WormlikeMicellesFormedbySynergisticSelf-AssemblyinMixturesofAnionicand - PDF document

WormlikeMicellesFormedbySynergisticSelf-AssemblyinMixturesofAnionicandCationicSurfactantsSrinivasaR.Raghavan,*GerhardFritz,andEricW.KalerCenterforMolecularandEngineeringThermodynamics,DepartmentofChemicalEngineering,UniversityofDelaware,Newark,Delaware19716ReceivedOctober15,2001.InFinalForm:January28,2002Self-assemblyinmixturesofcationicandanionicsurfactantsoccurssynergisticallybecauseofattractiveinteractionsbetweentheoppositelychargedheadgroups.Here,sucheffectsareexploitedtoobtainhighlyviscoelasticfluidsatlowtotalsurfactantconcentration.ThesystemsconsideredaremixturesoftheCtailedanionicsurfactant,sodiumoleate(NaOA),andcationicsurfactantsfromthetrimethylammoniumbromidefamily(CTAB).Inparticular,mixturesofNaOAandCTABshowremarkablyhighviscosities:for3%surfactant,thezero-shearviscositypeaksatca.1800Pasforaweightratioof70/30NaOA/TAB.Thehighviscositiesreflectthegrowthofgiant,entangledwormlikemicellesinthesolutions.MixturesofNaOAwithashorter-chainanalogue(CTAB)havemuchlowerviscosities,indicatingaweakmicellargrowthandhenceaweakattractionbetweenthesurfactants.Ontheotherhand,increasingtheTABtaillengthto10or12leadstomuchstrongerinteractionsbetweenthesesurfactantsandNaOA.Consequently,bothmicellarandbilayerstructuresareformedinthesemixtures,andthesamplesseparateintotwoormorephasesoverawidecompositionrange.Thus,thesynergisticgrowthofwormlikemicellesincationic/anionicmixturesismaximizedwhenthereisanoptimalasymmetryinthesurfactanttailOverthepast15years,theself-assemblyofionicsurfactantsintowormlikemicelleshasbeenwidelyWormlikemicellesaretypicallyformedbyaddingsalttosolutionsofacationicsurfactantsuchascetyltrimethylammoniumbromide(CTAB).Thesaltfacilitatesmicellargrowthbyscreeningtheelectrostaticrepulsionsbetweenthechargedsurfactantheadgroups.Muchlikepolymers,wormlikemicellestendtobelongandflexiblechains(contourlengthsofca.1m)thatbecomeentangledintoatransientnetwork,therebyimpartingviscoelasticitytothesolution.Themicellescanthusbeexploitedforthickeningandrheology-controlapplicationsinaqueoussystems.Mixturesofsurfactantscanexhibitsynergisticgainscomparedtotheparentsurfactantsinbothsurfaceandbulkproperties.Thisisparticularlytruewhenthereareattractiveinteractionsbetweenthesurfactants,asisthecaseinmixturesofanionicandcationicsurfactants.Forexample,addingsmallamountsoftheanionicsurfactantsodiumdodecylbenzenesulfonate(SDBS)tosolutionsofthecationicsurfactantcetyltrimethylammo-niumtosylate(CTAT)causesasynergisticenhancementoftherheologicalproperties.NotethatCTAThasahydrophobiccounterionthatbindsstronglytothemicelle,therebypromotingmicellargrowth.Thus,CTATbyitselfformswormlikemicelles,andtheresultingsolutionsareAddingSDBStoCTATincreasesthezero-shearviscosity10-fold,withthepeakoccurringataweightratioof4/96SDBS/CFurtheradditionofSDBSlowers,andforSDBS/CTATratiosexceeding9/91,thesamplesseparateintotwophases.AtevenhigherSDBSfractions,bilayerphases(vesiclesorlamellae)coexistwithmicellarsolutions,andacrystallineprecipi-tateformsattheequimolarcomposition.Theabovestudiesillustratehowstronginteractionsbetweenheadgroupsincationic/anionicmixturescanfacilitatemicellargrowthaswellascausephasesepara-tion.ThemicellesgrowatlowSDBScontentbecauseSDBSreducesthesurfacepotentialofthemixedmicellesviachargeneutralization,whilethereleasedcounterionsincreasetheionicstrength.AthigherSDBScontent,however,thestronglyboundcationicanionicpairactslikeadouble-chainedzwitterionandpackspreferentiallyintobilayergeometries.Attheequimolarratio,thecatanionicsurfactantC(i.e.,thecompoundformedbyeliminatingthecounterionsfromthetwosurfactants)precipitatesoutofthesolution.Previousstudiesofcationic/anionicmixturesgestthatphaseseparationandprecipitationcanbeavoidedifthesurfactantscontainonelongandoneshortalkyltail.Forexample,dodecyltrimethylammonium *Towhomcorrespondenceshouldbeaddressed:e-mailsraghava@eng.umd.edu;Tel(301)405-8164;FAX(301)405-0523.