Electrochemical study of nanometer Co3O4, Co, CoSb3 and Sb thin films toward lithium

www.elsevier.com/locate/ssi

ElectrochemicalstudyofnanometerCo3O4,Co,CoSb3andSb

thinfilmstowardlithium

V.Pralong*,J.-B.Leriche,B.Beaudoin,E.Naudin,M.Morcrette,J.-M.Tarascon

´activite´etdeChimiedesSolides,UMR6007Universite´Picardie,33rueSt.Leu,80039Amiens,FranceLaboratoiredeRe

Received20June2003;receivedinrevisedform14November2003;accepted15November2003

Abstract

Therecentinterestinreversibleconversionreactionprocessversuslithiumintransitionmetaloxidesraisesafundamentalquestion

regardingtheroleplayedbythemetalliccobaltortheelectrolyte.Inthiswork,wereportontheelectrochemicalreactivityofcobaltoxideandantimonide,metalliccobaltandantimonythinfilmstowardlithium.Nanometerthinfilmswerepreparedbypulsedlaserdeposition(PLD)usingastoichiometrictarget.Theobtainedfilmswerecharacterizedfromtheirchemicalandelectrochemicalpropertiesbymeansofvariouscomplementarytechniquessuchasx-raydiffraction(XRD),electronicmicroscopyandelectrochemicalquartzcrystalmicrobalance(EQCM).D2004ElsevierB.V.Allrightsreserved.

Keywords:Co3O4;Co;CoSb3;Sb;PLD;EQCM;Lithiumionbatteries;Thinfilm

1.Introduction

Thestudyofnegativeelectrodematerialsforrecharge-ablelithiumbatteriesremainsamajortopicintheareaofenergystorage.Alternativestocommercialcarbonaceousmaterialsareeagerlysoughtaftertomatchthehighestcapacitiesofnewemergingpositiveelectrodematerials.Explorationsoftransitionmetaloxidesasnegativeelectrodematerials,firstreportedbyIdotaetal.[1],Tarasconetal.[2,3]andLerouxetal.[4],werepursuedandledtothedevelopmentofnewconceptsregardingLiuptake/removalmechanism.RecentinteresthasalsobeenarousedbyreportsonLiuptakeintransitionmetalcompoundsofGroupV,suchasphosphides(CoP3)andantimonides(CoSb3),whichshowasimilarreversibleLiuptake[5,6].Itwasshownthataswellasthereductionofmetaloxidewithlithium,besidesthegrowthofmetalnanoparticlesandLi2OorLi3P,Li3Sb,therewasalsotheformationofapolymericinterfacegrowthbetweentheactivematerialandtheelectrolyte.

Hopingtofurtherthrowlightonthesereportednovelmechanisms,webeganastudyofthesecompoundsintheirthinfilmforms.Electrodematerials,namelycathodes,havelongbeenstudiedasthinfilms.Attheearlystage,

*Correspondingauthor.Tel.:+33-2-31-45-26-05;fax:+33-2-31-95-16-00.

E-mailaddress:valerie.pralong@ismra.fr(V.Pralong).0167-2738/$-seefrontmatterD2004ElsevierB.V.Allrightsreserved.doi:10.1016/j.ssi.2003.11.018

sputteringtechniques[7]werepreferablyusedoverpulsedlaserdeposition(PLD),butoverthelastfewyears,thetrendhasbeenmovingmoretowardslaserablation,whichisamoreversatiletechnique[8–11].Moreover,thebene-fitsofstudyingthinfilmstoaddressthisissueresideinthefactthatnoelectronicadditives(suchascarbon)arerequiredtoperformelectrochemicalquartzcrystalmicro-balance(EQCM)studiesthatareessentialtostudylithiumreactivity.

Herein,weextendedourexpertiseonthinfilmgrowthofelectrochromicmaterialsWO3[8],NiO[9],Sn–Sb–O[10]tothatofbatteryelectrodematerials.

