<%BANNER%>

In-situ Spectroscopic Studies of Single-Walled Carbon Nanotubes and Conjugated Polymers in Electrochromic Devices


PAGE 2

Thepastveyearshavebeenanintenselychallenging,howeverenjoyable,lifejourney.Duringthattime,Ihavebeenveryfortunatetohavebeensurroundedbyexceptionalpeoplethatgreatlyinuencedmycareerandatthesametimeinspiredme.TherstpersontowhomIwouldliketoexpressmysinceregratitudeismyresearchadvisor,ProfessorDavidB.Tanner,whowelcomedmeinhisgroup,supportedmeandencouragedmefromtheverybeginning.ThetimeIhavespentworkingunderhissupervisionhasprovedtobeaninvaluableexperience.Furthermore,Iwouldliketothankhimforhistrust,andforallowingmetopursuearesearchpath,whichrequiredclosecollaborationwiththeChemistryDepartment. Inthatdierentuniverse,apersonthatalsohadamajorimpactonmywork,andtowhomIwouldliketoexpressmysinceregratitude,isProfessorJohnR.Reynolds,whoinrealityhasbeenasecondadvisortome.Forthelastthreeyears,hehaswelcomedmeinhisgroup,andhehastirelesslysupportedme,guidedme,beingalwaysinterestedinandappreciativeofmywork.IhavetoadmitthatIfeelextremelyprivilegedhavingcollaboratedwiththesetwoexcellentresearchersand,aboveall,exceptionalpeople. Iwouldalsoliketoexpressmyappreciationtomanypeople:ProfessorRichardWoodard,forhishelporganizingmytransitionfromGreeceandwithmyinitialad-justmenttoanewenvironment;ProfessorKatalinKamaras,forherhelpwiththeex-periments,dataanalysis,guidance,input,andusefuldiscussionsinvolvingthewholeprojectdescribedinChapter6;also,ProfessorAndrewG.Rinzlerforprovidinghighqualitysamples,andforalwayspatientlyansweringallmyquestionsconcerningthesubjectofnanotubeswhichiscoveredinChapters6and7.Iamalsoindebtedtomysupervisorycommittee,ProfessorsAlanT.Dorsey,ArthurF.Hebard,andPierreSikivie,fortheirinterestinservingonmycommittee.ii

PAGE 3

FromthePhysicsDepartment,IwouldliketothankmycollaboratorsfromProf.Rinzler'sgroup:DrZhihongChenforprovidingthesamplesfortheresearchoutlinedinChapter6andformanyusefuldiscussions,ZhuangchunWuforprovidingthesamplesfortheresearchoutlinedinChapter7,andalsoJenniferSippelOakleyforhelpwiththeAFMimages.FromProf.Kamaras'group(ResearchInstituteforSolidStatePhysicsandOptics,HungarianAcademyofSciences,inBudapest,Hungary),IthankFerencBorondicsforcollaborationontheresearchprojectoutlinedinChapter6.Lastbutnotleast,IwouldliketothankthepeopleinProf.Tanner'sgroup,mypastandpresentcolleagues,thepeopleIworkedsidebysidewiththroughoutmygraduateyears,fortheircooperation,usefuldiscussions,andtheirfriendship.IwouldalsoliketothankMattCornick,anundergraduatestudentwhojoinedourgroupthroughtheREUprogram,forhishelpwiththeexperimentsoutlinedinChapter5,andNathanHestonforhishelpwiththelastmid-infraredexperimentoutlinedinChapter7. Iwouldalsoliketoacknowledgethehelpofthemembersofthemachineshop,especiallyMarcLinkandJohnVanLeer,thepeopleintheelectronicsshop,andthecryogenicsteam.iii

PAGE 4

ThisworkisdedicatedtothememoryofMariaB.Panousi.iv

PAGE 5

TABLEOFCONTENTS Page ACKNOWLEDGMENTS.........................ii ABSTRACT..............................viii CHAPTER 1INTRODUCTION...........................1 2CONDUCTING,ORCONJUGATED,POLYMERS............5 2.1Non-conjugatedandConjugatedPolymers.................5 2.2ClassicationofConjugatedPolymers....................9 2.2.1DopingMechanismsinDGSPsandNDGSPs............11 2.3TheoreticalModels..............................17 2.4ConductivityinConjugatedPolymers(CPs)................19 2.5Metal-Insulator(M-I)Transition.......................21 2.6DopingInducedPropertiesinConjugatedPolymers............22 2.7DopingMethodsforConjugatedPolymers.................24 2.8FundamentalsofElectrochromism......................28 2.9SynthesisMethodsofCPs..........................31 2.10CharacterizationMethodsofElectrochromicCPs.............33 2.11ElectrochromicDevices(ECDs)BasedonCPs...............36 2.12GeneralApplicationsofCPs.........................38 3THINFILMOPTICS........................40 3.1OpticalProcessesandOpticalConstants..................40 3.2InteractionofElectromagneticWaveswithMatter.............42 3.3LightPropagationThroughaPlanarInterface...............49 3.4LightPropagationThroughaSingleLayerStructure...........53 3.5LightPropagationThroughaMulti-LayerStructure............58 3.5.1MatrixMethod............................61 3.6Kramers-Kronig,orDispersion,Relations..................62 3.7ModelsfortheDeterminationofOpticalConstants............64 3.7.1LorentzModel.............................64 3.7.2DrudeModel.............................68 3.7.3Drude-LorentzModel.........................71 3.7.4SumRules...............................72 v

PAGE 6

4INSTRUMENTATIONANDEXPERIMENTALTECHNIQUES......73 4.1DryBox....................................74 4.2ElectrochemicalMethods...........................76 4.2.1ElectrochemicalPolymerization...................77 4.2.2CyclicVoltammetry(CV)......................78 4.3OpticalMethods...............................79 4.3.1Spectroelectrochemistry.......................79 4.3.2InterferometricorFTIRSpectrometer...............81 4.3.3MonochromaticSpectrometer....................93 5REFLECTIVEELECTROCHROMICDEVICES(ECDS).........104 5.1ECDsFabrication...............................105 5.2In-situReectanceMeasurementsandAnalysis..............109 5.3ResultsandDiscussion............................110 5.4EnhancingthePerformanceofECDs....................114 5.5Conclusions..................................120 6FREE-STANDINGSINGLEWALLEDCARBONNANOTUBEFILMS..122 6.1CarbonNanotubes..............................122 6.1.1ElectronicStructureofCarbonNanotubes.............125 6.1.2DensityofStates(DOS)ofSWNTs.................127 6.1.3CarbonNanotubeSynthesisandPurication............131 6.1.4GeneralApplicationsofCNs.....................132 6.2Free-StandingSWNTFilms: Transparency,andOpticalConductivity..................133 6.2.1SamplePreparation..........................133 6.2.2In-situTransmittanceMeasurements................136 6.2.3OpticalConstants,DataAnalysisandModelFit..........146 6.2.4InfraredSpectralWeights......................152 6.2.5Conclusions..............................152 7TRANSMISSIVE/ABSORPTIVEECDEVICES.............155 7.1SamplePreparation..............................155 7.1.1PuricationofLaboratoryChemicalsandMaterials........155 7.1.2ElectrodePreparation........................157 7.1.3ElectrochemicalPolymerization...................157 7.1.4GelElectrolytePreparation.....................159 7.2PEDOTonSWNTElectrodes........................160 7.2.1SpectroelectrochemistryofPEDOT.................160 7.2.2ThicknessDeterminationofPEDOTonSWNTElectrodes....164 7.2.3AtomicForceMicroscopy(AFM)Images..............166 7.3PBEDOT-Hx 2 -Pyr-PyronSWNTElectrodes...............168 7.3.1CyclicVoltammetryofPBEDOT-Hx 2 -Pyr-Pyr...........168 7.3.2SpectroelectrochemistryofPBEDOT-Hx 2 -Pyr-Pyr.........169 vi

PAGE 7

7.4Transmissive/AbsorptiveECDeviceConstruction.............172 7.5In-SituTransmittanceMeasurementsofECDevices............174 7.5.1AnalysisoftheDierentLayersintheECDevice.........175 7.5.2AnalysisofaPEDOTECDevice(withhole)............176 7.5.3ResultsonPBEDOT-Hx 2 -Pyr-PyrECDevice(withhole).....183 7.6TowardstheconstructionofaDualDevice.................189 7.6.1CyclicVoltammetryofPBEDOT-Hx 2 -Pyr-Pyrdualdevice....189 7.6.2In-SituAbsorbanceMeasurements..................191 7.6.3SwitchingTime............................195 7.7Conclusions..................................197 APPENDIX ANEXTGENERATIONOFREFLECTIVEECDSUSINGMICROPOROUSGOLD ELECTRODES...........................200 BDRUDE-LORENTZFITTINGPARAMETERS.............205 CPBEDOT-HX 2 -PYR-PYR......................208 REFERENCES.............................210 BIOGRAPHICALSKETCH.......................216 vii

PAGE 8

Ourinitialfocushasbeentheoptimizationoftheperformanceofreectiveelec-trochromicdevices.In-situreectancemeasurementsandextensivesubsequentanaly-sisenabledustoconstructdeviceswithsignicantlyenhancedperformance.VariablereectiveelectrochromicdevicesbasedonPProDOT-Me2andPBEDOT-NMeCzwereoptimizedtoexhibitcontrastratiosof60-70%inthemid-infraredregion.Inthelatterregion,twostrongabsorptionpeaks,O-HstretchingmodeandC-Hstretchingmode,werehinderingtheperformanceofthedevices.Anoveltechniquewasdevelopedinordertobuildwater-freedevices.Inadditionanewsampleholderwasdesignedthatpermitscontroloftheelectrolytegelthickness,andreducestheC-Hsignature. Buildingupontheaforementionedtechniques,wedesignedaninfraredtransmis-sive/absorptivetypeofelectrochromiccell.Theconstructionofsuchadevicehadbeenpreviouslyimpossibleduetotheabsenceofasuitable,infraredtransparent,conductingmaterial.Anextensivestudyofsingle-walledcarbonnanotubesasfree-standinglmsshowedthattheselmshavethehighesttransmittanceamongtransparentconductorsinthe2-5mspectralrange,atverylowsheetresistance.Transmittancemeasure-mentsonpuried(p-doped)andvacuumannealed(de-doped)carbonnanotubelmsviii

PAGE 10

Themainfocusofthisworkistheopticalcharacterizationofconjugatedpoly-mersindierentelectrochromicdeviceplatforms,inaneorttooptimizetheperfor-manceofthesedevices.Furthermore,anattempttodeveloptherstinfraredtrans-missive/absorptiveelectrochromicdevice,ledtoextensivestudiesonsingle-wallcarbonnanotude(SWNT)lms.Thismaterialprovedtobethebestcandidateforthereplace-mentoftheconventionalindium-dopedtinoxide(ITO),whichservedastheconductinglayerinelectrochromicdevices,fortheinfraredregion.Preliminaryresultsoftheper-formanceofthistypeofdevicesarealsopresentedinthiswork. Morespecically,thesecondchapteroersanintroductiontothetheoreticalback-groundthatisnecessaryfortheunderstandingofconjugatedpolymersystems.Thisincludesapresentationonthedierenttypesofconjugatedpolymersandtheiruniquecharacteristics,suchasthedopingmechanisms,andtheinducedstructuralandopti-calpropertychanges.Abriefdescriptioninthechargetransportmechanisms,andthemetal-insulatortransitionoccurringinconjugatedpolymersisalsogiveninthischapter.Dierentpolymerization,doping,andcharacterizationmethods,andseveraltheoreticalmodelsthatattempttoexplainthebehaviorofthesesystemsareintroduced.Finally,thephenomenonofelectrochromismisdened,andgeneralapplicationsbasedoncon-jugatedpolymersarepresented. Thethirdchapterisareviewofthebasiclawsthatgoverntheinteractionofelectro-magneticwaveswithmatter.Opticalprocessesandcommontechniquesforextractingtheopticalconstants,whichcharacterizeeachsystem,areintroducedandexplained.Lightpropagationthroughdierentstructuresisdescribed,andtheformulasforthecalculationofthereectanceandtransmittancethroughthesestructures,aswellastherelatedopticalparametersarederived.Thechapterconcludeswiththepresentationof1