(1)Cates,M.E.;Candau,S.J.J.Phys.:Condens.Matter(2)Rehage,H.;Hoffmann,H.Mol.Phys.,933.(3)Candau,S.J.;Oda,R.ColloidsSurf.,5.(4)Kaler,E.W.;Herrington,K.L.;Murthy,A.K.;Zasadzinski,J.A.N.J.Phys.Chem.,6698.(5)Herrington,K.L.;Kaler,E.W.;Miller,D.D.;Zasadzinski,J.A.;Chiruvolu,S.J.Phys.Chem.,13792.(6)Brasher,L.L.;Herrington,K.L.;Kaler,E.W.,4267.(7)Yatcilla,M.T.;Herrington,K.L.;Brasher,L.L.;Kaler,E.W.;Zasadzinski,J.A.J.Phys.Chem.,5874.(8)Koehler,R.D.;Raghavan,S.R.;Kaler,E.W.J.Phys.Chem.B,11035.(9)Holland,P.M.,Rubingh,D.N.,Eds.;MixedSurfactantSystemsACSSymposiumSeries501;AmericanChemicalSociety:Washington,DC,1992;Chapters1and2.(10)Rosen,M.J.InMixedSurfactantSystems;ACSSymposiumSeries501;Holland,P.M.,Rubingh,D.N.,Eds.;AmericanChemicalSociety:Washington,DC,1992;pp316 (11)Soltero,J.F.A.;Puig,J.E.,2654.(12)Regev,O.;Khan,A.J.ColloidInterfaceSci.,95.(13)Edlund,H.;Sadaghiani,A.;Khan,A.,4953.(14)Barker,C.A.;Saul,D.;Tiddy,G.J.T.;Wheeler,B.A.;Willis,J.Chem.Soc.,FaradayTrans.1,154.(15)Kato,T.;Iwai,M.;Seimiya,T.J.ColloidInterfaceSci.,439.(16)Huang,J.B.;Zhao,G.X.ColloidPolym.Sci.,156.(17)Zhao,G.X.;Xiao,J.X.ColloidPolym.Sci.,1088.(18)Talhout,R.;Engberts,J.B.F.N.,5001.10.1021/la0115583CCC:$22.002002AmericanChemicalSocietyPublishedonWeb04/17/2002 chloride(CTAC)andsodiumnonanoate(SN)formisotropicmicellarsolutionsatallcompositionsupto40wt%surfactant.Taillengthasymmetrydoesnotnecessarilyprecludephaseseparation,though,forex-ample,mixturesofsodiumdodecylsulfate(SDS)andoctyltrimethylammoniumbromide(CTAB)formtwophasesoverarangeofcompositions,asdomixturesofCTACandsodiumperfluorononanoate(thefluoro-carboncounterpartofSN).Significantly,manyoftheabovemixturesdoshowahigherviscosityrelativetothepurecomponentsolutions.Forinstance,intheCsystem,addingSNtoCTACat10%totalsurfactantincreasedtheviscosity50-foldtoapeakattheequimolarTheviscosityrisecorrelatedwiththegrowthofmicellesfromspherestorodsofaxialratioabout10.Here,wereportcationic/anionicmixturesthatexhibitmillion-foldincreaseinviscosityrelativetothesingle-componentsolutions.Theanionicsurfactantinthiscaseissodiumoleate(NaOA),whichhasaCalkyltailcontainingacisunsaturation.Thecationicsurfactantsarefromthealkyltrimethylammoniumbromide(Cfamilyhavinglinearalkyltailsrangingfrom6to12carbons.WeshowthatthephasebehaviorandrheologyofNaOA/CTABmixturescanbetunedbyajudiciouscombinationofsurfactanttaillengths.ExperimentalSectionAlkyltrimethylammoniumbromidesurfactants(CBr)with-decyl,and-dodecyltailswereobtainedfromTCIAmerica(allwereof99%purity).Sodiumoleate(-9-octadecanoate,CCOONa),alsofromTCIAmerica,wasof97%purity.Allsurfactantswereusedasreceived.Distilleddeionizedwaterwasusedinpreparingthemixedsurfactantsolutions.Sampleswereequilibratedat25ÉCforphasebehaviorobservations,andphaseboundarieswereidentifiedvisually.Rheologicalexperiments(steadyanddynamic)wereconductedat25ÉConthesingle-phasesamples.TheexperimentswereperformedonaBohlinrheometerusingacouettecell(cupof27.5mmo.d.,bobof25mmi.d.,and37.5mmlength)oracone-and-plateapparatus(40mmdiameter,4Éconeangle).Formildlyviscoussamples,theviscositywasmeasuredusingaCannon-Ubbelohdecapillaryviscometerimmersedinaconstanttem-peraturewaterbathat25ÉC.Flowtimesweresufficientlyhighinallcasessothatkineticenergycorrectionswerenotnecessary.Small-angleneutronscattering(SANS)experimentswereperformedattheNationalInstituteofStandardsandTechnology(NIST)inGaithersburg,MD,ontheNG-3beamline.Theincidentneutronwavelengthwas6…witha15%spread.Quartzcellswith2mmpathlengthswereused,andthesampleswereplacedinatemperature-controlledchambermaintainedat25ÉC.Threedifferentsample-to-detectordistanceswereusedtospanarangeof0.0040.4…inthescatteringvector.Thescatteringspectrawerecorrectedforbackgroundradiation,detectorefficiency,emptycellscattering,andsampletransmission.Thespectrawereradiallyaveragedandplacedonanabsolutescaleusingcalibra-tionstandardsprovidedbyNIST.Thedataareshownintermsoftheabsolutescatteredintensityasafunctionofthescattering)sin(/2),whereisthewavelengthofincidentneutronsandthescatteringangle.PhaseBehaviorandRheology.ConsidermixturesofNaOAandoctyltrimethylammoniumbromide(Catatotalsurfactantconcentrationof3wt%.Becauseofitslong(C)alkyltail,NaOAself-assemblesreadilyinaqueoussolution,anditscriticalmicelleconcentration(cmc)islow(0.06wt%).Ontheotherhand,theshortertailinCTABleadstoamuchhighercmc(3.5wt%).Thus,ataconcentrationof3wt%,CTABdoesnotmicellizeinsolutionwhereasNaOAtendstoformsmallmicelles.Globularorshort,rodlikemicelleshaveaweakeffectontherheologythesolutionsareNewtonian,andtheviscosityriseslinearlywiththevolumefractionofthemicelles()accordingtoistheviscosityofthesolventmedium(forwaterat25ÉC,is1mPas).Theintrinsicviscosity[]is2.5foradispersionofhardspheres,whileforshortrodsbelowtheoverlapconcentration*,[]isafunctionofthemicelleTheviscosityofthe3%NaOAsolutionis1.5s,indicatingthepresenceofshortrodlikemicelles.AddingasmallamountofCTABtoanNaOAsolutionincreasestheviscositydramatically(Figure1a).Thezero-shearviscosityof3%surfactantsolutionsrisesexpo-nentiallywithincreasingCTABfractionandpeaksatavalueof1800Pasforacompositionof70/30NaOA/CTAB.Thisismorethan6ordersofmagnitudehigherthantheviscositiesofthepurecomponentsolutions.Theincreaseinviscosityreflectsthegrowthofmicellesfromshort,nonoverlappingrods(diluteregime,below*)tolong,flexibleworms(semidiluteregime,abovewormsareexpectedtohaveaveragecontourlengthsmorhigher.Itistheentanglementofthesewormsinto (19)Jonsson,B.;Lindman,B.;Holmberg,K.;Kronberg,B.andPolymersinAqueousSolution;JohnWiley&Sons:NewYork, (20)Mukerjee,P.;Mysels,K.J.CriticalMicelleConcentrationsofAqueousSurfactantSystems;NSRDS-NBS-36;U.S.GovernmentPrint-ingOffice:Washington,DC,1971.(21)Larson,R.G.TheStructureandRheologyofComplexFluidsOxfordUniversityPress:Oxford,1999. Figure1.RheologyofNaOA/CTABmixturesasafunctionofcompositionforatotalsurfactantcontentof3wt%:(a)zero-shearviscosity;(b)relaxationtimeandplateauplateauè])(1)Langmuir,Vol.18,No.10,2002Raghavanetal. atransientnetworkthatimpartsviscoelasticitytotheThesynergisticformationofelongatedmicellesin3%TABmixturesisapparentevenwhenthemixturecontainsmostlyCTAB.Forexample,theviscosityofa20/80NaOA/CTABsolutionis5mPas(i.e.,5timesthatofCTABalone),whichsuggeststhatrodlikemicellesofmoderatelength(*)arepresent.Inturn,thisconfirmsthatthecmcofthe20/80mixtureissignificantlylowerthanthecmcofCTABalone;i.e.,asmallamountofNaOAinducestheCTABtocomicellize.The3%NaOA/CTABmixturesalsoshowviscoelasticbehaviorindynamicshear.Dynamicrheologicalmea-surementsoftheelastic()andviscous()moduliasfunctionsoffrequencyyieldvaluesoftheplateaumodulusandtherelaxationtime,whereistheintersectionpointof).Figure1bshowsthattheplateaumodulusisindependentofcomposition,whilepeaksatthecompositionof70/30NaOA/CTAB,atwhichthezero-shearviscosityalsoshowedapeak.DynamicrheologyalsorevealsthatsamplesnearthemaximumareMaxwellfluids,i.e.,fluidswithasinglerelaxationtime.Figure2ashowsthefrequencyspectrumforthe70/30NaOA/CTABsample,andaColeColeplotofthesedata(Figure2b)revealsthesemicircleexpectedofaMaxwellfluid.Steady-shearrheologicaldataforthesamesample(Figure3)showsaNewtonianplateaufortheviscosityatlowshearrates,followedbyshearthinningathighershearrates.Thezero-shearviscosityis1800s,whichcorrelateswellwiththeproductof(100s)(18Pa;Figure3a),asexpectedforaMaxwellfluid.Intheshear-thinningregime,theshearstressapproachesaplateau.Theeffectoftotalsurfactantconcentrationontherheologyisstudiednextforafixedcompositionof70/30TAB(Figure4).Thezero-shearviscosityanonmonotonicbehaviorasafunctionofsurfactantconcentration,peakingataconcentrationofabout4wt%.Forconcentrationsof10%andhigher,dropstoalowandnearlyconstantvalue(thedynamicresponseisweakinthesecases.)Theplateaumodulus,ontheotherhand,monotonicallyincreaseswithconcentration.Thepower-lawexponentforis2.2,whichisexactlythevalueexpectedforentangledwormlikemicelles.Concentrationeffectscanalternatelybeprobedbyvaryingthemixturecompositionatdifferentvaluesofthetotalsurfactantcontent.PlotsofvstheCfractioninthemixtureat3%,6%,and9%surfactantareshowninFigure5.Ineachcasethereisapeakin,andthepeakpositionshiftstocompositionsricherinNaOAwithincreasingconcentration.Fora50/50NaOA/Csample(equimolarratio),ispracticallyunalteredovertheconcentrationrange.Asmallexcessofeitheroneofthesurfactantscausestodropwithincreasingconcentration.OnlyatlowCTABcontentdoestheviscosityincreasewithconcentration.TheeffectofaddedsaltontherheologyisexaminedinFigure6,wheretheviscositiesof3%surfactantsolutions Figure2.Dynamicrheologyofa3wt%surfactantsolutionwiththecomposition70/30NaOA/CTAB:(a)dynamicfre-quencyspectrum;(b)ColeColeplot. Figure3.Steady-shearrheologyofa3wt%surfactantsolutionwiththecomposition70/30NaOA/CTAB.Theviscositytheshearstressareplottedasafunctionoftheshearrate. Figure4.RheologyofNaOA/CTABmixturesasafunctionoftotalsurfactantconcentration.TheratioofNaOA/CTABisheldconstantat70/30.Dataforthezero-shearviscositytheplateaumodulusareshown. Figure5.Zero-shearviscosityofNaOA/CTABmixturesasafunctionofcompositionforthreedifferentsurfactantcon-centrations(3,6,and9wt%).WormlikeMicellesLangmuir,Vol.18,No.10,2002 inthepresenceof0.1MKClarecomparedwiththoseshownpreviouslyforsolutionswithnoaddedsalt.ThepresenceofKClincreasestheviscosityatlowCcontent,whiledepressingitatorbeyondtheoriginalviscositypeak(i.e.,70/30NaOA/C8TAB).Beyondtheequimolarratio,theinthepresenceofKClapproachestheearliervalue.Thus,theviscositypeakapparentlycorrespondstoanoptimalextentofelectrostaticscreening,asdiscussedpresently.Finally,cationicsurfactants(CTAB)ofvaryingalkyltaillengthsareexaminedinmixtureswiththeCNaOA.First,theviscositiesof3%NaOA/CTABmixturesarecomparedwiththosefor3%NaOA/CTAB(Figure7).TheNaOA/CTABsystemalsodescribesaviscositymaximum;however,thevaluesarefarlowerthanthoseoftheNaOA/CTABsystem.Thepeakviscosityfor3%TABmixturesisonlyabout50mPas,i.e.,50timestheviscosityofwater.Thus,decreasingthealkyltaillengthfromeighttosixunitsgreatlyreducesthesynergisminself-assembly.Also,thepeakinisshiftedtoacompositionof60/40NaOA/CTAB,whichisclosetotheequimolarratioforthismixture.Inthecaseof3%NaOA/CTABand3%NaOA/Cmixtures,samplesatintermediatecompositionsseparateintotwoormorephases(Figure8).Thenatureofthesephaseshasnotbeendeterminedsystematically;however,inmostsamples,oneofthephaseslikelyinvolvesalamellarmorphology.Upondilution,manysamplesturnintononviscous,isotropic,one-phasesolutionswithabluethesearelikelytocontainunilamellarvesicles100nmindiameter),andthebluecolorisduetothescatteringoflightbytheseentities.NotealsothatthemultiphaseregionspansaslightlylargercompositionrangefortheNaOA/CTABsystemthanfortheNaOA/TABsystem.Asregardsthesolutionviscosity,divergesasthephaseboundaryisapproachedfromeitherside,withhighervaluesofbeingobservedintheNaOA/TABsystemthanintheNaOA/CTABsystem.Small-AngleNeutronScattering.Tofurtherprobethemicrostructure,SANSmeasurementswereperformedonselectedNaOA/CTABsamplesinDO.SANSspectra)forthreeNaOA/CTABsamplesat3wt%totalsurfactantareshowninFigure9.Thecompositionswerechosentomirrorthenonmonotonicrheologythe70/30samplecorrespondstoaviscositypeakwhiletheothersfalloneithersideofthispeak(therheologyinDOissimilartothatinHO).TheSANSdata,however,showonlyamonotonictrend.The80/20and70/30samplesshowapeakatintermediateanddepressedscatteringatlow,reflectingtheelectrostaticrepulsionsbetweenthechargedaggregatesinsolution(notethatthereisnosaltaddedtoscreentherepulsions).Astheproportionofcationicsurfactantincreases,thepeakbecomeslesspronounced,andforthe50/50sample(closetoequimolarratio),thepeakisabsent.The70/30samplealsoshowsaslightupturninintensityatlow.Thescatteringatisapproximatelyidenticalforthethreesamples.SANSdataforthreedifferentsurfactantconcentrationsatacompositionof70/30NaOA/CTABareshownin (22)Chen,S.H.;Sheu,E.Y.;Kalus,J.;Hoffmann,H.J.Appl.,751. Figure6.RheologyofNaOA/CTABmixturesinthepresenceofsalt.Thezero-shearviscosityisplottedasafunctionofcompositionformixturescontaining0.1MKCl(filledcircles).Thedataintheabsenceofsaltarealsoshownforcomparison(opencircles).Thetotalsurfactantcontentis3wt%inall Figure7.Zero-shearviscosityofNaOA/CTABmixturesasafunctionofcomposition.Forcomparison,dataforNaOA/CTABmixturesarealsoshown.Thetotalsurfactantcontentis3wt%inallcases. Figure8.Zero-shearviscosityasafunctionofcompositionformixturesofNaOA/CTABandNaOA/CTAB.Thetotalsurfactantcontentis3wt%inallcases.Forintermediatecompositionsinbothsystems,thesamplesseparateintotwo(ormore)phases. Figure9.SANSspectrafor3wt%NaOA/CTABsolutionsinO.Thesolidlineisafittothe50/50samplefor0.04…usingamodelformonodispersecylinders.Fromthefit,amicelleradiusof20…isobtained.Langmuir,Vol.18,No.10,2002Raghavanetal. Figure10.Apeakappearsatintermediateinallcases,andthepeakpositiondictatesthecorrelationlengthwhichrepresentsthemeandistancebetweenthechargedcylindricalmicelles.Theshiftinwithincreasingconcentrationmeansthatcorrespondinglydecreases,asexpected.Theothernotableaspectistheupturnatlow,observedinboththe3%and6%samples,butnotinthe9%case.Suchanupturnusuallysignifiestheonsetofattractiveinteractionsorcriticalfluctuations.AnalysisofSANSData.Toobtainstructuralinfor-mation,theSANSdatahavetobefittedtoamodel.Thescatteredintensity,ingeneral,willhavecontributionsfromboththestructurefactor)andtheformfactor)oftheelongatedmicelles.Structurefactoreffectscannotbeneglectedherebecausethemicellarchargeisnotscreenedcompletelyandthemicellarconcentrationissemidilute.AnalysisusingtheindirectFouriertransform(IFT)methodshowsthat)effectsarepresentevenforthe50/50sample,forwhichthereisnointeractionpeak.Thisisalsowhytheslopeofatlowforthissampleisnotequalto1aswouldbeexpectedforrodlikemicelles.Athigh)isapproximately1,thedataadmittoanalysisbasedontheformfactor).Forthe50/50sample,afitofthehighdatatotheformfactorformonodispersecylinders(solidline,Figure9)yieldsamicelleradiusofca.20….Themicelleradiusisfoundtovarynegligiblywithcomposition,anditsvaluecompareswellwiththelengthofanextendedoleylchain(25…).Moresophisticatedanalysisofthehighdatausingthecross-sectionIFTmethodconfirmsthatthemicellesarehomogeneouswithanaverageradiusof191….However,theanalysisalsosuggestsapolydispersityof915%inthemicelleradius,whichmayactuallyimplyslightdeviationsfromacircularcrosssection,asreportedforothercationic/anionicsystems.Takentogether,theSANSanalysisconfirmsthepresenceofmicellarchains,whosecrosssectionmaybecircularorslightlyellipsoidal.Thelengthofthesechainscannotbeobtainedunambigu-ouslywithoutaccountingfor)effects.WormlikeMicellesinMixedSurfactantSystems.FromtheSANSmeasurementsaswellasfromtherheology,itisevidentthatlongwormlikemicellesarepresentinthesemixedsurfactantsolutions.SignaturesofwormlikemicellesincludetheMaxwellianbehaviorindynamicrheology(Figure3),andthehighviscosityandstressplateauinsteady-shearrheology(Figure4).observations(throughcross-polars)alsorevealthattheseisotropicsolutionsbecomebirefringentunderflow,whichisanothersignatureofelongatedmicelles.Thepresenceofwormlikemicellesisfurtherconfirmedbycryo-TEM,asreportedelsewhere.Anoteworthyaspectoftheseviscoussolutionsisthatthemajorcomponentisananionicsurfactant(NaOA).Mostreportsofhighlyviscouswormlikemicellarsolutionshaveinvolvedalong-tailedcationicsurfactant.surfactantshavebeenshowntoformelongatedmicellesbutthezero-shearviscositiesofthosesolutionswerequitelow(10Pas),andmoreover,themicellesgrewonlywhenhighconcentrationsofsimplesalts(e.g.,KCl)wereadded.HerewehaveshownthatitissufficienttoaddafewmMofCTABtotheanionicsurfactantNaOAtoproducehighsolutionviscosities(comparabletoaddingafewmMofthehydrotropesodiumsalicylatetoacationicsurfactantsuchasCMoreover,thezero-shearviscosityisremarkablyhigh1000Pas)formanyoftheNaOA/CTABsamples.Thesevaluesarecomparabletothehighestviscositiesobtainedusingcationicsurfactantsatequivalentsurfactantcon-Whydolongwormlikemicellesformincationic/anionicmixtures?Theoriginliesinthestrongelectrostaticinteractionsbetweentheoppositelychargedheadgroups,asdiscussedabove.Torationalizemicrostructures,itisusefultoconsiderthesurfactantpackingparameter,whereisthetailvolume,thetaillength,andtheareaperheadgroup.Thepairingofoppositelychargedheadgroupsdecreasestheheadgroupareatherebyincreases.Thus,asthefractionofoppositelychargedsurfactantinthemixtureincreases,variesfromforasinglesurfactantto1fortheequimolarmixture.Correspondingly,themicrostructurechangesfromspheri-calmicellestowormlikemicellestobilayerstructures.ThisentireprogressioncanbeobservedforNaOA/CTABandNaOA/CTABmixtures.Thestronginteractionbetweencationicandanionicheadgroupsalsoinducesthecatanionicsalttoprecipitateinmixturesclosetotheequimolarcomposition.larlywhenthetaillengthsofthecationic()andanionic)surfactantsarenearlyequal,thetailspackwellinacrystallinelattice,andtheprecipitateisexceptionallyConversely,asymmetrictailsdonotpackef-ficiently,therebyinhibitingsaltformation.Here,thecatanionicprecipitateisformedattheequimolarcom-positionformixturesofNaOA(18)withCTABandTAB.However,fortheNaOA/CTABandNaOA/CTABsystems,noprecipitateisformed,andthesamplesarehomogeneousmicellarsolutionsoverthecompositionrange.Apparently,theasymmetrybetweenthetailsinthelattercasesissufficienttoinhibitprecipitationofthecatanionicsalt.Similartrendsinphasebehaviorhavebeenreportedpreviouslyforothercationic/anionicsurfactantmixtures.Forexample,mixturesofCTABandsodiumalkyl (23)Bergstrom,M.;Pedersen,J.S.J.Phys.Chem.B,8502.(24)Bergstrom,M.;Pedersen,J.S.Phys.Chem.Chem.Phys.,4437.(25)Supportinginformationforthispaper. (26)Edlund,H.;Norgren,M.;Raghavan,S.R.;Kaler,E.W.Manuscriptinpreparation.(27)Gamboa,C.;Sepulveda,L.J.ColloidInterfaceSci.(28)Almgren,M.;Gimel,J.C.;Wang,K.;Karlsson,G.;Edwards,K.;Brown,W.;Mortensen,K.J.ColloidInterfaceSci.,222.