2.Experimental

2.1.Characterizationtechniques

X-raydiffraction(XRD)patternswerecollectedusingeitherafour-circleScintagdiffractometerwithCuKa(k=0.15406nm)oraBrukerdiffractometerD8PSDdetectorwithCoKa(k=0.178897nm).Ascanningelectronmicroscope(SEM)fieldeffectgun(FEG)PhilipsXL-30witharesolutionofabout1nmwasusedtostudytexturalchanges,whileelementalcompositionsweredeterminedbyenergy-dispersivespectroscopy(EDS)onaLink-Isisana-lyzer(ATW6650detector).Thesamplemorphologieswere

296V.Pralongetal./SolidStateIonics166(2004)295–305

Fig.1.XRDpatternsoftheCo3O4thinfilmsdepositedina10À1mbaroxygenatmosphereat25jC(a),300jC(b)and600jC(c).

determinedbymeansofatransmissionelectronmicroscope(TEM)PhilipsCM12instrument.Filmthicknessesrangingfrom100to700nmweredeterminedbyprofilometry(DektarStInstrument).

CobaltoxideCo3O4thinfilmsweredepositedona1.37-cm2nickelsubstratebypulsedlaserdeposition(PLD),usingKrFexcimerlaserbeam(Lambdaphysik,Compex102,k=248nm)withalaserfluencyof2J/cm2.Ablationwasperformedwith80%ofthetheoreticaldensity(d=4.07g/cm3)counterrotatingtargetandwithatarget–substratedistanceof4cm.ForCo3O4thinfilmgrowth,thepressureinthechamberwasfixedat10À1mbarofoxygen,usedalaser-shootingrepetitionrateof5Hz(0.45A

˚andwe

/shot)for30min.Thesubstratetemperaturevariedfrom25to700jC.ThegrowthofCo,CoSb3andSbthinfilmswasperformedinasimilarfashionfor30minusinghigh-puritytargets(Alphaeaser99.95%)thatwereablatedunder10À5mbarvacuumatroomtemperatureusingalaserenergyof200mJ,atafrequencyof5Hz,whilethetarget–substratedistancewasfixedat4cm.Topreventthespontaneousoxidationofametalliccobaltthinfilm,weequippedourPLDsystemwithagloveboxenablingustohandleandcharacterizesampleswithoutanyairexposure.

Thephysicochemicalcharacterizationandelectrochemi-calstudyofthegrownthinfilmswereperformedbymeansofElectrochemicalQuartzCrystalMicrobalance(EQCM)monitor.APM-700(Maxtek)QCMapparatuswasusedwitha5MHzquartzcrystalcoveredwithatitaniumdeposit(S=1.37cm2).SuchapparatuscoupledwithaVMP(bio-logic)enabledus,oncetheablationwasperformed,todeterminethedepositfilmmassthroughresonancefrequen-cyaccordingtotheSauerbreyequation[12].Afterwards,theQCMsubstratewasplacedinanelectrochemicalcellspeciallydesigned(describedinRef.[13])forinsitumonitoringmasschangeslinkedtotheredoxprocessduringthecyclingofourelectrodeversusLi.Duetothecellconfigurationandtheconfinementoftheelectrolyteagainstthequartz,theslopeisgivenwithanaccuracyofF1g/mol.Therefore,inthispaper,emphasisisbeingplacedontrendsinsteadofprecisevalues.

Theelectrochemicalstudywasperformedusingcoincells.Forthenegativeelectrodes,0.6cm2disksweredirectlycutoutoflithiumcommercialfoil(Aldrich,0.05cmthickness).Thepositivediskwasseparatedfromthelithiumnegativeelectrodebyaborosilicatemicrofiber(Millipore)filmimpregnatedwith0.5mlofaLiPF6(1M)inEC-DMC(1–1)solution(Prolabo,LP30).TheelectrochemicaltestswerecarriedoutusingaVMPsystem(Biologic,Claix,France)thatcanbeoperatedinbothpotentiodynamicandgalvanostaticmodes.AllreportedpotentialvaluesaregivenversusLi/Li+.