PAGE 11

Thefourthchapterintroducestheinstrumentation,andexperimentaltechniquesusedthroughoutthiswork.Adescriptionofsystemssuchasthedrybox,orglovebox,andthedierenttypesofspectrometers,interferometricandmonochromatic,isprovided.TheconceptsofFourierspectroscopyarepresented.Techniquesfortheproductionand/orcharacterizationofsamples,suchaselectrochemicalpolymerizationandcyclicvoltammetryareintroduced. Inchapterve,thefabricationofreectiveelectrochromicdevicesispresented.Detailsonthedesign,theassembling,andthematerialsusedareprovided.In-situreectancemeasurements,andanalysisoftheexperimentalresultsarediscussed.Theproblemsthatwereencountered,andthewaystoovercometheminaneorttoenhancetheperformanceofthesedevicesarealsopresented.Finally,adesignofanewtypeofsampleholder,whichispartoftheproblemsolutionisdescribedhere. Chaptersixincludesanintroductiononthetheoreticalbackgroundthatisneces-saryfortheunderstandingofcarbonnanotubesystems.Thisincludesapresentationonthedierenttypesofnanotubesandtheiruniquecharacteristics,suchasthede-pendenceoftheirelectronicstructureonthetubediameterandwrappingangleofagraphenesheet,withnodopingimpuritiespresent.Furthermore,theelectronicdensityofstates(DOS)incarbonnanotubesisnotcontinuous,asitisingraphite,butitdi-videsintoaseriesofspikescausedbythequantumconnementofelectronsintheradialandcircumferentialdirections.ThesespikesareknownasvanHovesingularities,andtheycompriseatypicalsignatureof1Dsystems.Typicalsynthesis,puricationtech-niquesandgeneralapplicationsofcarbonnanotubesarementioned,aswellasabriefdescriptionofthesamplepreparation.Transmittancespectraofas-prepared(puried),p-doped,andvacuumannealed,de-doped,free-standingSWNTlmsofdierentthick-nesses,arepresentedfromthefarinfraredthroughtheultravioletregion,atdierenttemperatures.TheanalysisofthedataincludestheuseofKramers-Kronigrelationsfor

PAGE 12

InChapterseven,transmissive/absorptiveECdeviceswerebuiltusingtwodier-entelectroactivepolymersand,forthersttimetothebestofourknowledge,SWNTlmswereusedastheconductivelayersinECdevices.Initially,weconductedacompar-isonstudybetweentheperformanceofSWNTlms,usedastheconductiveelectrodes,andthewellstudiedandextensivelyusedITO.Cyclicvoltammogramsandabsorbancemeasurementsinsolutionforbothpolymersusedinthedeviceswereperformed,andconclusionweredrawn.Furthermore,twodierentcongurationwereused,andtrans-mittanceorabsorbancemeasurements,dataanalysisbasedontheDrude-Lorentzmodel,aswellasstabilitystudieswereperformed.Therstconguration,isanECdevicewithholeonthecounterelectrode,employedPEDOTastheelectroactivelmwhichcanbeswitchedbetweenaneutralandap-dopedstate.Thesecondconguration,isadualECdevice,employedPBEDOT-Hx2-Pyr-Pyrastheelectroactivelmwhichisoneofafewpolymersthatcanbeswitchednotonlybetweenneutralandp-dopedstates,butalsobetweenneutralandn-dopedstates.Theconstructedtransmissive/absorptivedualECdevicethatexhibitselectrochromicchangesfromthevisibletotheinfraredregion.Thisisthersttransmissive/absorptivedualECdeviceexhibitselectrochromismintheinfraredspectralregion.Furtheroptimizationstudiesareneededforbothtypesofde-vices.Theconclusionsthatweredrawnfromthisextensivestudyarepresentedattheendofthechapter. Inspectroscopicstudies,dierentenergyunitsareoftenusedfordierenttech-niques,dierentpartsoftheelectromagneticspectrum,andbetweendierentdisciplines.Ingeneral,themostcommonlyusedunitfortheinfraredregionofthespectrumisthefrequencyorwavenumber,expressedincm1,whereasforthevisibleregionofthespec-trum,isthewavelengthexpressedinnm.Throughoutthisworktheseunitsareused

PAGE 13

Table1.1:Relationbetweenenergyunits. 1cm10.124meV1meV8.0658cm1

PAGE 14

Thediscoveryofconductingpolymersin1977byAlanJ.Heeger,AlanG.Mac-Diarmid,andHidekiShirakawa[1]createdanewresearcheldthathasbeenstudiedintensivelysince.Althoughconductingpolymershadbeeninitiallyintroducedin1862whenpoly(aniline),orPANI,wassynthesized[2]itwasthecollaboratingeortin1977ofthe2000Nobellaureates[3{7]whichledtothediscoveryofpoly(acetylene)(PAc),thatinvigoratetheinterestoftheresearchcommunity.Thisworkshowedthatorganicconjugatedpolymershavetheabilitytobedopedoverthefullrangefrominsulatortometal,andoeredthepromiseofanewtypeofpolymers:materialswhichexhibittheelectricalandopticalpropertiesofmetals,orsemiconductors,whileretainingtheattrac-tivemechanicalpropertiesandprocessingadvantagesofpolymers.Thisnewgenerationofpolymerscreatedaneweldofinterdisciplinaryresearch,ontheboundarybetweenchemistryandphysics,andnewenormouspotentialapplications.2.1Non-conjugatedandConjugatedPolymers

PAGE 15

Conducting,orconjugated,polymersarethemostrecentgenerationofpolymers,andtheirelectroniccongurationisfundamentallydierent.Inconjugatedpolymers,carbonatomshavethefollowinghybridization:sp2pz,whichleadstothreeequivalent-bondsandone-bondfromtheremainingpzatomicorbital.The-bonding,whichoccurswhentwopzorbitalsofsuccessivecarbonatomsalongthebackboneoverlap,re-sultsinelectrondelocalization,andhenceinchargemobilityalongthepolymerchain[8].Thisdelocalizationisresponsiblefortheuniquepropertiesofconjugatedpolymers.The-gap,orenergygap,Eg,isrelativelysmall,i.e.,1-4eV,andthisisthereasonforthesemiconductingbehaviorofconjugatingpolymers.Therefore,thepolymerchainsymmetryplaysanimportantroleindeterminingtheelectronicstructure,andconse-quentlycarefuldesignofthemonomerunitalongwithdopingcanresultinsystemswithmetallicproperties.ThebondformationofconjugatedpolymersisshownschematicallyinFigure2.1. Figure2.1:Formationof-bondbystrongoverlappingoftwosp2-orbitalsandformationof-bondsbyweakoverlappingofthepz-orbitalsoftwosuccessivecarbonatoms. Eachcarbonatomalongthebackboneofconjugatedpolymershasoneunpaired-electron,theorbitalofwhichoverlapsstronglywiththeorbitalsofthenearestunpaired-electronsandweaklywithorbitalsofunpaired-electronsindierentpolymerchains.Therstoneisknownasintrachaininteractionandthesecondasinterchaininteractions.Thesebonds,stronginsidethepolymerchainandweak,vanderWaalstype,between

PAGE 16

However,accordingtoPeierl(1955)thegroundstateofsuchaone-dimensionalmetalisunstablewithrespecttoastructuraldistortion,whichresultsinthecreationofalternatingdoubleandsinglebonds,asshowninFigure2.3.Asaresult,aspontaneoussymmetrybreakingoccurs,andanenergygapopensattheFermilevelrenderingthematerialsemiconductor[9].Thesesymmetrybreaking,orPeierl'sdistortion,doublestheunitcell,asshowninFigure2.4,andconcentratesthe-electronsbetweenalternatingpairsofcarbonatoms.Thisisconsistentwithconjugatedpolymershavingalternatingsingleanddoublebonds,orlongerandshorterbonds(1.446Aand1.346Arespectively

PAGE 17

Figure2.3:Peierlstransitionordimerization,trans-poly(acetylene)isshownasanex-ample. Theloweringofthesymmetrylowerstheenergyoftheoccupiedstatesandsta-bilizesthedistortion.Thus,thebandissplitintotwosubbands,afullyoccupiedband(alsocalledtheHOMO:HighestOccupiedMolecularOrbital,valenceband,or

PAGE 19

Forthesecondgroupofconjugatedpolymers,thesocallednon-degenerategroundstatepolymers(NDGSPs),theinterchangeofthedoubleandsinglebondsleadstotwoenergeticallyinequivalentcongurations.InFigure2.7poly(para-phenylenevinylene)isshownasanexampleinwhichthequinonoid(orquinoid)structure(B)islessstable

PAGE 20

Morespecically,upondopingofconjugatedpolymers,dierenttypesofexcita-tionsoccurdependingonwhichofthecategories,DGSPsorNDGSPs,thepolymerbelongs.Fordegenerategroundstatepolymers,e.g.,trans-polyacetylene,theintro-ductionofelectrons,and/orholes,tothepolymerchain,createsadomainwallthatseparatesregionsofdierentbondingstructure,i.e.,phaseAandBFigure2.8.Theseexcitationswereinitiallycalled\mists"butlater,inviewofthefactthatthedomainwallisanonlinearshapepreservingexcitationwhichpropagatesfreelyalongthepoly-merchain,theywerecalled\solitons"[9].Solitonsaretopologicalexcitations,andsincetheyconvertphaseAstructuretophaseBstructure,andviceversa,inaperfectinnitechaintheycanonlybecreatedordestroyedinpairs. Whilethesolitonhasanobviouseectonthelatticedistortionpatternofthepolymerchain,italsohasaremarkableeectontheelectronicstructure.Thelocalizeddistortiongivesrisetoasinglelocalizedelectronicstateinthemiddleoftheenergybandgapregion.Thismid-gapstateisasolutiontotheSchrodingerequationinthepresenceofastructuraldistortionandtherefore,canaccommodate0,1,or2electrons.

PAGE 22

Indegenerategroundstatepolymers,thesolitonexcitationisdelocalizedandfreetoextendoverseveralmonomerunits.Therefore,thechangefromonestructuralphasetotheotherisnotabruptasshowninFigure2.9,butrathergradualovertheextendedsolitonwherethelengthsofthesingleanddoublebondsarechangingslowlyuntiltheyarereversed.Ithasbeenshown[11]thatintrans-polyacetylenethesolitonexcitationextendsapproximatelyoverfourteenmonomerunits. Liftingthedegeneracyofthegroundstateenergyleadstosignicantchangesnotonlyforthegroundstatepropertiesbutalsofortheformationoftheexcitations.Aconsequenceofthedegeneracyliftingisthatsolitonsarenotthestableexcitationsanymore.Thus,upondopingthegroupofnon-degenerategroundstatepolymers,e.g.,poly(para-phenylenevinylene)orpoly(thiophene),formsadierenttypeofexcitations,orchargestoragespecies,calledpolaronsandbipolarons.Polaronsandbipolaronsarenon-topologicalbecause,incontrastwithsolitons,bothsidesofthechainareinthesamebondingphase(AorB)whentheyarecreatedFigure2.11. Theseexcitationsaresimplychargesinanextendedlattice,whicharestabilizedbyalocaldistortion.Thislocalizeddistortiongeneratestwosymmetricallyplaced,with

PAGE 23

Inmoredetail,theformationofachargedpolaron,orradicalcationoranion,isthestableexcitationforNDGSPswhenthesystemisslightlydoped.Apolaroncanbethoughtofasaboundstateofachargedsolitonwithaneutralsoliton,andthusitpossesseschargeandhasspin1/2.Ithasbeenshown[9,12]thatbyincreasingthedopingleveloftheconjugatedpolymersystem,theformationoftwopolaronsisenergeticallymorefavorablethanthecreationofabipolaron.ThisisbecausethelatticedistortionrequiredforthecreationofabipolaroncorrespondstogreatershiftsinthelocalizedHOMOandLUMO(valencebandandconductionband).Therefore,thesystembecomessaturatedwithpolaronsrst,beforethebipolaronsbegintoform.ThisissupportedbyElectronSpinResonance(ESR)experiments[12],whereatlowdoping

PAGE 25

Athigherdopinglevels,theformationofachargedbipolaron,orradicaldicationordianion,isthestableexcitationforNDGSPs.Abipolaronisaboundstateoftwochargedsolitonsofthesamecharge,ortwopolaronsofthesamechargewhoseneutralsolitonsannihilateeachother.Thus,bipolaronspossesschargebutcarrynospin.Thecreationofabipolarongeneratestwosymmetricalmid-gapenergystates,asinpolaronsbutlocatedclosertothecenterofthegap.Whenthesemid-gapstatesareemptythebipolaronispositive(p-typedoping),whereaswhentheyarefullyoccupiedthebipolaronisnegativelycharged(n-typedoping).Theseexcitationsaredelocalizedoverseveralmonomerunits,generallysixtoeight,andthuscanpropagatealongthepolymerchain. Asthedopinglevelisfurtherincreasing,theindividualbipolaronlevelsdescribedabove,showninFigure2.13,coalesceintobands.Thesebipolaronbandsarisefromthedepletionofelectronicstatesfromthevalenceandconductionbandedges,whichresultsinaconcomitantincreaseofthegap.Atthesehighdopinglevels,bipolaronscangiverisetohighconductivityupontheapplicationofanelectriceld.ESRexperimentshaveproventhatinhighlyconductiveconjugatedpolymers(CPs)thechargecarriersarespinless,andthereforebipolarons,sincenosignalcanbedetected[12]. Mostcasesconsideredinliteraturerefertopositivelychargedpolaronsandbipo-larons,i.e.,p-typedopingoroxidationoftheconjugatedpolymers,becausemostCPscannotformstablenegativelypolaronsorbipolarons. TheconceptofdopinginCPsisuniqueformanyreasons.First,thedopingprocessisreversible,andthede-dopingprocessproducestheneutralpolymerwithlittleornodegradation.Second,itinvolvestheformationofmobileexcitations,suchassolitons,polarons,and/orbipolarons,thatcoupletolatticevibrations.Third,unlikedopinginconventionalsemiconductors,inCPsthedopantatomsdonotsubstituteatomswithinthelatticebutarepositionedbetweenthepolymerchains,anddonateoracceptelectronsfromthepolymerbackbone[13].Finally,thedopinglevelcanbecontrollablyadjusted,andthustheconductivitycanbetunedoverseveralordersofmagnitudeinthesame

PAGE 26

Thehighlyanisotropiccharactershownbytheseonedimensionalmaterialscausestheelectronicmotiontobeeasyonlyalongthebackboneofthepolymerchain.There-fore,thepropertiesofthesematerialsaremainlygovernedbythefollowingtypesofinteractionsamongtheunpairedelectronsthatoccupythehighestmolecularorbitalsinthesolid:(1)theoverlapofthewavefunctionsoftheseelectronsbetweenadjacentsitesinthepolymericchain;(2)theinteractionsoftheelectronswiththeirsurroundings,inparticularlywiththelatticevibrations(electron-phononcoupling);and(3)theCoulombinteractionbetweentheelectrons(electron-electroncoupling).Varioustheoreticalmod-els,basedonthesimpletightbindingmodel,havebeendeveloped,takingintoaccountsomeoftheaboveinteractionsinaneorttoprovideamorerealisticdescriptionofthesesystem. Averywellknownmodel,usedforthetheoreticaltreatmentofconjugatedpolymersystems,istheSu-Schrieer-Heeger(SSH)model.Thismethodwasinitiallyappliedtopoly(acetylene),andyieldedusefulinformationfortheinterpretationofexperimentaldata.ThisSSHtreatmentofpolymericsystemsincludestheelectron-latticevibrationinteractions,omitstheelectron-electroninteractions,andmodelstheCPchain,orlat-

PAGE 27

2XnK(un+1un)2+1 2XnM_u2n(2.1) whereithasbeenassumedthattheelectronscanbetreatedinthetightbindingapproximationwithahoppingintegraltn+1;n,whichcanbeexpandedtorstorderaroundtheundimerizedstate(beforethePeierl'sdistortion,allbondsareofequallength): wheret0isthehoppingintegralfortheundimerizedstate,andistheelectron-latticedisplacement(phonon)couplingconstant.Thislinearapproximationisvalid,sincethebond-lengthchangesaresmall,oftheorderof0.08A.Inequation(2.1),cyn;sandcn;screateanddestroyelectronsofspin1/2onthenthrepeatedunit.Thesecreationandannihilationoperatorssatisfytheanticommutationrelationsoffermions.Thesecondterminthisequationcorrespondstothebondingenergy,whichhasbeenexpandedtosecondorderaroundtheundimerizedstate.Therstordertermcanbeneglectedduetosymmetry.Finally,thethirdterminequation(2.1)isthekineticenergyofnuclearmotion,whereMisthetotalmassoftherepeatedunit.ThistermcanalsobewrittenasM_un=pn,wherepnandunsatisfycanonicalcommutationrelations: [pn;un0]=~ Anothermodelusedforthetheoreticaltreatmentofconjugatedpolymersystems,isthePariser-Parr-Pople(PPP)model.Thismodelconsiderstheelectron-electroncou-pling,,orCoulombinteraction,whileneglectingtheelectron-phononcoupling.ThePPPHamiltonianiswrittenasfollows[15]:

PAGE 28

2Xi;i0Vi;i0(ni1)(ni01)(2.4) where isthenumberoperator.U,Vi;i0areawaytoparameterizetheelectron-electroninterac-tion.ThepotentialUiseectiveonlywhentwoelectronsoccupythesamesite,whereasVi;i0iseectiveforelectronsthatoccupydierentsites. Inadditiontotheonesgenerallydescribedabove,thereareotherwellknownmodelsusedforthetheoreticaltreatmentofconjugatedpolymersystems.TheseincludetheHubbardmodel,theextendedHuckelmodel,theValenceEectiveHamiltonian(VEH)method,thePeierl-Frohlichmethod,andmore.2.4ConductivityinConjugatedPolymers(CPs) Experimentalstudieshaveshown[5]thattheelectricalconductivityofCPsim-provesasthedegreeofchainextensionandchainalignmentisincreased.Thestudies

PAGE 29

ThemechanismofchargetransportinCPshasbeenthesubjectofresearchoveralongperiodoftime.Animportantfeatureofthesesystemsintheirdopedstateisthatthevariationofconductivitywithtemperatureisdierentcomparedtoconventionalmetals.Ingeneral,asthetemperaturelowerstheconductivitydecreases,althoughforsomeofthemosthighlyconductingpolymerstheconductivityremainshighevenatlowtemperatures[13,16].ThisdierenceinbehaviorbetweenCPsandmetalsseemstobetheresultofdisordereects[13,17].WhenthedopingofCPsdoesnotoccurinahomogeneousway,thelmsconsistofhighlyconductiveislands,orordereddomains,surroundedbylessconductive,oreveninsulating,material.Theselatter,lessorderedoramorphousregions,serveaspotentialbarriersforthechargetransport. Avarietyofdierenttransportprocessescontributetotheconductionmechanism.Theseincludeintrachainandinterchaintransportwithintheordereddomains,aswellashoppingortunnelingacrossdisorderedregions.Hence,severalconductionmodelshavebeendevelopedovertheyearsinaneorttounderstand,andaccountforthedierent

PAGE 30

Finally,giventhatconjugatedpolymersshowtunableconductivitiesintheirdopedstateswithvalues,insomecases,thatcanreachthoseofconventionalmetals,onecannothelpbutwonderifthereisapossibilitythatthesesystemswillexhibitsuperconductiv-ity.Althoughconjugatedpolymerssharemanyfeatureswithorganicmaterials,suchasthetetramethyltetraselenafulvalencefamily,(TMTSF)2X[18],andmethylenedithiote-traselenafulvalenefamily(MDT-TSF)X1:271:29[19]whereXisusuallyahalogen,whichexhibitsuperconductivity,thisphenomenonhasnotbeenobservedyetfordopedconju-gatedpolymericsystems.Workersintheeldappeartobeoptimistictowardsthisideabutthereisstillalotofprogressthatneedstobedone.Currentlyavailablematerialsarebarelymetallicwithelectronicpropertieswhicharedominatedbydisorder,render-ingthecharacteristicmeanfreepathsintheregionfordisorder-inducedlocalization.Therefore,therststepinthedirectionoftrulymetallicconjugatedpolymersistheimprovementofthematerialsquality,whichwilleventuallyresultinlongermeanfreepaths,ideallyoftheorderofthemonomericunit.2.5Metal-Insulator(M-I)Transition

PAGE 31

wherekFistheFermiwavenumber,andlisthemeanfreepath.ThemetallicregimeexistsforkFl1. Basedontheabovecriterion,Mottstatedthatforametal-insulatortransitiontooccur,thedisordershouldbesucientlylargethatkFl>1[20{22].InthelimitwherekFl1,theaveragedisorderpotentialbecomeslargecomparedtothebandwidthandthus,allstatesbecomelocalized.Inthiscase,thesystembecomesaninsulatorandiscalledaFermiglass.EventhoughFermiglasseshaveacontinuousdensityofstatesandnoenergygap,theybehaveasinsulatorsasaresultofthespatiallylocalizedstatesattheFermilevel. Inconjugatedpolymericsystems,thismetal-insulatortransitionisveryinterest-ing.This,isduetotheabilitytocontrolthecriticalregimebyvaryingthedegreeofdisorderofthesystem,orbyapplyingexternalpressureand/ormagneticelds.Thecriticalbehavior,closetothephasetransition,hasbeenobservedbyseveralworkersinanumberofconjugatedpolymers,suchaspoly(acetylene),poly(pyrrole),poly(para-phenylenevinylene),poly(aniline)[23],aswellasinpoly(3,4-ethylenedioxythiophene)[16],inarelativelywiderangeoftemperatures.Althoughmetallicbehaviorhasbeendemonstratedforconjugatedpolymers,thetrulymetallicregimeforwhichkFl1hasnotbeenachievedyet.2.6DopingInducedPropertiesinConjugatedPolymers

PAGE 32

ForthecaseofNDGSPs,therearealsosignicantsignaturesofchargedpolaronandbipolaronformation.First,thegenerationoflocalizedstructuraldistortionswhichisassociatedwithphononmodes.ThesepolaronandbipolaroninducedIRAVmodesareactiveinthemid-infraredregion,andcanbeobservedbyspectroscopicmeasurements,asitwillbeenseenlaterinthissection.Second,theelectronictransitionsassociatedwiththegenerationofthetwo,symmetricallyplaced,mid-gapenergystatescanalsobeobservedbyspectroscopicmeasurementsinthenearinfraredfrequencyrange.Fi-nally,thechargestorageinitiallyinchargedpolaronswithspin1/2and,asthedopinglevelincreases,incharged,butspinless,bipolaronscanbeveriedthroughelectron-spinresonanceexperiments.Theseexperimentsshowasmallsignalatlowdopinglevelsthatgrowsasthedopingincreasesandsaturatesatintermediatedopinglevels,consis-tentwiththeformationofchargedpolaronswithspin1/2.Athigherdopinglevelstheformationofbipolaronscommence,andthereforethesignalisgraduallylost[12]. Figure2.15showsaschematicbanddiagram,andtheabsorptionspectraofpoly(3,4-ethylenedioxythiophene),aNDGSP,intheneutral,slightlyp-doped,andheavilyp-dopedstates.Intheneutralstate,onlythe-,orEg,transitionispossibleandtherefore,onlyoneabsorptionbandappearsinthespectrum.Whenthepolymerisinitsslightlyp-dopedstate,the-transitiondiminishesand,duetotheformationofpositivelychargedpolarons,twosymmetricmid-gapstatesaregenerated.Thisresultsintwoadditionalabsorptionbandsinthespectrum.Finally,uponheavyp-doping,thecreatedpositivelychargedbipolarons,movethetwosymmetricmid-gapstatesclosertothecenterofthebandgapandtheHOMOandLUMO(thevalenceandconductionbands)furtherapart.Therefore,the-transitionistotalybleachedand,basedonthe

PAGE 34

Thismethodwastheoneusedintheinitialdiscoveryoftheabilitytodopeconju-gatedpolymersbycharge-transferredoxchemistry[1].Theoxidation,orp-doping,ofpoly(aniline),PANI,wasachievedbyexposingthepolymertoiodinevapors,andthereduction,orn-doping,involvedtreatmentwithsodiumnaphthalenide.Che-micaldopingcanalsobeachievedwithprotonationbyacid-basechemistry.Thistypeofdopingleadstoaninternalredoxreaction,andforthecaseofPANItotheconversionofemeraldinebase,asemiconductor,toemeraldinesalt,ametal[5]. Thechargetransferredoxchemistry,oxidationandreduction,isillustratedinthefollowingexamples:(CP)n+3 2nxI2![(CP)+x(I3)x]n(2.7) foroxidation,orp-typedoping,and(CP)n+[Na+(Naphthalide)]![(Na+)x(CP)x]n+(Naphthalide0)(2.8) forreduction,orn-typedoping. Materialsproducedbychemicaldopinghaveveryhighelectricalconductivities,andcanbeusedinapplicationsastransparentelectrodes,antistatics,electromag-neticinterference(EMI)shielding,andintrinsicconductingbers.

