricants in relation to a differential spectrum of moisturenitrile, illustrating the spectral variability produced in the OH region.. A linear regression calibration equation relating themoisture added (in ppm) to the absorbance value at(relative to a baseline point at 3738 cm1) in thespectra of the calibration standards was used to pre-isture content of oil samples in order to assess the

recovery, repeatability, reproducibility, and accuracyhod. Percent recovery was evaluated by spiking 10pes of oil samples with increasing amounts of waterppm in200 ppm increments) and assessed by divid-IR-predicted values by their respective theoreticalvalues and multiplying by 100. Repeatability and

ility were evaluated by running three replicates ofd used oil samples having high and low water con-cutively and on 3 different days and were assesseding the average standard deviation of the replicates.was evaluated using 58 validation samples, 23 ofe from JOAP with the balance being commercial sam-mples prepared in our laboratory. All these validationere analyzed by KF titration in the NETE labora-ccuracy of the FTIR method was assessed by linearof the FTIR-predicted values against the referenceand calculation of the standard error of prediction

and discussion

al considerations

trile is a highly suitable solvent for the extraction oflubricants and its subsequent quantitation by FTIRy for several reasons. First, it is sufficiently polarly immiscible with both mineral- and ester-basedand is capable of efficiently extracting water fromeover, because acetonitrile does not absorb stronglyions of the mid-IR spectrum where water absorbs,

of up to 1000m may be employed for the analy-evels of moisture, thereby providing high sensitivity.nts the OH stretching region in the spectra of a series

on standards prepared by addition of water to acetoni-hich the spectrum of acetonitrile has been subtractedto the binding of water molecules exclusively to ace-

ther than to each other at these high dilutions [7],tra of water exhibit distinct symmetric and asym-stretching bands (3630 and 3540 cm1), rather thanroad band normally observed. Both bands respondadded water and maintain consistent ratios relativeher over the concentration range examined (300.ion to extracting water from oils, acetonitrile is alsoxtracting some other polar constituents, such as alco-lic compounds, hydroperoxides, and organic acids,

ic constituents. Since formulated lubricants contain ay of added constituents and can accumulate variouscontaminants during the lubricants in-service life,f species may be co-extracted with water into ace-

Fig. 2. D300210and asym

tonitriexamia widepresenobtainfrom tconstit

Altpeak msubstainterfewas juthe spspikedobtain0.5% (the levin lubbandstowardwith th

Fig. 3. Dmercialin acetoy recording the spectra of acetonitrile extracts fromiety of commercial new and used lubricants. Fig. 3e 38503200 cm1 region of the differential spectrafter subtraction of the spectrum of dry acetonitrileectra of the standards and indicates that a variety ofs other than moisture were extracted into acetonitrile.h Fig. 3 shows that measurements at either of thea of the water absorption bands would be subject to

interference from various co-extracted components,e on the high-frequency side of the 3630-cm1 bandto be minimal. This is illustrated in Fig. 4, in which

m of the acetonitrile extract of a mineral-based oil286 ppm of water is superimposed on the spectra

fter extraction of the same oil spiked with 0.2 and) 2,6-di-tert-butyl-4-methylphenol, corresponding tot which phenolic antioxidants are commonly presentt formulations. In the case of ester-based oils, theH-containing constituents are broadened and shifteder wavenumbers as a result of hydrogen bonding

ter carbonyl group, thus generally being clear of the

F.R. van de Voort et al. / Talanta 72 (2007) 289295 293

Fig. 4. Differ0.5% (B) 2,6-a differentialextracted into

3630 cm1ing a locatifor the qua

In princare less relithey are moshapes resugen bondinare expectetonitrile sodispersed wselect the owere derivestandards acm1 interpredict thehigh level oThe resultscontributedthan 3676 cing from 3630-cm1of the phenwater contemal frequelubricating

A seconology wastypes of luin underestdilution ofbase oil typous synthetsynthetic hacetonitrileat the otherSynthetic ebility with

ester (C O) absorption at 1730 cm1 in the spectra of theiracetonitrile extracts. To compensate for the partial miscibility

cetonsub

liedthe

tedt 20n 21e sotonin coetoning

.en thfou

the ouentsifierrusters,ilityandessotly icertere

n effeste

alibr

alibring wtoof

yed.om tn wential spectrum of a mineral-based oil spiked with 0.2% (A) anddi-tert-butyl-4-methylphenol and extracted with acetonitrile andspectrum of 286 ppm water (C) spiked into the same base oil andacetonitrile.