3.Resultsanddiscussion3.1.ThinfilmsgrowthbyPLD

Fig.1showstheX-rayDiffractionpatternofthreeCo3O4thinfilmspreparedat25,300and600jC,respectively.Co3O4(311)and(220)reflectionsarebarelydetectedat25jCandgrowslightlyupto300jCtofinallybecomesharperfrom300to600jC.FromthewidthofthesepeaksandusingtheScherrerequation[14],wecandeduceanapprox-imatecrystallitesizeof10,40and70nmat25,300and600jC,respectively.TheSEMmicrographsrevealahomoge-neousdepositinthewholerangeoftemperatureconditions(Fig.2).ParticlesizemeasurementsfromtheSEMimagewereinthesameorderastheonesmeasuredbyXRD,

Fig.2.ScanningelectronicmicroscopymicrographoftheCo3O4thinfilmsat300jC(a)and600jC(b).

V.Pralongetal./SolidStateIonics166(2004)295–305297

suggestingthattheparticlespresentamonolithictexture(uniquecoherentdiffractionareafromXRD;Fig.3a).Filmthicknesswasmeasuredbymeansofaprofilometer,andtheeffectofdepositiondurationtimeonthefilmthicknessisreportedonFig.3b.Anexponentialincreaseinthethicknessisobservedfor5–30mindepositiontimes.

TheroomtemperatureTEMimageforaCothinfilm(5Hz,200mJ,30min,10À5mbar,4cm,25jC)isreportedonFig.4,togetherwiththecorrespondingelectronicdiffraction(SAED).Thecrystallitesizesareabout10nmdiameter.Inaddition,SAEDrevealsthepresenceofringscorrespondingtometalliccobalthavingcubicandhexagonalstructures;thisdoesnotcomeasasurprisebecausemetalliccobaltsynthesisalwaysleadstothecoexistenceofbothtypesofstructure.Fig.5showstheXRDpatternofaCoSb3thinfilmdepositedat25jCobtainedafterannealingfor2hat200jC.Itisworthnotingthata200jCannealingtemperature

wasnecessarytoformthisphaseasconfirmedbyarecentstudyindependentlycarriedoutbyanothergroup[15].Theaverageparticlesizewas400nm.Finally,tobeconsistent,aswillbedetailedbelow,Sbthinfilmsweregrownusingthesameconditionsastheonesusedforthemetalliccobaltthinfilm(5Hz,200mJ,30min,10À5mbar,4cm,25jC).AsreportedontheSEMimage(Fig.6),theSbfilmsarehomogeneouswitha200-nmthickness.3.2.Characterizationofthethinfilm

Alltheabovefilms,Co3O4,Co,CoSb3andSb,havebeencharacterizedfortheirelectrochemicalproperties.Forreasonsofclarity,thedatawillbepresentedseparatelyforeachcompound.

Fig.7ashowsthepotential/compositioncurveobtainedonthegalvanostaticcyclingofaCo3O4/Licellperformedat

Fig.3.Evolutionofthecrystallitesizesversusthetemperaturefor30-mindeposition(a),filmthicknessasfunctionofdepositiondurationforCo3O4filmgrownat300jC(b).

298V.Pralongetal./SolidStateIonics166(2004)295–305

Fig.4.TEMmicrographsofmetalliccobaltimageandcorrespondingelectronicdiffraction.

anominalCrate(10Li+/h)inthe0.05–3.0Vpotentialwindow.Thefirstdischargeprocessischaracterizedbyaquasilinearpotentialdecreasecorrespondingtotheuptakeof10lithium.Uponincreasingthedischargerate,1or10Li+/h,thevoltage–compositionprofilecurveevolvesfromaniceplateautowardaslowslopingvoltagecurveimplyingsomekineticslimitations(insetFig.7a).Thus,itcomesasnosurprisethatthevoltage–compositioncurveobtainedfor

Fig.5.XRDpatternsoftheCoSb3thinfilm.