PAGE 35

Electrochemicaldopingisillustratedinthefollowingexamples:(CP)n+[Li+BF4]solution![(CP)+x(BF4)x]n+Lielectrode(2.9) foroxidation,orp-typedoping,and (CP)n+Lielectrode![(Li+)x(CP)x]n+[Li+(BF4)]solution(2.10) forreduction,orn-typedoping. Materialsproducedbyelectrochemicaldopingcanbeusedinelectrochemicalbat-teriesforchargestorage,lightemittingelectrochemicalcells,andelectrochromicapplications,e.g.,\smartwindows",opticalswitches,andlowenergydisplays.

PAGE 36

Inbothmethodsdescribedabove,chemicalandelectrochemicaldoping,thein-ducedelectricalstructureispermanentuntilthesystemispurposely\un-doped",i.e.,thechargeisremovedorchemicallycompensated. wherexisthenumberofelectron-holepairs.Thisnumberdependsuponthecompetitionofthepumpratewiththerecombinationrate.The\photoconduc-tivity"lastsonlyuntiltheexcitationsareeithertrapped,orhavedecayedbacktothegroundstate.Followingthephotoexcitationfromthegroundstatetothelowestexcitedstatewiththepropersymmetry,therecombinationordecayofanelectron-holepairtothegroundstatecanbeeitherradiative(luminescence)ornon-radiative.Someconjugatedpolymers,e.g.,PPV,andPPP,showluminescencewithhighquantumeciencieswhileothers,e.g.,PAcandPth,donot[9]. Photodopingbyphotoexcitationproduceshigh-performanceopticalmaterials,whicharesuitableforphotovoltaicdevices,andalsoprovidesarouteformaterialswithtunablenonlinearoptical(NLO)responseforelectro-opticandopticaldevices,e.g.,waveguides.

PAGE 37

and([(CP)n+xe]![(CP)x]n:(2.13) Thismethodisfundamentallydierentfromchemicalandelectrochemicaldoping,becausetherearenocounterionsintroducedinthesystem,althoughthepoly-merbecomesoxidizedorreduced.Inthecaseofchargeinjectionatametal-semiconductorinterface,electronsresideintheband,and/orholesatthebandonlyaslongasbiasingvoltageisapplied.Thentheinjectedelectronsandholesrecombinewiththeemissionofradiation(electroluminescence). Thiselectronorholeinjectionatametal-semiconductingpolymerinterfaceisparticularusefulforapplicationssuchasorganiceldeecttransistors(FET's),andlightemittingdiodes(LED's)[27,28].2.8FundamentalsofElectrochromism Electrochromicmaterialschangecolorinareversiblewaybyanelectrochemicalreaction,andtheycanbeclassiedintothreecategoriesbasedontheirelectronicallyaccessibleopticalstates,orsimplystated,basedontheircapabilitytoaccessdierentcolors.Therstcategoryincludesmaterialsthathaveonecoloredstate,i.e.,inthisstatethematerialabsorbswithinthevisibleregion,andonetransparentorbleachedstate,i.e.,inthisstatetheabsorptionisoutsidethevisibleregion.Thisclassofmaterialsismostlyusedintransmissive/absorptivetypeofdevices.Thesecondcategoryincludesmaterialsthathavetwodistinctivecoloredstates,i.e.,bothstatesabsorbwithinthe

PAGE 38

Manydierentmaterials,inorganicandorganic,exhibitelectrochromism.Amongthemaretheinorganictransitionmetaloxidesystems,andespeciallythehighbandgapsemiconductortungstenoxide,WO3,which,hasbeenthemainfocusofresearchforthelastthreedecades,[30{34].Tungstenoxide,uponreduction,changesfrombeingtransparentinthevisibleregiontohavingabluecolor: WO3+xM++xe!MxWO3:(2.14) Thefabricationofthinlmsof,amorphousorpolycrystalline,WO3requiressputteringunderhighvacuum,whichisacomplicatedandexpensiveprocess.Thishighmanu-facturingcost,inadditiontootherreasonssuchastheverylonglifetimerequirements,aswellastheinsucientlyfastresponsetimes,proclaimsthatthereisstillworkthatneedstobedone. Otherinorganicmaterialsthatexhibitelectrochromismaremixedoxidesofvana-dium(V),molybdenum(Mo),niobium(Nb),titanium(Ti),nickel(Ni),cobalt(Co),andiridium(Ir),phthalocyaninemetalcomplexes,andalsotransitionmetalhexacyanome-tallates,suchasPrussianBlue(PB).PrussianBlue,anexampleofpoly-electrochromicmaterial,hasabluecolorinits\natural"form,anduponreductionitbecomestrans-parent,theso-calledPrussianWhite(PW)butalsoknownasEveritt'ssalt: [FeIIIFeII(CN)6]+e![FeIIFeII(CN)6]2;(2.15) whilepartialoxidationofPrussianBlueresultsinPrussianGreen(PG),which,asindi-catedbythename,hasagreencolor:

PAGE 39

3[FeIIIFeII(CN)6]![FeIII(FeIII(CN)6)2FeII(CN)6]+2e;(2.16) andfurtheroxidationyieldsPrussianBrown,whichhasayellow-goldencolor: [FeIIIFeII(CN)6]![FeIIIFeII(CN)6]+e:(2.17) Anotherfamilyofmaterialsthatexhibitstheelectrochromiceectincludessmallorganicmolecules,suchasbipyridiliumsalts,otherwiseknownasviologens.Themostwellknownviologensarethe1,1'-dimethyl-4,4'-bipyridiliumorotherwisecalledmethylviologen(MV)[32],andthe1,1'-di-n-heptyl-4,4'-bipyridilium,alsoknownasheptylviologen[35,36].Thistypeofmaterialsuponreductionundergoesachangefromatransparenttoacoloredstate,withthecoloredstatedependingonsuitablechoicesofnitrogen,oralkyl,orothergroupsubstitutions.Untilrecently,theonlywidespreadelec-trochromiccommercialapplicationistheautomaticrear-viewdimmingmirrorsystembyGentex,calledNightVisionSafety(NVS),whichutilizessolution-phaseelectrochromicviologens.Despitethesuccessofthisreectivedevice,thedevelopmentofotherelec-trochromicsystems,e.g.,\smartwindows"forbuildings,havenotshowntheexpectedbreakthroughinthemarketyet. Conjugatedpolymersarethethirdfamilyofelectrochromicmaterials,andtheonethathasgainedalotofattentionintherecentyears.Althoughnotasdevelopedasthematerialsdiscussedearlierinthissection,theirpopularityisbasicallyduetothefactthatconjugatedpolymersareeasiertoprocessthaninorganicelectrochromicmaterials,andtheyoerthemajoradvantageofcolortunability.Morespecically,theircolorcanbetailoredthroughstructuralmodicationoftherepeatunit,i.e.,monomerfunctionalizationandcopolymerization,orthroughtheuseofblends,laminates,andcomposites.Hence,onecanndconjugatedpolymersinanyofthethreecategoriesmentionedabovefortheuseintransmissive/absorptiveorreectivetypeofdevices,andinawideselectionofcolors.Conjugatedpolymersalsopromiserapidresponsetimes,

PAGE 40

Thefactthatelectrochromicmaterialshavebeenintensivelystudiedthelastdecades,andlatelyhavebeenemployedinnumerousapplications,forcedthedenitionofelectrochromismtobemodiedinordertotwithinthedemandsofthemodernworld.Therefore,althoughpreviouslyelectrochromismwasthereversibleandvisiblechangeofthecolorassociatedwiththereduction-oxidation,ordoping/de-doping,processofanelectrochromicmaterial,thisdenitionhasbeenextendedtoincludeawiderspectralrangemodulation.Thisspectralrangenowcoversultraviolet(UV),visible(Vis),nearinfrared(NIR),midinfrared(MIR),farinfrared(FIR),andmicrowave(MW)regions.Inthecaseoftheseregions,\color"correspondstotheresponseofthedetectorsatthedierentwavelengths,andcanbestudiedasthechangeinthetransmittance,and/orreectance,inducedbydopingorde-dopingoftheelectrochromicmaterial.2.9SynthesisMethodsofCPs Intheelectrochemicalcase,thepolymerizationisalsoinitiatedbythegenerationofaradicalion,whichoccursatanelectrodesurfacebyoxidationviaanappliedelectricpotential.Duetothefactthatintheelectrodevicinity,wherethereactionstakeplace,theconcentrationofradicalionsislarge,radical-radicalcouplingoccursandthis,leadsto

PAGE 41

ElectrochemicalpolymerizationisthemethodthathasbeenusedthroughoutthisworkbecauseitprovidesaquickandeasywayforthedepositionofCPlmsonvari-oussubstrates.Thisprocessemploysathree-electrodecongurationthatconsistsofaworkingelectrode(WE),anauxiliary(counter)electrode(CE),andareferenceelectrode(RE).Theworkingelectrodeistheelectrodewherethepolymerizationofinteresttakesplace,andthereisanumberofdierentmaterials,solidorexiblesubstrates,thatcanbeused.Thechoicesincludeindium-dopedtinoxide(ITO)onglassorPET,PEDOT-PSSonPET,andthinmetallmsongrids,whentransmittancemeasurementsneedtobeperformed,whereas,solidplatinum(Pt)orgold(Au),andgoldonMylar,areusedwhenreectancemeasurementsneedtobeperformed.Thecounterelectrodeprovidestherequiredcurrenttosustainthedevelopingprocessesattheworkingelectrode.Inordertoensurethattheelectrochemicalreactionstakingplaceonitssurfacearenotlimiting,theareaofthecounterelectrodeneedstobelarger,oratleastsimilar,totheareaoftheworkingelectrode.APtag,whichisapieceofPtfoilandPtwireweldedtogether,isusuallyemployedasacounterelectrode.Finally,thereferenceelectrodeprovidescontroloftheappliedelectricalpotential.Thereareseveraldierenttypesofreferenceelectrodesthatarecommerciallyavailable,suchassaturatedcalomel,Hg2Cl2(SCE),Ag/AgClsaturatedinKCl,andAg/Ag+electrodes.Inadditiontothesestan-dardreferences,quasi-orpseudo-referencescanbeused,e.g.,silverwire.Thelatteronesneedtobecalibratedeverytimebeforetheiruse.Forthepolymerization,theWE,CE,andREelectrodesareplacedinamonomersolution,andtheyareconnectedtoapotentiostat/galvanostat.Thisthree-electrodearrangementpreventslargecurrentsfrompassingthroughthereferenceelectrode,andchangeitspotential. TheelectrochemicaldepositionoftheCPonaconductivesurfaceofchoicecanbeachievedinseveraldierentways.Inthecasewheretheappliedelectricpotential

PAGE 42

Cyclicvoltammetry(CV)isanothermethodthatcanbeusedforCPdeposition.Inthismethod,thepotentialisrepeatedlycycledoveraspeciedvoltagerange,whiletheresultingcurrentismeasured.Theobtainedvoltammogramisadisplayofcurrentdensityasafunctionoftheappliedvoltage.Thescanrate,expressedinmV/s,isadynamicparameterandcanbechangedinorderfordierenttypesofreactions,fastorslow,tobefollowedeectively.Thismethodresultsinlmsofcomparablequalitytothepotentiostaticmethod.2.10CharacterizationMethodsofElectrochromicCPs Spectroelectrochemistrycanbeperformedusingthethreeelectrodeconguration,describedintheprevioussection,toinitiallycharacterizetheconjugatedpolymerlminmonomer-freesolution.Inthismethod,dierentpotentialvaluesareapplied,andafterthesystemreachesanequilibrium,theabsorptionspectrumismeasured.Figure2.16showsthespectraofaPEDOTlmdepositedonITO/Glasssubstrate,atdier-entdopinglevels,fromwhichtheabsorptionofthemonomer-freesolution,andoftheITO/Glasssubstratehavebeensubtracted.