(29)Israelachvili,J.IntermolecularandSurfaceForces;AcademicPress:SanDiego,1991. Figure10.SANSspectrafor70/30NaOA/CTABsolutionsinOatdifferentsurfactantconcentrations.Thedataforthe3%sampleareabsolutevalueswhiletheremainingcurvesareoffsetbyfactorsof5forclarity.WormlikeMicellesLangmuir,Vol.18,No.10,2002 carboxylatesformasinglemicellarphaseoverthecompositionrangefor8butseparateintotwocoexistingliquidphasesforMixturesofCTACandsodiumnonanoate(9)formasinglemicellarphaseatallcompositionsupto40wt%surfactant.Thus,asymmetryinthealkyltaillengthspromotesastablemicellarphaseattheexpenseofbilayersandsaltprecipitates.Foraconstantratiooftaillengths,increasingthetotalnumberofcarbonsinthetails(i.e.,)enhancesprecipitation.Forexample,bilayersandmultiphaseregionsarepredominantintheternaryphasediagramofCTABandsodiumoctylsulfate(ThecomparativestabilityoftheNaOA/CTABsystem(26)isprobablyduetothecisdoublebondintheCofNaOA,whichhinderscrystalpacking.MixedSurfactantPhaseBehaviorBasedontheFromamacroscopicstandpoint,interactionsbetweensurfactantscanbequantifiedusingtheparameterfromregularsolutiontheory.Formixturesoftwosurfactants,1and2,isdefinedby,andarethemolarfreeenergiesofinteractionfor11,22,and12pairs,respectively.Idealbehaviorimpliesnonetinteractionsbetweenthesurfac-tants,i.e.,andhence0.Ontheotherhand,attractiveinteractionsbetweenthesurfactantsimpliesthatwillexceedwillhencebenegative.Inthatcase,thecmc'softhemixtureswillbelowerthanthoseofthepuresurfactants.parameterforamixtureoftwosurfactantscanbeestimatedfromcmcdataasafunctionofmixturecomposition.Asexpected,cationic/anionicmixturesshownegativevaluesof,andthemorenegativethe,thestrongertheattractiveinteractions.Asthetotalnumberofalkylcarbons()increases,becomesmorenegative.Forexample,symmetricmixturesofCNa(alkylsulfates,anionic)andCTABs(cationic)exhibit18.5for10,and25.5forInthisseries,precipitationofcatanionicsaltoccursattheequimolarcompositionfor10and12,butnotfor8.Somoderatelynegativeleadstomicellargrowth,whilehighlynegativeleadstobilayersandprecipitation.WecannowclassifythebehaviorofNaOA/Cmixturesfordifferentvaluesofintermsofexpectedvaluesof.First,considertheNaOA/CTABsystemwherethereisalargedisparitybetweenTAB,duetoitsweakamphiphilicity(itdoesnothaveacmc),shouldbeviewedasahydrotropicsaltratherthanasurfactant.Consequently,NaOAandCTABwillinteractratherweakly(expectedtobeca.5).AddingCTABtoNaOAwillthereforehaveaweakeffectonthesurfactantpacking,whichmayincreaseslightlyfromtoless.Thiscorrelateswithweakmicellargrowthandhenceasmallriseinviscosity(Figure9).Notethatthepeakinviscosityoccursneartheequimolarratio,wherethemicellarchargeisscreenedthemost.SimilarresultshavebeenreportedforCTAC/sodiumnonanoatemix-wherealsothepeakviscositywaslow(30mPaat10wt%surfactant),andthepeakoccurredattheequimolarcomposition.Next,considertheNaOA/CTABandNaOA/Csystems.Here,arerelativelyclose,andsothesurfactantpairsshouldinteractstrongly(expectedtobearound20).Inturn,thepackingparameterattainvaluescloseto1overarangeofmixturecomposi-tions.Thisexplainstheexistenceofbilayerphasesinthesemixtures(Figure10).Notethatbilayersoccuroverawidercompositionrangeinthecaseofthelonger-tailedCTAB.Thecatanionicsaltalsoprecipitatesattheequimolarcompositionineachcase.Theviscosityincreaseonapproachingthemicellebilayerphaseboundaryreflectsachangeinatthosecompositions.Finally,considertheNaOA/CTABsystem.AsintheTABcase,thearesufficientlydifferentsothatthemixturesareabletoavoidphaseseparationorprecipitation.However,CTABisfarmoreamphiphilicthanCTABandshowsacmcaround3.5wt%.Therefore,relativelystronginteractionsshouldoccurbetweenNaOAandCTAB(theparameterisexpectedtobearound15).Accordingly,thepackingparameterwillassumevaluesaround(butwillnotapproach1).Thiscanexplainthegrowthofgiantwormlikemicelleswithinthesingle-