Fig.6.SEMimageofSbthinfilm.Co,CoSb3andSbfilmsweredepositedinvacuumatRT.

athinfilmataslowrateissimilartotheonepreviouslyobtainedwithCo3O4powdertestedintheSwagelokcell[16].Then,followingthecharge–dischargecycle,sevenlithiumcanbereversiblyremoved.

Regardingtheevolutionofthecapacityretentionuponcycling,weperformedthesubsequentcharge/dischargeofa100-nm-thickfilm,usinga20Li+/hrate(Fig.7b).Thecapacityrapidlydecreasedduringthefirst10cyclesandstabilizedoverthenext150cyclestofinallyincreaseandreach800mAh/gafter350cycles.SuchaphenomenonhadalreadybeenobservedforCoO/Licell,andwasassociatedwiththeoccurrenceofasecondredoxprocessinvolvingtheelectrolyte[17].

TheTEMmicrographofthestartingCo3O4film(Fig.8a)showsthatitconsistsofparticlesofdimensionsrangingfrom10to100nm,andthecorrespondingSAEDpatternsuggeststhepresenceofwell-crystallizednanoparticlesinaccordancewiththeXRDresults.WhentheCo3O4isfullyreduced(Fig.8b),thebrightfieldimagerevealsthepresenceof10–20nmnanoparticlesembeddedwithinthecrystalliteandthepreservationoftheoverallshapeofthestartingparticletogetherwiththepresenceofanamorphouspoly-mericlayer.Moreimportantly,reflectionscorrespondingtoCoandLi2OareclearlyevidencedintheSAED.

Uponthefollowingcharge(3.5V),theSAEDpattern(Fig.8c)clearlyshowsthevanishingoftheCoandLi2OreflectionsattheexpenseofthoseofCo3O4.ThisconfirmsthepreviouslyreportedmechanismthatentailsthereductionofCo3O4inCoduringthedischargeconcomitantwiththeformationofLi2O(reaction(1)),thereverseoccurringuponcharge.

Themechanismcanbewrittenasfollows:Co3O4þ8Liþþ8eÀf3Coþ4Li2O

þelectrolytereductionð1Þ

V.Pralongetal./SolidStateIonics166(2004)295–305299

Fig.7.(a)VoltageprofileofaCo3O4/Licellcyclingatroomtemperaturebetween0.05and3VversusLi/Li+witha10Li+/hrate.Inset:comparisonvoltage–compositionprofilebetweentworates,1and10Li+/h.(b)Gravimetricchargecapacityasfunctionofcyclenumbersof0.1mgCo3O4/Licellunderaconstantcurrentof1100mA/gcorrespondingtoarateof2C(20Li+/h).

Thevoltage–timecurvereportedonFig.9afortheCothinfilmrevealsalineardecreaseinthepotentialduringthecourseofthefirstdischarge.Itisworthnotingthatsincenoreductionofthetransitionmetalcouldoccur,electrolytedecompositionmustberesponsiblefortheobservedcapac-ity.Interestingly,thiscapacityisreversibleandmostlikelyduetotheinsituformationofanelectroactivesurfacelayer.Unfortunately,SAEDofthecycledfilmisfeatureless,implyingatotalamorphousactivematerial,whichisdiffi-culttoidentifywiththecurrentcharacterizationtools.

ConcerningtheCoSb3andSbfilms,theirvoltagecom-positioncurvesareshowninFig.9bandc.Theprofilesareidenticaltothoseobservedinthebulk,withthewell-pronouncedroundingofthepotentialwhenitreaches0.05

VforCoSb3insteadofasharpdropofpotentialforSbelectrode[18].Fromtheelectrochemicalcurves,wecanhypothesizethatthereactingpathoccursonthethinfilmsthesamewayasinthebulk,asdescribedbelow:CoSb3þ9Liþþ9eÀfCoþ3Li3SbþelectrolytereductionSbþ3Liþþ3eÀfLi3Sb

ð2Þð3Þ

Tofurthercharacterizethereactionprocess,wedecidedtomonitorthemasschangeofthethinfilmsduringthefirstdischargebymeansoftheEQCM.