PAGE 43

Ascanbeclearlyseenfromthespectrum,whenthepolymerisinitsneutral(un-doped)stateitbehavesasasemiconductorwithanenergybandgapof1.65eV.Thebandgapisdeterminedbyextrapolatingtheonsetofthestrong-absorptionbandinthevisibleregion.Uponelectrochemicaldoping(oxidationorp-typedoping),thebandstructureofthepolymerismodied;lowerenergymid-gapstatesaregeneratedandchargecarriers,polaronsandbipolaronsarecreated.Therefore,theabsorptionspectraevolvesaccordingly.Uponlowdoping,thestrengthofthe-absorptionbandisreduced,whileabsorptionbandsduetopolaronsemerge.Asthedopinglevelcontinuetoincrease,the-andpolaronbandsareslowlydepleted,andabroadbipolaronabsorptionbandatlowerenergiesforms.Atthehighestlevelofdoping,only

PAGE 44

Spectroelectrochemistrymeasurementscanalsobeperformedinreectiveortrans-missive,electrochromicdevices.Fordevices,thethreeelectrodecongurationisreplacedbyasuitableholder.Inordertoapplyavoltageacrossthedevices,thecounterandthereferenceleadshavetobeconnectedtooneanother.Dierentpotentialvaluesareappliedand,afterthesystemreachesequilibrium,theabsorptionspectrumismea-sured.Fromtheresultingspectra,usefulinformationcanbededucedconcerningtheelectrochromicmechanismswithinthedevice. Anotherimportantcharacterizationmethodconsistsofakineticsexperiment,whichallowsformeasuringoftheswitchingtimesbetweentwoextremeelectrochromicstates,neutralanddopedforthepolymerlmsinsolutionorindevices.Inthisex-periment,asquare-wavepotentialisappliedatspeciedtimeintervals,whileconcur-rentlytheabsorbanceatmaxismonitored,wheremaxisthewavelengthofmaximumelectrochromiccontrast,andcanbedeterminedfromthespectroelectrochemistrymea-surementsdescribedabove.Furthermore,thesameexperimentalsetupcanbeusedtoperformalong-timeredoxswitchingstability,orlifetime,test.Byapplyingthesquare-wavepotentialoveralongperiodoftime,hundredsoreventhousandsofcycles,onecanstudythedegradationoftheperformanceofthepolymerlmitself,orinthedeviceunderinvestigation.Theswitchingtimesandthelifetimescanvarysignicantlynotonlybetweendierentpolymers,butalsoforthesamepolymer,e.g.,whendierentdopantsareused. Othercharacterizationmethodsincludecyclicvoltammetry;atechniquethatpro-videsinformationaboutthepotentialvaluesatwhichoxidationandreductionoccurforeachpolymer,in-situconductivitymeasurements,andin-situcolorimetricanalysis.Thislattermethodaccuratelydenesthecolor,takenintoaccountthesensitivityofthehumaneye,andtheelectrochromiccontrastratiosinconjugatedpolymers.Although

PAGE 45

TherearetwobasiccategoriesofECDsbasedontheirfunctionmodeor,inotherwords,thetypeoflightmodulationtheyperform.Therstofthesecategories,isthereectiveECDsinwhichthelightincidenttothedeviceiscontrollablyreected.ThetypicaldesignofanECDthatoperatesinreectancemodeisanoutward-facingplatform,andwasoriginallydevelopedbyBennett[39]andChandrasekhar[40].Morespecically,thedeviceconsistsofanoutward-facingworkingelectrode,usuallygoldcoatedonaslittedMylaroronaporousmembrane,ontowhichanactiveelectrochromicCPisdeposited.Theporousmembranecanbepolycarbonate,polysulfone,polyester,ornylon,andcomparedtotheslittedMylaritallowsforfasterandmoreuniformiondiusion[41,42].Thecounterelectrode,usuallygoldonMylar,doesnotcontributedirectlytothemodulationoflightbutthecounterelectrodeECpolymerservestobalancethecharge.Toassemblethedevice,thecounterelectrodeisplacedfacingthe

PAGE 46

MoredetailsaboutthedesignofthistypeofdevicesaregiveninChapter5.An-othertypeoffabricationofreectiveECDsincludespatterningoftheworkingelectrode,anduseofmorethanoneelectrochromicCPsinthesamedevice.Thisdesignprovideshighresolutionpixeldevicesinwhicheachpixelcanbeaddressedindividually[38,43]. ThemostcommonapplicationsofreectiveECDsaredisplaysalthough,thereisstillalotofworktobedoneforthesedevicestobewidelycommercialavailable.Manyworkersareputtingaseriouseortintomakingdeviceswithfasterswitchingtimes,longerstability,andhigherelectrochromiccontrast.ReectiveECDsbasedonCPsoermanyadvantagescomparedtoLCDs(LiquidCrystalDevices).Theyareeasytoprocess,havelowproductioncost,lowvoltagerequirements,oerawideselectionofcolors,andthequalityoftheimagedoesnotdependontheviewingangle.Furthermore,reectiveECDscanbemadeexibleand,inanysizeandshape.Otherapplications

PAGE 47

ThesecondcategoryofECDsisthetransmissive/absorptivedevice,inwhichtheincidentlightcontrollablypassesthroughthedevice.ThedesignofthistypeofECDsconsistsoftwoelectrochromicCPsdepositedontransparentelectrodeswhicharefacingeachotherandareseparatedbyanelectrolytelayer,asshowninFigure2.18.Theelec-trochromicCPsemployedintransmissive/absorptiveECDs,mustbecomplimentarytoeachother,acathodicallycoloringpolymerandananodicallycoloringpolymer,inorderforhighcontrastvaluestobeachieved.Acathodicallycoloringpolymerhasalowbandgap,ideallyaround1.8-2.2eV,iscoloredinitsneutral(undoped)state,anduponoxida-tionitbecomestransparentinthevisibleregion.Whereas,ananodicallycoloringpoly-merhasahighbandgap,ideallyhigherthan3.0eV,istransmissiveinthevisibleregionwhenitisinitsneutralstate,anduponoxidationitabsorbsthevisiblelight.Therefore,theECDcanbereversiblyswitchedbetweenacolored,absorptive,andatransmis-sive,bleached,state.Theconductingelectrodesemployedinthesedevices,dependontheregionoftheelectromagneticspectrumthedeviceisneededtooperate.Typicaltransparentelectrodesforthevisibleregion,consistofindium-dopedtinoxide(ITO)lmsdepositedonglass,andITOorpoly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate)(PEDOT-PSS),depositedonpoly(ethyleneteraphthlalate)(PET)forexibleECDs.Thistypeofdevicescanbeusedas\smartwindows"forcars,e.g.,rear-viewmirrors,sunroofs,andforbuildings,providinghugeenergysavingcost.2.12GeneralApplicationsofCPs

PAGE 48

Inparticular,conductingpolymersexhibitinterestingandimportantpropertiesnottypicallyavailableinothermaterials.Theiradvantagesoverothermaterials,or-ganicandinorganic,layontheirlightweight,relativelyinexpensivefabrication,easyprocessibilityandabilitytobeformedintooddshapesandsizes,lowvoltagerequire-ments,compatibilitywithmanyorganicliquidandsolidelectrolytes,andfastswitchingcolorchanges.Inaddition,awidevarietyofcolorsisavailablewithconjugatedpoly-mers,aswellastheabilityofcolortunabilitybasedforexampleonthedopinglevel,orthechoiceofdopant.Theseuniquepropertiesmakepossibleanumberofappli-cationssuchastransparentelectrodes,antistatics,electromagneticinterference(EMI)shielding,conductingbers,electrochemicalbatteries,anti-corrosioncoatings,sensors,\smartwindows",opticalswitches,thermalcontroldevices,lowenergydisplays,lightemittingelectrochemicalcells,photovoltaicdevices,organiceldemissiontransistors,andmore[5,17,26,27,32,33,36,38,44].

PAGE 49

Opticalexperimentsonthinlmsprovideavaluablemethodofunderstandingthepropertiesofmaterialsthatareinasolidstateform.Thematerialsspecicparameters,directlyavailablethroughtheseexperimentsarethefrequency-dependentreectanceR,ortransmittanceT,spectrumoveraregionofinterest.Thisregiondoesnotreferonlytotherelativelynarrowvisibleregionbutextendsandcoversthewholerangeoftheelectromagneticspectrumfromfarinfraredtoultra-violet.Fromtheexperimentalmeasurements,Rand/orT,andcalculationsbasedontheoreticalmodels,thedielectricfunctioncanbededucedwhichisthepropertymostdirectlyrelatedtotheelectronicstructureofthematerialunderconsideration.Thisway,theopticalphenomenacanbequantied,andanumberofopticalconstantscanbederiveddescribingtheresponseofthemediumtolight.3.1OpticalProcessesandOpticalConstants

PAGE 50

whereiscalledtheabsorptioncoecientandisastrongfunctionoffrequency,andz0istheinterfaceplane. Anotherphenomenonthatcausesattenuationofthelightintensitywithinamediumisscattering.Scatteringisresponsibleforre-directingthelighttoallpossibledirections.Consequently,itdiminishestheamountoflightthatproceedsintheforwarddirectionandtherefore,theamountoflightwhichtransmitsthroughthematerial.Thus,scat-teringhasaneectsimilartoabsorption,andtheresultedattenuationoflightcanbeexpressedinananalogousformastheequationabove: whereNisthenumberofscatteringcentersperunitvolume,andisthescatteringcross-sectionofthescatteringcenter.Anumberofdierentfactorscanberesponsibleforthescatteringprocess,suchasinhomogeneities,impurities,defects,etc. Furthermore,thelightwavestravelwithinamediumwithasmallervelocitycom-paredtothevelocityoflightinfreespace,andwithavelocitythatdiersfordierentmaterials.Thisphenomenon,calledrefraction,isdescribedbytherefractiveindexn.Therefractiveindexis,ingeneral,afunctionoffrequencyandthiseectisknownasdispersion.Therefractiveindexnisdenedastheratioofthevelocityoflightinfreespacec(wherec=2.998x108ms1),tothevelocityoflightinamedium: :(3.3) Thevelocitychangecausesthelightraystobentwithrespecttothenormalontheinterfaceasthelightgoesthroughtwodierentmaterials,anditisdescribedbySnell'slaw[45]. Thetwoquantitiesthatdescribethepropagationofthelightwavewithinamedium,theabsorptioncoecientandtherefractionindexn,canbecombinedintoasingleparametercalledthecomplexrefractiveindex:

PAGE 51

~n=n+i;(3.4) wheretherealpartofthecomplexrefractiveindexistherefractiveindexn,asdenedinequation(3.3),andtheimaginarypart,isrelatedtotheabsorptioncoecient,asitwillbeshownlater.Theimaginarypartiscalledtheextinctioncoecient. Intheabovedescriptionofthephenomenathatoccurwhenlighttravelsthroughamedium,ithasbeenassumedthattheresultofthisinteractionisindependentoftheintensityofthelightbeam.Thereforetheanalysisisbasedonlinearopticswherepropertiessuchastherefractiveindexnandtheabsorptioncoecientaretakentobeindependentoftheopticalpowerofthesource.Thisisindeedthecasewhenconventionallightsourcesareused,andourresearchthroughoutthisdissertationfallsintothisdomain.However,nowthathighpowerlasersareavailable,thereareanumberofotherphenomenathatcanoccurasalightbeamofhighintensitypropagatesthroughamedium.Thesearedescribedbynonlinearoptics.Nonlineareectscausethepropertiesofthematerialunderinvestigationtodependontheintensityofthelightsource.Anexampleofanonlineareectisthedoublingofthefrequency,wherethefrequencyofpartofthebeamdoublesasitinteractswithamedium.3.2InteractionofElectromagneticWaveswithMatter Theapplicationofanexternalelectriceldtoanisotropicandhomogeneousmediumtendstoalignthemicroscopicdipolemomentswithintheatomsalongtheexternalelddirection.Thisproducesanetdipolemomentwithinthemedium,andthereforeapolarization.Morespecically,thepolarizationPisdenedasthenetdi-

PAGE 52

whereeeistheelectricsusceptibilityofthemedium. TheelectricdisplacementDofthemediumisdenedas andbycombiningthelasttwoequationswecanwrite where ~1+i2=1+4ee(3.8) isthecomplexdielectricfunction,whichisaveryimportantparameterfortheunder-standingoftheinteractionoflightwithmatter. ThecurrentdensityJ,whichcomesasaresponseofthemediumtotheapplicationofE,isrelatedtotheelectriceldasfollows where ~1+i2:(3.10) isthecomplexconductivityofthemedium.Ingeneral,thecurrentdensityistheresultofthecontributionsarisingfromtheboundandfreecharges:J=Jbound+Jfree,where

PAGE 53

~=1+i4~ !:(3.11) Thelastequationcanalsobeexpressedasfollows and 4:(3.13) Anotherusefulexpressionthatrelatesthecomplexdielectricfunctionandtherealpartoftheconductivityisthefollowing: ~=1(!)+4i !1(!)(3.14) where1(!)istherealdielectricfunction,while1(!)istherealpartofthecomplexconductivity,alsocalledopticalconductivity.Aninterestingcaseisthezerofrequencylimitoftheaboveequation,whichspeciestheresponseofthemediumtostaticelds.Inthislimit,1(!=0)becomesthestaticdielectricconstant,while1(!=0)becomesthedcelectricalconductivity. Inananalogousway,astheapplicationofanexternalelectriceld,theapplica-tionofanexternalmagneticeldtoanisotropicandhomogeneousmediumproducesamagnetizationMtothemedium,whichisproportionaltoappliedmagneticeldH: whereemisthemagneticsusceptibilityofthemedium.