water band. Accordingly, the possibility of employ-on on the high-frequency tail of the 3630-cm1 bandntitation of water was investigated.iple, absorption measurements on the tail of a peakable than measurements at the peak maximum sincere likely to be affected by changes in spectral bandlting from intermolecular interactions such as hydro-g. However, in the present case, such matrix effectsd to be minimal, as the samples are very dilute ace-lutions in which water has been shown to be wellithout extensive hydrogen bonding (see Fig. 2). To

ptimal measurement frequency, calibration equationsd by measuring the absorbance of water/acetonitrilet 15 frequencies between 3700 and 3590 cm1 at 8-vals. These calibration equations were then used toamount of water in a mineral oil sample spiked with af phenol antioxidant (0.5%, w/w) and 286 ppm water.showed that the absorption of the phenol antioxidantto the predicted water content at wavenumbers lowerm1, with the magnitude of this contribution rang-40 ppm at 3660 cm1 to 300 ppm at the peak of the

1

with atainingmultiption ofcalculaband abetweetonitrilthe acedilutiowith acdependlength

Givmonlyaffectconstitdemultives,modifimiscibbasedcomprnificanabove,tives wdilutiofor the

3.2. C

A cby addOwingregionemplotrile frequatiowater absorption band. At 3676 cm , the presenceol antioxidant contributed

294 F.R. van de Voort et al. / Talanta 72 (2007) 289295

Table 2Average percent recovery and S.D. for 10 different oil samples

Oil type Spiked additive Average recoverya (%) S.D. (%)Mineral base oil 97.9 0.80Mineral base oil Heavy-duty diesel engine oil additive (13.5%) 95.9 5.25Mineral base oil Antiwear hydraulic oil additive (3%) 97.7 1.46Mineral base oil Industrial gear oil additive (3%) 100.2 4.29Ester base oil 101.1 0.52Synthetic air compressor lubricant 101.9 5.65Engine oil 92.9 2.78Hydraulic fluid 104.9 5.27Turbine fluid 96.5 4.54Gear fluid 94.4 6.41

a Average recovery: average of % recovery values for same sample spiked with 200, 400, 600, 800 and 1000 ppm water.

into 10 samples of different types of oil, mixing vigorously ona reciprocal shaker to ensure that the water was well dispersedin the oil, and then analyzing the samples by the acetonitrileextraction/FTIR method. For certain types of oil samples, recov-ery increased as the time of mixing increased, as exemplifiedby the follohydraulic o80%; 10 m10 differen200 ppm4 out of therecovery wconfirmingwater prese

For the asamples wedifferent daTable 3. Thaverage S.DS.D. of themoisture lethe results,sample pre

Generalmethod remethod, su

case is in fact a primary method, traceable to gravimetricallyprepared water/acetonitrile standards, and, as such, the FTIRmoisture values predicted for lubricant samples should be accu-rate. A performance comparison against the KF procedure wasundertaken to determine the validity of this assumption. Thus,

f useelectmet

d. Fithero-re

2O

.950

2O

.950

tercein t

(Eq.biasthe

me

ref

Table 3Repeatability S.D. o

Sample no.

123456789

101112wing recovery data obtained for a fully formulatedil spiked with 600 ppm water: 2 min, 62%; 5 min,in, 84%: 20 min, 92%. Average recoveries for thet oil samples spiked with 2001000 ppm water inincrements and mixed for 20 min were >90%, with10 samples having 100% recovery (Table 2). Thus,

as deemed satisfactory with a 20-min mixing time,the capability of acetonitrile to extract most of thent in lubricating oils.ssessment of repeatability and reproducibility, 12 oilre analyzed in triplicate within the same run and on 3ys, respectively. The data obtained are summarized ine average S.D. for repeatability was 8 ppm and the. for reproducibility was 80 ppm. The much higherdata for reproducibility suggests that atmospheric

vel variations on different days may have affectedindicating a certain level of moisture ingress duringparation.ly, FTIR spectroscopy is considered a secondaryquiring calibration against a recognized primarych as KF titration; however, the FTIR method in this

a set ofully sthe KFmethoversus

and ze

FTIR H

R= 0

FTIR H

R= 0The inchangeoriginare nottracksthe KFthe KF

and reproducibility data for quantitation of water in 12 oil samples (mean andType of oil RepeatabilityMean (ppm) S.D.New motor oil 361.7 1.5Used motor oil 755.7 10.1Used motor oil 552.3 2.5New bearing oil 5.0 8.5New turbine oil 31.0 10.5New engine oil 473.3 4.0Polyol ester oil 342.7 10.6New hydraulic oil 77.3 8.0New gear oil 96.7 10.1New turbine oil 86.3 8.3Used motor oil 547.0 11.7Used motor oil 590.7 9.2d military vehicle oils (n= 23) and a further 35 care-ed samples were analyzed in the NETE laboratory byhod and were also analyzed by the acetonitrile FTIRg. 5 presents a plot of the FTIR moisture predictionsKF results which yielded the following conventionalgression equations, respectively:

= 36.6 + 0.987 KF H2O,, S.D. = 80 ppm (2)

= 1.063 KF H2O,, S.D. = 83 ppm (3)pt in Eq. (2) is well within the regression S.D., and thehe S.D. when the regression was forced through the(3)) was minimal, confirming that the FTIR results

ed with respect to the KF data. Thus, the FTIR methodmoisture content of the oils in the same manner asthod, with an overall SEP of 80 ppm relative toerence values. The SEP is comparable to the S.D.

f three replicates)Reproducibility(ppm) Mean (ppm) S.D. (ppm)3 324.2 50.342 666.6 108.312 511.4 84.754 39.3 50.954 67.3 52.084 504.4 102.609 467.6 207.438 83.4 64.722 104.2 62.349 92.1 53.249 554.0 74.759 584.2 65.24

F.R. van de Voort et al. / Talanta 72 (2007) 289295 295

Fig. 5. Regression of FTIR moisture predictions vs. the KF results for 58 usedand new oil samples.

for reproducibility of the FTIR method and indicates excellentagreement between the FTIR and KF results.

Overall, the results of the validation study indicate that theFTIR analysis of lubricant acetonitrile extracts is a viable andaccurate means of determining moisture content in a wide rangeof lubricants. The rate of automated analysis on the COATsystem employed in this work is 72 acetonitrile extracts/h; inaddition, ththe oil, canof samplesprovides aallowing foa high degmented toglaboratoryacid numbemonitoring

physical and chemical methods by FTIR methodology. As such,FTIR analysis offers substantial benefits to commercial oil anal-ysis laboratories in terms of both sample turnaround and costsavings.

Acknowledgements

F.R. van de Voort would like to thank Thermal-Lube Inc. forsupporting his sabbatical sojourn at their research facilities inPointe-Claire, Quebec, Canada. Thanks also go to Dr. E. Urban-ski of the Joint Oil Analysis Program run out of the U.S. NavyTechnical Support Center in Pensacola, Florida, for providinglubricant samples. The very helpful collaboration of the NavalEngineering Test Establishment, facilitated by its Director, Mr.Mike Davies, is much appreciated. The authors acknowledge theNatural Sciences and Engineering Research Council (NSERC)of Canada for partial financial support of this research.

References

[1] ASTM D6304: Standard Test Method for Determination of Water inPetroleum Products, Lubricating Oils, and Additives by Coulometric KarlFischer Titration, ASTM International Standards.

[2] F.R. van de Voort, J. Sedman, R.A. Cocciardi, D. Pinchuk, Tribol. Trans. 49(3) (2006) 410.

[3] ASTM E2412: Standard Practice for Condition Monitoring of Usedricanttromvan dl. Spel-Ala.l-Ala

e Narvan d11) (2e sample preparation steps, apart from weighing outbe performed in batch mode while the previous setis being processed. Accordingly, this FTIR methodpractical instrumental alternative to KF analyses,

r the rapid, automated determination of moisture withree of accuracy. This methodology may be imple-ether with other FTIR methods developed in ourfor the automated analysis of lubricants, includingr and base number [8] as well as lubricant condition[2], thereby replacing several conventional, tedious

LubSpec

[4] F.R.App

[5] A. A1295

[6] A. A23.

[7] A. L[8] F.R.

57 (s by Trend Analysis Using Fourier Transform Infrared (FT-IR)etry, ASTM International Standards.e Voort, J. Sedman, V. Yaylayan, C. Saint Laurent, C. Mucciardi,ctrosc. 58 (2) (2004) 193.wi, F.R. van de Voort, J. Sedman, Appl. Spectrosc. 59 (10) (2005)

wi, F.R. van de Voort, J. Sedman, A. Ghetler, JALA 11 (1) (2006)

vor, E. Gentric, P. Saumagne, Can. J. Chem. 49 (1971) 1933.e Voort, J. Sedman, V. Yaylayan, C. Saint Laurent, Appl. Spectrosc.003) 1425.

An automated FTIR method for the routine quantitative determination of moisture in lubricants: An alternative to Karl Fischer titrationIntroductionExperimentalMaterialsInstrumentationAnalytical protocolKF analysisCalibration and validation

Results and discussionGeneral considerationsCalibrationValidation and comparison to Karl Fischer titration

AcknowledgementsReferences