300V.Pralongetal./SolidStateIonics166(2004)295–305

Fig.8.TEMmicrographsandcorrespondingelectronicdiffractionofCo3O4thinfilm(a)asdepositedfilmat25jC,(b)dischargedmaterial0.05V,(c)chargedmaterial3V.

3.3.Electrochemicalquartzcrystalmicrobalance(EQCM)Unlessotherwisespecified,EQCMmeasurementswillbereportedbyplottingthevariationoftheelectrodemassasafunctionofthenumberofelectrons(nelectrons=ISt/FwithS=1.37cm2,F=96,480C/mol).Furthermore,toeliminateanypossibleinterferencefromtheTi-coatedquartzmicro-electrode,wefirstrunablank.Thatistosay,abareTi-coatedelectrodewastestedinanEQCMexperimentwithasolutionof1MLiPF6inEC-DMCaselectrolyteandlithiumascounterelectrode.Forthisexperiment,weusedthesamecurrentvalueasthatusedforthethinfilmexperiment(0.05mA/cm2),andundertheseconditions,nomasschangeswerenotedunderpolarization.Thisimpliesthatfromnowonalltheobservedchangeswillbeintrinsictothethinfilmmaterial.Foreachinvestigatedcompound,thereareseveralslopesthataremostlikelyreminiscentofdifferentprocesses,implyingthewealthoftheoverallmechanism.TheEQCMtracesforCo3O4/LiandCo/Lihalf-cellsduringthefirstdischargeareshowninFig.10.RegardingCo3O4,themassevolutionduringfirstdischargereveals,althoughpoorlydefined,twoslopingvoltagerangesof6.5and9g/mol,suggestingcompetingprocesses,respectively.OncethedischargepotentialoftheCo3O4/Licellhasreached1.1V,theCo3O4phasedecomposesintoLi2O+CoaccordingtotheredoxreactionCo3O4+8-Li++8eÀ!3Co+4Li2O(Eq.(1)).Assumingthatlithiumaloneisinvolvedintheprocess,weshouldmeasureaslopeequivalentto6.94g/mol.Thisvalueisinthesamerangeasthecalculatedone(6.5g/mol),suggestingthat,atthebeginningofthedischarge,onlytheCo3O4decompo-sitionoccurs,asshowninreaction(1).Moreover,inordertoconfirmtheexperimentalslopeof6.5g/mol,measuredduringtheearlystageoftheCo3O4reductionversuslithium,thesameEQCMexperimentwasperformedonaCo3O4thinfilmelectrochemicallyreducedinNaClO4(1M)PCelectrolyte.Tooursatisfaction,theresultingslope

V.Pralongetal./SolidStateIonics166(2004)295–305301

Fig.9.Cyclingcurvesat20Li+/hrateofthe(a)Cofilms,(b)CoSb3films,and(c)Sbfilms.

wasequalto21g/molcomparedwiththetheoretical23g/molvalue.

Concerningthesecondslope,9g/mol,greaterthantheoneexpectedforpurelithium,itishighlyprobablethatwearefacingatleasttwocumulativereactions.BesidestheelectrochemicaldecompositionoftheCo3O4,wearebegin-ningtovisualizetheelectrolytereduction.Thus,alegitimatequestionconcernsthepossibilityofdeterminingthenatureofthereducedspeciesbyEQCM.ThereductionofthesaltLiPF6andthesolventEC-DMCisaverycomplexprocess,alreadystudiedbyAurbach[19,20]andEvansandKwon[21].Nevertheless,basedonthesepreviousstudies,somehypothesescanbeputforward;forexample,theformationofLi2CO3byreductionofthecarbonates(EC,DMC;slope