PAGE 54

ThemagneticuxdensityBofthemediumisdenedas: andbycombiningthelasttwoequationswecanwrite where ~=1+4em(3.18) isthecomplexmagneticpermeability. Theinteractionofamediumwithanappliedelectromagneticeld,isdescribedbyMaxwell'sequation's: cJf+1 wherefisthefreechargedensity,andJfisthefreecurrentdensity. Intheabsenceofexternalcharges,f=0,andcurrents,Jf=0,Maxwell'sequationscanbewrittenasfollows:

PAGE 55

EliminatingD,andHfromtheaboveequationsresultsin: c@E Usingthefollowingvectoridentity: weobtainthenalresult: c2@2E whichhastheformofawaveequationwithvelocity: 1 c2=)=c Comparingthelastequationwithequation(3.3),andtakingintoaccountthefactthatforopticalfrequenciesthemagneticpermeabilitycanbesetequaltoone,=1,therefractiveindexcanbeexpressedasfollows

PAGE 56

Thisequationallowsustorelatetherefractiveindex,aparameterassociatedwiththewayanelectromagneticwavepropagateswithinamedium,tothedielectricfunction,whichisaconstantdirectlyrelatedtotheelectronicstructureofthemedium. Basedontheaboveanalysis,thesolutionstoMaxwell'sequationswillhavetheformofaplanewaveofangularfrequency!:2 whereE0,H0areconstantamplitudes,ingeneralcomplexnumbers,andkisthewavevector,whichinanon-absorbingmediumisgivenby: =n=! n;(3.39) whereisthefreespacewavelength=2c=!. Inthemostgeneralcaseofanabsorbingmedium,kisacomplexnumberanditrepresentstheenergydissipationasthewavepropagatesthroughthemedium: ~k=! c~n=! c(n+i):(3.40) Substitutingequation(3.40)toequation(3.37),andtakingEalongthezdirection,weobtain

PAGE 57

Thus,inthecaseofanabsorbingmedium,thereisanextraterminthewaveequationwhichrepresentstheexponentialdecayofthewavewithinthemedium.IfwetakeintoaccountthefactthattheintensityofalightbeamisproportionaltothesquareoftheelectriceldI_EE[45{47],andwecomparethiswithequation(3.1),itcanbeshownthattheextinctioncoecientisproportionaltotheabsorptioncoecient: c=4 ;(3.42) whereisthefreespacewavelength.Combiningequations(3.8),(3.34),and(3.42),wecanrelate~,~n,,andtoeachother: 2(3.45) 2(3.46) Thisanalysisprovesthattherefractiveindexanddielectricconstantarenotindependentparameters,andthus,ifweknowoneofthemwecancalculatetheother. Intheanalysisabove,onlyisotropicandhomogeneousmediawereconsidered.Inthecaseofhighlyanisotropicmedia,thepolarizationandinducedcurrentslieinadierentdirectionfromthatoftheelectriceldoftheelectromagneticwave.Inthissituation,thedielectricfunctionbecomesatensorquantity.Theresponseofthemediaiswellcharacterizedbythistensor,howeverifthedirectionoftheelectriceldisnotalongoneoftheprincipaldirectionsofthedielectrictensor,theanalysiscanberathercomplicated.

PAGE 58

Weconsideraplanewave,incidentonasurfaceatz=0,asshowninFigure3.1.Theamplitudesoftheincident,reected,andtransmittedwavesoftheelectricvectorsarethefollowing3 ])(3.50)E0p=[E0pcos'0^x+E0psin'0^z]ei(!t[2n0xsin'0 wherek=2n=,andthesubscriptsp,sdenotethetransversemagneticpolarization,alsoknownasTM,andthetransverseelectricpolarization,alsoknownasTE,respec-tively.Theincident,reected,andtransmittedwavesofthemagneticeldvectorscanbecalculatedfrom: (! c)H=(kE)(3.55)

PAGE 59

Thetotalcomponentsoftheelectricandmagneticvectorsatthesurface,atz=0,fortheincident,reected,andtransmittedwavesare Theboundaryconditionsrequirethetangentialcomponentsofboththeelectricandmagneticeldtobecontinuousattheinterface,atz=0,thus:

PAGE 60

Substitutingthetotalcomponentsoftheelectricandmagneticvectorsatthesurface,z=0,fortheincident,reected,andtransmittedwavesintothelatterequations,weobtainthewellknownFresnelcoecientsforreectionandtransmission: ThereectanceR,denedastheratiooftheintensityofthereectedlighttotheintensityoftheincidentlightonthesurface,andthetransmittanceT,denedastheratiooftheintensityofthetransmittedlighttotheintensityoftheincidentlightonthesurfacearegivenby:

PAGE 61

or,intermsoftherefractiveindices,andthereectedandtransmittedangles,thereectanceandtransmittancearegivenby: Furthermore,incaseofnormalincidence,onanisotropicnon-absorbingmedium,theaboveequationsbecome Theaboveequationscanbeusedevenwhentheangleoftheincidentlightdoesnotmeetthemediumsurfaceatexactlyrightangle.Forexample,comparingtheaboveequationsforreectancefornormalincidentandanarbitraryangle,wecancalculate,4themagnitudeoftheerrorapproximation.Usingn0=1andn1=2wendthatfor

PAGE 62

Incaseofnormalincidence,onanisotropicbutabsorbingmedium,therefractiveindexn1,intheaboveequationsshouldbereplacedbythecomplexrefractiveindex~n1=n1+i1,inorderfortheabsorptiontobetakenintoaccount.Therefore,reectanceandtransmittancearegivenby: (n0+n1)2+21:(3.83) Forotherthannormalincidenceinanabsorbingmediumthecalculationsforthereectanceandtransmittancebecometedious,andapproximationsneedtobeusedforeachcase.Moredetailsonthissubjectcanbefoundinthereferences[48{52]. Althoughinourexperimentalpartweareconsideringthecaseofnormalincidence,wearenotdealingwithsinglesurfacesbutwithmulti-layeredstructures;thisfactmakestheproblemalotmorecomplicated.3.4LightPropagationThroughaSingleLayerStructure

PAGE 63

Weconsideranincidentlightbeamonalayerofrefractiveindexn1betweentwosemi-innitemediaofrefractiveindicesn0,andn2,asshowninFigure3.2. Theelectricandmagneticvectorsoftheincident,reected,andtransmittedwavesattherstinterface,atz=0,arethefollowing:

PAGE 64

andtheelectricandmagneticvectorsoftheincident,reected,andtransmittedwavesatthesecondinterface,atz=d,arethefollowing: Theboundaryconditionsrequirethetangentialcomponentsofboththeelectricandmagneticeldtobecontinuousattheinterfaces,atz=0andz=d,thusweobtain (E+0p+E0p)cos'0=(E+1p+E1p)cos'1(3.97)(E+0pE0p)n0=(E+1pE1p)n1(3.98)E+0s+E0s=E+1s+E1s(3.99)(E+0s+E0s)n0cos'0=(E+1s+E1p)n1cos'1(3.100)

PAGE 65

(E+1peik1d1+E1pe+ik1d1)cos'1=E+2peik2d1cos'2(3.101)(E+1peik1d1E1pe+ik1d1)n1=E+2peik2d1n2(3.102)E+1seik1d1+E1se+ik1d1=E+2seik2d1(3.103)(E+1seik1d1+E1se+ik1d1)n1cos'1=E+2seik2d1n2cos'2:(3.104) UsingtheFresnelcoecients,asweredenedintheprevioussection,thelatterequationscanbewrittenasfollows Thesuxespandshavebeenomittedintheaboveequationsbecauseithasbeenshownthat,therelationsbetweenthevectorspolarizedintheplaneofincidenceandthosepolarizedintheperpendicularplanearethesame[49].OnethingthatshouldbekeptinmindwhileusingtheseequationsisthattheFresnelcoecientsthemselvesdependonthetypeofpolarization.Basedontheserelations,theamplitudesofthelightbeamineachmediumcanbeexpressedintermsoftheamplitudeoftheincidentbeam,andhencethereectedandtransmittedamplitudescanbedeterminedby:

PAGE 66

where =n1d1cos'1(3.111) isthephasechangeuponthelightbeamtraversingthelmonce. Thus,thereectanceandtransmittancearegivenby: Forthesimplecaseofnormalincidence,theFresnelcoecientsexpressedintermsoftherefractiveindiceshavethefollowingcompactform: SubstitutingtheaboveFresnelcoecientsintheexpressionforreectanceandtransmit-tance,weobtaintheexpressionsforreectanceandtransmittanceatnormalincidenceintermsoftherefractiveindicesofthemedia.Incaseofabsorbingmedia,thereal

PAGE 67

where =nmdmcos'm(3.120) isthephasechangeuponthelightbeamtraversingthelayerm=1;2;or3. Thus,thereectanceandtransmittancewillbegivenby: and AnotherwaytoexpresstheFresnelcoecientsr1,r2,andr3isbyusingtheamplitudeandthephasechangeofthelightreectedataninterface.Forexample,theFresnelcoecient,knownalsoaseectiveFresnelcoecient,correspondingtothelightreectedatthebacksurfaceofthesecondlmFigure3.3canbewrittenintheform:

PAGE 68

Therstlm,withrefractiveindexofn1andthicknessd1,canthenberegardedaslyingontopofthesurfaceofamediumwhoseFresnelcoecientis%2ei2,andthereforewecanwrite Ifwenowconsiderasystemofmlayers,theeectiveFresnelcoecientofthelastonecanbeexpressedas:

PAGE 69

where%mandmaregivenby: and where tanm=rm+1sin2m tanm=rmrm+1sin2m Forthe(m1)thlayertheeectiveFresnelcoecientcannowbecalculatedfrom: Thus,theamplitudeandthephaseofthereectedlightbyamulti-layerstructurecanbecalculatedfromtheFresnelcoecientsateachinterfaceandthethicknessesofthelms,bysimplyrepeatingthisprocessuntilalllayershavebeentakenintoaccount. Inthecaseofamulti-layerstructureofabsorbingmedia,therealrefractiveindices,introducedthroughtheFresnelcoecientsmustbereplacedbycomplexquantitiesinordertotakeintoaccountabsorption.Inthiscase,theproblembecomesverycompli-cated.Severalmethodsofapproachingthisproblemhavebeendeveloped,andthereareafewapproximationsthatcanbeappliedbasedonspeciccases.Fortheanalysisofourexperimentaldata,thematrixmethodhasbeenused.

PAGE 70

wherem=kmdm,andrm,tmcanbederivethesamewayasequations(3.114),and(3.116).Theaboverecurrencerelationscanbewritteninamatrixform: ForasystemwithnlayerswerequiretoknowtherelationsbetweenE+n+1andE+0,whichwillallowustoobtainthetransmissioncoecient,andbetweentheE0andE+0,forthereectioncoecient.Basedonequation(3.133),wecanwrite whereEn+1=0sincethereisnonegative-goingwaveinthe(n+1)thlayer,and Therefore,E+n+1,andE0canbeexpressedintermsofE+0,andthisisthewaywecalculatethetransmissionandreectioncoecients.ThismethodisdescribedinmoredetailinChapter4ofreference[49].