phase,homogeneoussolutions.Fromarheologicalper-spective,suchamixtureprovidestheoptimalsynergism.InterpretationofViscosityMaxima.Wenowad-dresstheviscositymaximainNaOA/CTABmixturesasafunctionofcomposition(Figures1,5,and6).Generally,thesepeaksappearatlowCTABcontents,wellbeforetheequimolarratio(e.g.,thepeakat3%surfactantoccurswhenthereisroughlyoneCTABmoleculeforeverytwoNaOAmoleculesinthemicelle.)Thus,thepeakdoesnotreflectanoptimalextentofchargeneutralization.Whythendoestheviscositygothroughamaximum?Aplausiblehypothesisisthatthereisatransitionfromlineartobranchedmicellesatthepeak.Accordingtothishypothesis,micellesgrowlinearlybeforethepeakcom-positionandarelongestatthepeak.Thereafter,themicellestendtoformbranchesratherthancontinuegrowingaxially.Asbranchingproceeds,theviscositydropsbecausethebranchpointsarenotfixedbutarefreetoslidealongthemicelle(therebyprovidinganadditionalmodeforstressrelaxation).Anattractivehypothesisisthatthepointofmicellarbranchingrepresentsadelicatebalanceintheelectrostaticcharacterofthesystem.Forbranchingtooccur,a3-foldmicellarjunctionmustbeenergeticallymorefavorablethanamicellarend-cap.Thisseemstobethecasewhentheelectrostaticinteractionsaresufficientlyscreened.Notethatelectrostaticscreeningcanoccurintwowaysinthesesurfactantmixtures:(a)duetopairingoftheoppositelychargedheadgroupswithinthemicelle,whichreducesthechargedensityatthemicellarsurface,and(b)duetothereleaseofsurfactantcounterionsintothesolution,whichservestoreducetheelectrostaticdoublelayeraroundeachmicelle.Theformereffectincreasesalongaconcentrationpathfrompuresurfactanttoequimolarmixtures,whilethelattereffectincreaseswithtotalsurfactantconcentration.Ifindeedtheviscositypeakcorrespondstotheonsetofmicellarbranching,theshiftinthepeakpositionwithsurfactantconcentration(Figure6)andsaltaddition(Figure7)canberationalized.Addingsalt(KCl)uniformlyincreasestheionicstrength,sotheextentofelectrostaticscreeningisincreasedatallcompositions.Theonsetofmicellarbranchinghenceoccursforalowermicellarsurfacecharge,i.e.,foralowerCTABcontent.Increasingthesurfactantconcentrationincreasestheconcentrationofcounterionsreleasedintothesolution.Theeffectisthussimilartosaltaddition,andbranchingisinitiatedatalowerCTABcontentforahighersurfactantconcentration. (30)Lequeux,F.Reptationofconnectedworms.Europhys.Lett.,675. Langmuir,Vol.18,No.10,2002Raghavanetal. Attractiveinteractionsinmixturesofcationicandanionicsurfactantsaregovernedchieflybythesurfactanttaillengths.Generalizingtheresultsfromthisstudy,threepossiblescenarioscanbedistinguishedintermsofthetaillengths.First,ifthereisonelongandoneveryshorttail(e.g.,NaOATAB),weakinteractionsensuebetweenthesurfactants,leadingtoweakmicellargrowth(andasmallriseinviscosity).Second,ifbothtailsarelong(e.g.,NaOATAB),strongattractiveinteractionsoccur,leadingtoprecipitationofthecatanionicsaltattheequimolarcompositionandtheformationofbilayerstructuresoverasignificantwindowofcompositions.Last,andmostimportant,isthescenariowhereonetailislongandtheotherisofmoderatelength(e.g.,NaOAinthiscase,theattractiveinteractionsarestrongenoughtocausedramaticmicellargrowth,butnotstrongenoughtoinducebilayerstructuresorphaseseparation.Theresultisadramaticenhancementoftherheologicalpropertiesatintermediatecompositions.WeacknowledgeNISTforfacili-tatingtheSANSexperimentsperformedaspartofthiswork.HelpfuldiscussionswithDr.HakanEdlund(Mid-SwedenUniversity,Sundsvall)arealsoacknowledged.SupportingInformationAvailable:AnalysisofSANSdatausingtheindirectFouriertransformationtechnique.ThismaterialisavailablefreeofchargeviatheInternetatWormlikeMicellesLangmuir,Vol.18,No.10,2002