Àof34g/mol).ThePF5salt,formedafteradisproportion-ÀationofLiPF6!LiF+PF5,canbereducedintoLiF(slopeof26g/mol),someH2OandO2tracesintheelectrolytemaybereducedintoLi2O(slopeof15g/mol).Forinstance,aslopeof9g/molcanbeexplainedbythecompetitionofseveralredoxprocessesincludingthereductionofthe

302V.Pralongetal./SolidStateIonics166(2004)295–305

Fig.10.ComparisonbetweenpotentialandcorrespondingmassoftheelectrodeversusmolenumberduringthefirstdischargeofCoandCo3O4thinfilmsversuslithiuminLiPF6(1M)EC-DMC.

electrolyte.Indeed,ifweassumethat10%ofthecurrentisusedtoreducetheelectrolyteleadingtotheformationofLi2O(M=15g/mol),thenacalculatedslopewillbeabout9g/mol,whichisofthesameorderastheexperimentalone(9g/mol).InthecaseofLi2CO3formation,aslopeof9g/molwillbeobtainedif5%ofthecurrentisusedfortheelectrolytedecomposition.Thus,itwouldbehazardoustoconcludesolelyfromEQCMontheexactnatureoftheslope.AdditionalmeasurementslikeXPSneedtobeperformed.

Forthesakeofcomparison,EQCMmeasurementswereperformedonCothinfilmsthatweregrownunderthesameconditions(5Hz,200mJ,30min,4cm,25jC)astheCo3O4thinfilms.Then,arateof10Li+/hwasusedforthisexperiment.

OntheCofilmresponse(Fig.10),theevolutionofthepotentialduringthefirstdischargeshowsalineardecrease,andwecanobservealinearincreaseintheweightuptoabout19g/mol,whichmostlikelyisreminiscentofelectrolytedecomposition.Startingfromnanometricpar-ticlesofmetalliccobalt,theonlyelectrochemicalreactionthatcouldoccuratthispointisthedecompositionoftheelectrolyte.Again,theexactinterpretationofthesloperemainshazardous.

Theseconddischargeresponsebringsoutmoreinforma-tion(Fig.11).Whereasthepotentialevolutionisquasi-linearwecanobservetwodifferentmassuptakeregimechangesof7.5and13.5g/mol,respectively.Fromsuchexperiments,onecanhypothesizethatthematrix‘‘Co–Li2O’’isformedafterthefirstdischarge,suggestingthattheLi2Ophase,whichisnecessarytotheredoxprocess,couldbeformedinsitufromtheelectrolytedecomposition.Forsuchascenario,amassslopeof15g/molwouldbeaccepted,nottoofarfromtheexperimentalvalue.

Finally,itisworthpointingoutthatthederivativecurvedx/dVversusvoltageofthecharge–dischargecurveofthe

Fig.11.Potentialandcorrespondingmassoftheelectrodeversusmolenumberofa0.1mgCothinfilm(5Hz,200mJ,4cm,10À1mbarO2,15min)versuslithiuminLiPF6(1M)EC-DMC,(10Li+/hrate)ontheseconddischarge.

V.Pralongetal./SolidStateIonics166(2004)295–305303

Fig.12.Comparisonofthederivativedx/dVversuspotentialcurveofCo3O4(.)andCo(5)atthe30thcycleand10thcycle,respectively.

Co/Licellafter10cyclesissimilartotheoneobtainedforthe30thcycleoftheCo3O4/Licell(Fig.12).Thispointsimplyimpliesthat,afteralong-termcycling,theCo3O4electrodepresentsthesameelectrochemicalandstructuralcharacteristicsasaCoelectrode.Thus,itcouldagainbespeculatedthat,whateverourstartingprecursor,thereisafterafewcyclesthroughelectrolytedecompositionforma-tionofapolymermatrixcontainingatleastCoandLi2O.Inshort,throughEQCMmeasurementsweconfirm:(1)theuptakeoflithiumorsodiumintotheCo3O4thinfilm,leadingtothedecompositionofthecobaltoxideandtheformationofConanoparticles;(2)theformationofapolymer/gel-likefilmbyelectropolymerizationoftheelec-trolyteattheendofthedischarge.However,cautionhastobeexercisedininterpretingthemechanismbywhichthepolymerfilmformsortheoriginofthethinfilmcapacityafterlongcycling,whilebeingmainlyformedfrom‘‘Co,Li2O’’imbeddedinapolymermatrix.