PAGE 71

P1Z0n(!0)1 and P1Z01(!0)1 wherePdenotestheprincipalpartoftheintegration.Ascanbeseenfromthelatterequations,therealandimaginarypartsofalinearresponsefunctionarenotindepen-dentwitheachother.Therefore,theKramers-Kronigrelationsallowsus,forexample,tocalculatetherefractiveindexfromtheabsorptioncoecientandviceversa.Thismethodprovidesuswithaveryimportanttoolbecausewecanperformonemeasure-mentwhichwillprovideus,i.e.,thefrequencydependenceoftheopticalabsorption,andthencalculatethedispersion,withouttheneedtoperformaseparatemeasurement. Itshouldbenotedthat,foraphysicalsystemtheresponsefunctiontakenasanexamplethedielectricfunction,shouldsatisfythefollowingrelation:

PAGE 72

~(!)=~(!)(3.140) whichforadielectricsystemcanbeexpressedas: anditrequirestherealpartofthedielectricfunction1,tobeanevenfunctionofthefrequency!,andtheimaginarypartofthedielectricfunction2,tobeanoddfunctionofthefrequency!. ThemostcommonroutetodeterminetheopticalparametersofasystemistoperformreectancemeasurementsandmakeuseoftheKramers-Kronigrelations.ItwasshownintheprevioussectionthattheFresnelreectivecoecientcanbeexpressedintermsoftheamplitudeandphasechangeasfollows where and tan=Im[R] whichforthesimplecasewheretheincidentlighttravelsinvacuum~n0=1beforeitmeetsanabsorbingmedium~n1=n+i,thephasechangeisrelatedtonandofthemediumasfollows

PAGE 73

tan=2 Equation(3.143)canbewrittenas: lnR=ln(!)+i(!)(3.147) andsincethereectanceshouldalsoobeythelawofcausality,wecanusetheKramers-Kronigrelationstocalculatethephasechangedispersion: 1Z0lnR(!0)lnR(!) 21Z0ln!0+! !0!dR(!0) Thedeterminationofthephasechange(!),enablesustoobtaintherefractiveindexnandextinctioncoecientthroughequation(3.146). AfewdrawbacksoftheKramers-Kronigtechniquearethefactsthatonlyasin-glebounceistakenintoaccount,andtherequirementthatthemeasurementsshouldbeperformedoverthecompletefrequencyspectralrange0
PAGE 74

wherem,earethemassandthechargeoftheelectronrespectively,andristheelectrondisplacementfromtheequilibriumposition.Thesecondtermonthelefthandsideoftheaboveequationrepresentsthedampingtermandprovidesforenergylossmechanisms,whichinthecaseofsolidsarevariousscatteringprocesses.Thethirdtermonthelefthandsiderepresentstheharmonicrestoringforce(Hooke'slaw)withwhichtheelectronisboundtothecore.Thetermontherighthandsiderepresentsthedrivingforce,whereElocisthelocalelectriceldactingontheelectron.WeassumethelocalelectriceldvariesintimeasEloc(t)/ei!t,andthatthedisplacementrhasthesametimedependence.Thereforethesolutiontotheequationofmotionis ~r(t)=e m1 (!20!2)i!Eloc(t):(3.150) Theinduceddipolemomentisgivenby: ~p=er=e2 (!20!2)i!Eloc:(3.151) Assumingthatthedisplacementrissucientlysmall,thedipolemomentcanbewritten ~p=er=e(!)Eloc(3.152) wherethefrequencydependentquantityeistheatomicpolarizability.Combiningthelasttwoequations,wecanwriteforthepolarizabilityofoneelectronatom: (!20!2)i!:(3.153)

PAGE 75

Itisobviousfromtheaboveexpressionthat,thepolarizabilityisacomplexquantityduetotheinclusionofthedampingterm,andthusithasaphasedierencecomparedtothelocalelectriceldatallfrequencies. IncasethereareNatomsperunitvolume,themacroscopicpolarizationis Inordertorelatethemicroscopicatomicpolarizabilitytothemacroscopicelectricsus-ceptibility,weshoulddeterminetherelationshipbetweenthemicroscopicelectriceldEloc,andthemacroscopicelectriceldE.Ingeneral,6=Ebecausethelocalelectriceldistheaverageoveratomicsitesandnotoverregionsbetweensites.Forsim-plicity,inthecaseofboundelectronscanbeassumedthatthetwoeldsareequalsincetheinclusionoftherestoringforcecontainsallthenecessaryfeaturesforthedescriptionoftheopticalpropertiesofthesystem.Therefore,itcanbewritten fromwhicheecanbededuced (!20!2)i!:(3.156) Sincethedielectricfunctionisrelatedtothemacroscopicelectricsusceptibilityas~=1+4ee,itcanbeexpressedintermsofthedampedharmonicoscillatorsas: ~=1+4Ne2 (!20!2)i!;(3.157) or ~=1+!2p where!pistheplasmafrequencyinLorentzmodel,denedas:

PAGE 76

Therealandimaginarypartsofthedielectricfunctioncanalsobewrittenas: and Incasetherearemorethanonecharacteristicresonantfrequenciesduetotheoscillationoftheboundelectronwithintheatomandtothelatticevibrations,ortherearemorethanoneelectronsperatom,thedielectricfunctionisclassicallyexpressedas: ~=1+4e2 where Quantummechanically,thedielectricfunctionisexpressedas: ~=1+4e2 The!jrepresentsthetransitionfrequencyofanelectronbetweentwoatomicstatesofenergydierenceE=~!j,andtheparameterfjiscalledoscillatorstrengthandisameasureoftherelativeprobabilityofaquantummechanicaltransition.Incaseoffree

PAGE 77

Inthisanalysis,itwasassumedthattheelectronsareinvacuum,where1=1.1isthecontributionfromthehighfrequencyabsorption,beyondthemeasuredrange.However,insolidstate16=1andthusinthiscase,therstterminequation(3.162),or(3.164),shouldbereplacedby1.3.7.2DrudeModel wherek=m!20isthespringconstantwhichischosensothat!0coincideswiththenaturalfrequencyofanatom.Therefore,theequationofmotionfortheDrudemodelis wheremistheeectiveelectronmass,andthedampingconstanthasbeenreplacedby1=.Therelaxationtime,,characterizestheenergylossduetoscatteringinawaysimilarto.Morespecically,isassociatedwithcollisionsbetweenthefreechargecarriersandimpurities,latticevibrationalphonons,orotherscatteringcentersinmetals. Assumingthatthevelocityvariesintimeasloc/ei!t,thesolutiontotheequationofmotionhastheform:

PAGE 78

~=e m1 1i!E:(3.168) IncaseofNfreeelectronsperunitvolume,thecurrentdensitycanbeexpressedasfollows m1 1i!E:(3.169) Fromthislastequationandequation(3.9),weobtaintheACconductivity: ~D(!)=Ne2 m1 1i!=0 where m(3.171) iscalledtheDCconductivity,anditisthezerofrequencylimitoftheDrudeconductivity.TherealandimaginarypartsoftheconductivityintheDrudemodelare ThedielectricfunctioninDrudemodelisgivenby: ~D(!)=1!2Dp where!DpistheplasmafrequencyinDrudemodel,denedas:

PAGE 79

andtypicallyliesinthevisibleorultravioletspectralregion.Therealandimaginarypartsofthedielectricfunctioncanbeobtainedeitherfromequations(3.12)and(3.13)andequations(3.172)and(3.173),orfromequations(3.160)and(3.161)bysimplysetting!0=0,=1=,andm=m: !(1+!22):(3.177) Inthelimitoflowfrequency,where thedielectricfunctionfromequation(3.174),canbewrittenas ~D(!)=140 !!i (3.179) thus,therealandimaginarypartsofthedielectricfunction,are Forsucientlylowfrequencies!1,wehavej1jj2j,andtherefore,equations(3.45)and(3.46)willgive Usingequations(3.41),(3.42),(3.180),(3.181),and(3.182)anexpressionfortheskin,orpenetration,depthcanbewrittenasfollows

PAGE 80

Thepenetrationdepthprovidesameasureofthedecayoftheelectriceldwithinthemedium,andascanbededucefromtheaboveequation,thehigherthedcconductivityofthemediumtheshorterthepenetrationdepthoftheACeldatafrequency!.3.7.3Drude-LorentzModel ~=1+~D+~L(3.184) where1isthecontributionfromthehighfrequencyabsorption,beyondthemeasuredrange;invacuum1=1,~DistheDrudedielectricfunctiongivenbyequation(3.174),and~ListheLorentzdielectricfunctiongivenbyequation(3.162).Therefore,thedielectricfunctioncanbeexpressedas: ~(!)=1+Xj!2pj Thisexpression,obtainedbycombiningtheDrudeandLorentzmodelsisusedtottheexperimentalreectancedataandprovidesanalternativetotheKramers-Kronigmethodforextractingtheopticalproperties.TheadvantageoftheDrude-LorentztechniqueovertheKramers-Kronigtechniqueisthefactthatitdoesnotrequirethedatatobeobtainedoverawidespectralrange.Thettingprocedurecanbeperformedoveranitespectralrange,andnoextrapolationprocessesareneeded. AnothermajoradvantageoftheDrude-LorentztechniqueovertheKramers-Kronigisthefactthat,thistechniquecanbeemployedfortheanalysisofthinlmsandmulti-

PAGE 81

MoredetailedinformationabouttheDrudeandLorentzmodelscanbefoundinthereferences[48,53].3.7.4SumRules with!pgivenbyequation(3.159),whilefortheDrudemodelthef-sumrulecanbeexpressed SumrulescanfrequentlybefoundtobeexpressedintermsofaneectivenumberofelectronsperatomNe,contributingtotheopticalpropertiesoveranitespectralrange,asshownforexampleinreference[48].

PAGE 82

Thischapterdescribestheexperimentalequipmentandthetechniquesusedtohandlethematerials,togrowthelms,tomonitortheirstability,andtoperformmea-surementsofabsorbance,reectance,and/ortransmittanceatnearnormalincidenceoverawidefrequencyrange,from20cm1to40,000cm1,orequivalently2.5meVto5eV.Inordertocoverthisentireregion,varioustypesoflightsources,detectors,grat-ings,beam-splitters,lters,anddierenttypesofspectrometersmustbeused.Narrowregionsoftheelectromagneticspectrumaremeasuredseparately,usingtheappropri-ateopticalcomponents,andthenmergedtogethertoformthespectrumoftheentirefrequencyrangeofinterest. Thefollowingdierentspectrometersareusedtomeasuretheopticalpropertiesofoursamples: Amoredetaileddescriptionofthesespectrometerswillfollowlaterinthischapter.73

PAGE 83

Averyimportantpartofthissystemisthechemicalpuriermountedonthebacksideofthedrybox,showninFigure4.2.Thegaswithintheboxcirculatescontinuouslythroughthepurier.Thepuriercanistercontainsamoistureabsorbent(molecularsieves)andanoxygenreducingagentQ1,whichisamaterialconsistingofnelydividedcopperonanAluminaMatrix(madebyDowChemicalCompany)[54].Asthegasfromthegloveboxpassesthroughthepurier,theabsorbentremovesthewatervapor,andtheoxygenreducingagentremovestheoxygenbeforethegasentersthecontrolledatmosphereofthedryboxagain.Theoxygenreducingagent,copper,combinesche-

PAGE 84

Figure4.1:SchematicdiagramofDryorGloveBox.micallywiththeoxygenformingcuprousorcupricoxide.Whentheagentissaturatedwithoxygen,thepurierisregeneratedwiththeuseofaspecialregeneration/forminggasthatcontains3to10percenthydrogen.However,itshouldbenotedthattherearecertainchemicals,likesulfurandsulfurcompounds,suchasH2S,SO2,SO3,etc.,thatwillpoisonthereactantmaterialandtheyshouldnotbeusedinthedrybox.Iftheiruseisnecessary,asuitabletrapshouldbeinstalledinadvanceinthecirculatingpathtopreventthesechemicalsfromgettingintothepurier. Otherpartsthatcanbemountedonthedryboxareanoxygenanalyzer(modelAO-316-C,VAC,mountedonthedryboxinourlaboratoryinthePhysicsdepartment)tomonitorthetraceoxygeninthecontrolledatmosphere,andapedatrol(modelHE-63-P,VAC,mountedonthedryboxinourlaboratoryintheChemistrydepartment)toprovidebothautomaticandmanualpressurecontroloftheatmosphereinsidethedrybox.DuetotheabsenceofanoxygenanalyzermountedonthedryboxintheChemistrylaboratory,othermethodshadtobeemployedforoxygenandmoisturetestingoftheinertatmosphere.Sinceitisrecommendedthattheatmosphereinsidethedrybox