Inthehopeofprovidingmoredatabaseonthedecom-positionreaction,wealsoinvestigatedtheantimonidesys-tem,whichpresentsasimilarreactivityasthetransitionmetaloxide,namelyadecompositionprocess.ThisleadstoaCo/Li3Sb–CoSb3+polymersystem,toalesserextentthantheCoOsystem.Ontheopposite,theSbsystemdoesnotinvolveapolymericfilmgrowth(voltagesharpdropandnoextracapacity).

Figs.13and14showthemelectrodeversusnelectronscurvesforCoSb3andSb,respectively.First,wecanclearlyobservethatthemassevolutioniscompletelydifferentfortheantimony-basedfilmcomparedtothecobalt-basedfilms.

Fig.13.Potentialandcorrespondingmassoftheelectrodeversusmolenumberofa0.4mgCoSb3thinfilm(5Hz,200mJ,4cm,10À6mbar,15min)versuslithiuminLiPF6(1M)EC-DMC,(10Li+/hrate)onthefirstdischarge.

304V.Pralongetal./SolidStateIonics166(2004)295–305

Fig.14.Potentialandcorrespondingmassoftheelectrodeversusmolenumberofa0.1mgSbthinfilm(5Hz,200mJ,4cm,10À6mbar,15min)versuslithiuminLiPF6(1M)EC-DMC,(10Li+/hrate)onthefirstdischarge.

Indeed,afteramassincreaseofabout8g/mol,whichisconsistentwithalithiumuptakeintotheCoSb3andSbelectrodes,adecreaseofaboutthesamerangeisobserved.Afterthat,fortheCoSb3film,themassincreasestoabout53g/mol,suggestingthatanotherreactionprocessinaddi-tiontothelithiumuptaketakesplace,mainlytheelectrolytereduction.Curiously,theSbfilmmasschangeevolutionisdifferentbecauseaslopecloseto8g/molisobservedattheendofthedischarge.Purelycoincidentalornot,strictlyforSb,noextracapacitywasobservedandthepotentialsud-denlydecreased.Atfirst,thisexperimentalresultcouldsuggestadifferentreactivitybetweenintermetallicalloysandoxides.However,onebelievesthatsucheffectsarenotintrinsictothematerialbutratherreflectchangesinthefilmmorphology(texture,cracking,etc.).Indeed,itshouldberecalledthatintermetallicanodematerialsSbandCoSb3,undergoa20–30%volumechangeuponlithiumuptake[Sb

˚3),Li3Sb(70.9A˚3)andCoSb3(92.2A˚3)],resulting(29.9A

inaprogressivealterationofboththeelectronicpathwithintheelectrodeandthequartzcrystalfrequencyduringtheredoxprocess.Suchavolumechangeisnotobservedinthecaseofthetransitionmetaloxidereaction.Wearepresentlyfurtherexploitingsuchanunexpectedeffecttoknowwhetherwecould,fromthesefrequencyanomalies,graspmoreinformationaboutlithiumelectrochemicallydrivensurfacetexturinginintermetallicalloys.

involvinglithiumuptakeonpolymericgrowthorevenLi-alloyingreactions.Nevertheless,inthelattercase,EQCMmaypresentnewopportunitiesthatarebeingpursuedbyourgroup.

Acknowledgements

V.PralongisgratefultoA.Rougier,F.LeMarrecforenlightendiscussionsandtheirformationtothedepositiontechniques(PLD).TheauthorsthankM.Nelsonforhertechnicalhelpinthemanuscriptpreparation.

References

[1]Y.Idota,T.K.A.Maekawa,T.Miyasaka,Science276(1997)1395.[2]P.Poizot,S.Laruelle,S.Grugeon,L.Dupont,J.-M.Tarascon,Nature

407(2000)496.

[3]J.-M.Tarascon,M.Armand,Nature414(6861)(2001)359.

[4]F.Leroux,G.R.O.W.P.Power,L.F.Nazar,Electrochem.SolidState

Lett.1(6)(1998)255.

[5]V.Pralong,D.C.S.Souza,T.Leung,L.F.Nazar,Electrochem.Com-mun.4(2002)516.

[6]J.-M.Tarascon,L.Dupont,Y.Chabre,C.Payen,M.Morcrette,V.

Pralong,J.Electrochem.Soc.150(6)(2003)A732.

[7]S.D.Jones,J.R.Akridge,F.K.Shokoohi,SolidStateIonics69(1994)

357.

´de´,M.ElMarssi,Appl.Surf.Sci.153[8]A.Rougier,F.Portemer,A.Que

(1999)1;

´de´,J.Electrochem.Soc.148(2)(2001)7.A.Rougier,A.Blyr,A.Que

[9]I.Bouessay,A.Rougier,J.M.Tarascon,J.Electrochem.Soc.,

2004,inpress.

´ry,J.-M.[10]C.Marcel,M.S.Hegde,A.Rougier,C.Maugy,C.Gue

Tarascon,Electrochim.Acta46((13–14)2)(2001)2097.

[11]B.J.Neudecker,R.A.Zuhr,B.S.Kwak,J.B.Bates,J.D.Robertson,J.

Electrochem.Soc.145(12)(1998)4148.

[12]TheresultsofhisworkareembodiedintheSauerbreyequation,

whichrelatesthemasschangeperunitareaattheQCMelectrodesurfacetotheobservedchangeinoscillationfrequencyofthecrystal:

4.Conclusion

Throughthispaper,wehaveshowntheadvantagesofpulsedlaserdepositionsinrapidlygrowingthinfilmsmimickingtheelectrochemicalpropertiesofbulkpowders.Nevertheless,weexperiencedthatEQCMusedforelectro-depositionstudiesormonitoringmodelreactionhavesomelimitationswhenappliedtonovelreactivitymechanisms

V.Pralongetal./SolidStateIonics166(2004)295–305

f=ÀCfÂm(f=theobservedfrequencychangeinHz,Cf=thesensitivityfactorofthecrystalinHz/ng/cm2,0.056Hz/ng/cm2fora5-MHzcrystalat20jC).

B.Laik,P.Poizot,J.-M.Tarascon,J.Electrochem.Soc.149(3)(2002)A251.

P.Scherrer,Nachr.Go¨tt.Gesell2(1918)98.B.Schupp,I.Ba¨cher,M.Hecher,N.Mattem,V.Savchuk,J.Shu-mann,ThinsSolidFilms434(2003)75.

D.Larcher,G.Sudant,J.-B.Leriche,Y.Chabre,J.M.Tarascon,J.Electrochem.Soc.149(3)(2002)A234.

305

[13][14][15][16]

[17]S.Laruelle,S.Grugeon,P.Poizot,M.Dolle,L.Dupont,J.-M.Taras-con,J.Electrochem.Soc.149(5)(2002)A627.

[18]J.-M.Tarascon,M.Morcrette,L.Dupont,Y.Chabre,C.Payen,D.

Larcher,V.Pralong,J.Electrochem.Soc.150(6)(2003)A732.[19]D.Aurbach,J.PowerSources89(2000)206.[20]D.Aurbach,J.Electroanal.Chem.393(1995)43.

[21]K.Evans,J.W.E.Kwon,Electrochem.Solid-StateLett.5(3)(2002)

A59.