PAGE 85

Figure4.2:Schematicdiagramofthechemicalpurierandthegasowthroughthechemicalcirculation.becontinuouslymonitored,abottleofdiethylzincandabottleofborontriuoridediethyletheratearekeptinsideatalltimes.Diethylzincisusedforoxygendetectionandborontriuoridediethyletherateisusedforwaterdetection.Toperformthetest,thebottlesshouldbeuncoveredandincasesmokeisobserved,thentheatmosphereiscontaminatedandaregenerationofthepurierisneeded;incasesmokeisnotobserved,thenthelevelsofoxygenand/orwaterarewithintheacceptablelimits.However,itshouldbenotedthatthistestonlyprovidesanindicationoftheconditionofthedryboxatmosphere.4.2ElectrochemicalMethods

PAGE 86

Animportantparameterintheelectrochemicalexperimentsisthemeasurementandcontrolofthecellpotential,whichisthedierenceinelectricpotentialbetweentheelectrodesinanelectrochemicalcell.1Themagnitudeofthepotentialdierenceataninterfacedeterminesthedirectionandtherateofchargetransfer[55].4.2.1ElectrochemicalPolymerization

PAGE 87

Figure4.3:Schematicdiagramofthree-electrodecongurationcell.4.2.2CyclicVoltammetry(CV)

PAGE 88

SpectroelectrochemicalexperimentswereperformedusingaStellarNetPhotoDiodeArray(PDA)UV-Vis-NIRspectrophotometer(madebyStellarNet,Inc.)locatedinside

PAGE 89

TheStellarNetspectrophotometer(EPP2000CUV-VisandEPP2000NIRIn-GaAs)isaportableberopticinstrumentusedforabsorbance/transmittancemea-surementsintheUV-Vis-NIRrange.Thisspectrophotometerutilizesaspecialconcaveholographicgratingforabberationcorrectioninordertoprovidebetterimaging.With-outmirrors,thegratingmaintainslowstraylight,andtheinstrumentlossesaremini-mized.Inaddition,theabsenceofmovingpartsmakestheinstrumentmorestableandveryreliable.ThissystemisequippedwithalinearPhotoDiodeArraydetectorandatungstenhalogen(tungstenkryptonbulb)source.Thespectrometercovers190-890nmfortheUV-Visregion,and900-1600nmfortheNIRregion. TheUV/Vis-NIRVarianCary500spectrophotometerisadoublebeammono-chromatorthatcoversthewavelengthrangefrom175-3300nm.TheinstrumentisequippedwithahighperformanceR928photomultiplierdetectorfortheUV-Visregionandanelectrothermallycontrolledleadsulde(PbS)photocellfortheNIRregion.TheavailablelightsourcesforthissystemareatungstenhalogensourcewithquartzwindowforNIRandvisibleregions,andadeuteriumarcUVsource. Fortypicalthinlmpolymersamplesdepositedpotentiostaticallyonanumberofdierentsubstrates,athree-electrodecellcongurationhasbeenused,astheonede-scribedabove,toallowpotentialapplicationwhilemonitoringtheabsorption/transmissionspectra.Thesamplesareplacedinmonomerfreesolution,andtheelectrodesarecon-nectedtoanEG&PARmodel273Apotentiostat/galvanostat.ForECdevices,onlytheUV-Vis-NIRVarianCary500spectrophotometerisused,inwhichthethree-electrodecellisreplacedbyasuitableholderavailableintheCary5kit.Inordertoapplyavolt-ageacrosstheECdevices,thecounterandthereferenceleadshavetobeconnectedtooneanother.Finally,forbothcases,ECpolymerlmsonelectrodesandindevices,theexperimentisperformedbysequentiallysteppingtheappliedpotentialineither0.1Vor0.2Vincrements,startingfromapotentialatwhichthepolymerisinitsneutral

PAGE 90

TheUV/Vis-NIRVarianCary500spectrophotometerprovidesonemoremethodofECpolymeronelectrodeorindevicecharacterizationinadditiontotheoneoutlinedabove.Usingthesameexperimentalsetupwecanperformakineticsexperimentandmeasurethespeedwithwhichthematerialisabletoswitchbetweentwoextremastates.Inthiscase,asinglewavelengthisused,thewavelengthofmaximumcontrastwhichcanbedeterminedfromtheexperimentdescribedabove,andasquarepotentialwaveformisappliedatdesiredtimeintervals.Thepercentagetransmittance(T%)ismonitoredatthewavelengthofmaximumcontrastmaxasafunctionoftime,whiletheECpolymeronelectrode(insolution),orindevice,isrepeatedlyswitchedbetweenthetwoextremastates.4.3.2InterferometricorFTIRSpectrometer Asaresultoftheseadvantagesaninterferometerhaslargeresolvingpower,fastsamplingtimes,reducedstraylight,andhighsignal-to-noiseratio(S/N)[58{60].How-ever,eventhoughtheJacquinot(orthroughputoretendue)advantageholdsforlowandhighfrequencies,theFellgett(ormultiplex)advantageislostforhigherfrequencies.Thisisduetothefactthatalthoughfortheinfraredregionthenoiseisusuallydetector

PAGE 91

Morespecically,intheinfraredregionthesignal-to-noiseratioforaninterferom-eterisproportionaltothesquarerootofthetotaltimeT,requiredforascanofabroadband: (S=N)I/T1=2;(4.1) butinthevisibleregion,thesignal-to-noiseratioisalsoproportionaltothesquarerootofthesourceintensityI: (S=N)I/[(T=M)I()]1=2;(4.2) whereMisthenumberofspectralelementsofwidthinabroadband=21.Comparingtheaboveresultswiththesignal-to-noiseratioforamonochromator,whichintheinfraredregionis (S=N)M/(T=M)1=2;(4.3) andinthevisibleregionis (S=N)M/[(T=M)I()]1=2;(4.4) weconcludethatalthoughtheinterferometerhasanadvantageoverthemonochromatoratlowfrequencyregionsduetotheFellgettadvantage,equations(4.1)and(4.3),athigherfrequenciesthisadvantageislostandthetwospectrometersprovidethesamesignal-to-noiseratio,equations(4.2)and(4.4)[58,59]. Thereareseveraladvantagesandlimitationsforbothtypesofspectrometers.TheFellgett(ormultiplex)advantagemakestheinterferometeranexcellentchoiceforthelowfrequencyregionsbutthelossofitathigherfrequenciesincombinationwiththe

PAGE 92

Inmoredetail,assumethatthesourceisemittingmonochromaticplanewavesoftheform: InaMichelsoninterferometer,aftertheamplitudesplittingtakesplace,therearetwobeamsthathavetraveleddierentdistancesz1andz2,beforetheyrecombinedatthebeamsplitter.Aswasmentionedabove,eachbeamhasundergoneonereectionandonetransmissionatthebeamsplitter.

PAGE 93

Figure4.4:SchematicdiagramoftheMichelsonInterferometer. Ifrbsisthereectioncoecientandtbsthetransmissioncoecientofthebeamsplitter,thentheresultanteldoftherecombinedbeams,usingthesuperpositionlaw,willbe andtheintensity(irradianceoruxdensity)IR(z1;z2;)atthedetectorwillbe wherez1z2=istheopticalpathdierencebetweenthetwobeams.Hence,equation(4.7)canbewritten

PAGE 94

Inthecaseofapolychromaticsourceemittingacontinuousspectrumfrom=0to=1,thetotalintensityatthedetectorwillbe wherethersttermontherightsideoftheaboveequationisconstantanditrepresentsthetotalintensityemittedbythesource.Atzeropathdierence=0theintensityatthedetectoris Thus,equation(4.9)canbewritten [IR()1 2IR(0)]=2jrbstbsj21Z0E20()cos(2)d;(4.11) inwhichthequantity[IR()1 2IR(0)]isknownasinterferogram.TheFouriercosinetransformoftheinterferogramprovidestheactualspectrum: 2IR(0)]cos(2)d:(4.12) However,inpracticeitisnotpossibletomeasureaninterferogramoveraninnitepathdierence.Theniteopticalpathdisplacementresultsintheintroductionofnumerouspeaksintothetransformedspectra,inadditiontothemain,whichisonecenteredapproximatelyat=0.Thesepeaksneedtobecorrectedbecausetheycausea\leakage"ofspectralintensity;theintensityisnotlocalizedtothemainpeakat=0anymorebutitisdistributedtoallpeaks.Thesepeaksarecalledsidelobes,or\feet",

PAGE 95

andthetriangular,alsoshowninFigure4.5: Lj,ifjzjL0,ifjzj>L(4.14) whereListhemaximumpathdierence.OtherapodizationfunctionsthatcanbeusedarethetrapezoidalwhichisshowninFigure4.5,theHapp-Genzel,theBlackman-Harris,andtheNorton-Beer(week,medium,orstrong).Furtherdetailsontheeectsandthechoicesofotherapodizationfunctionsused,canbefoundintheliterature[58,59,61].Itmustbenotedthough,thattheconvolutionoftheapodizationfunctionwiththeinterferogramwillresultinreductionoftheresolution<,where<1=L. Anotherassumptionthatwasmadeinthissection,whichisnotalwaystrueinarealexperimentisthattheinterferogramissymmetricwithrespecttothezeropathdierence(ZPD)=z1z2=0.Usually,thesamplingoftheinterferogramstartsat=0,andtheinterferogramfunctionhasamaximumvalueatthispoint;butifthisisnotthecase,thenallsamplingpointsaredisplacedbythesamesmallamount.Thisresultsinanasymmetricinterferogramwithrespecttothe=0point,anditintroducesaphaseerror.

PAGE 97

Furthermore,inthecalculationofthespectrum,equation(4.12),asymmetricalinterferogramwasassumed.However,inthecaseofanasymmetricalinterferogram,acompletecosineandsineFouriertransformationisneededforthecalculationofthespectrum.TheapplicationofthecosineFouriertransformationalonewillresultinadistortedspectrum,andthecreationofspuriouslines.UnfortunatelyusingthecompleteFouriertransformation,twiceasmanypointsneedtobecollectedinordertoachievethesamespectralresolutionwhichtranslatesintolongertimesfordataacquisition.Furthermore,theuseofaphasecorrectionmodeisnecessaryinorderforthephaseerrortobecorrected,andforthescanningtimetobekeptshort. Finally,thedataacquisitionrequirestheuseofadigitalcomputerforthecalcu-lationofthespectrumfromtheinterferogramfunction.Inorderforthiscalculationtotakeplace,therecordeddatahavetobedigitizedintoanumberofdiscretevalues.Forthisreason,theinterferogramissampledatstepsofpathdierence.ThediscretenatureoftherealexperimentcanbehandledmathematicallybyusingDiracdeltafunc-tions.Theinterestedreadercanndmoredetailsaboutthisprocessintheliterature,references[59,62].Thespectrumthatisobtainedfromthesampledinterferogramisperiodic,itrepeatsitselfatmultiplesof.Iftherepeatedspectraoverlapthenanerroreect,calledaliasingor\folding",isintroduced.Forthiserrortobeavoided,themaximumfrequencyofthetruespectrummustobeythefollowingcondition: max= Othererrorscanoccurfromelectronicltering,misalignment,opticaleectscausedbyvariouspartsoftheinstrumentoptics,suchastheuseofnon-idealmirrors(ideal:100%reective),andnon-idealbeamsplitters(ideal:non-absorbing,50%transmissiveand50%reective),andnotaccurateadjustmentofthemovablemirror.Theseerrorsintroduceaphasefactorthatusuallycanbetakencareofbytheuseofaphasecorre-ctionmode.Themostcommonlyusedcorrectionmode,whichisalsotheoneusedinourequipment,istheMertzbutthereareseveralmorethatcanbeused.

PAGE 98

TheprincipleoftheFT-IRBruckerinterferometerisbasedontheMichelsoninter-ferometerdescribedabove.Thelightbeamgeneratedbythesource,asshowninFigure4.6(a),goesthroughanautomatedcircularaperture3tothebeamsplitter,mountedonanautomaticchanger4.6(d),whereisdividedintotwoparts.Eachpartofthebeamisreectedbyxedmirrors,andafterimagedontothefacesofthemovabledouble-sidedmirror4.6(e),thetwobeamsrecombineatthebeamsplitter.Partoftherecombinedbeamisreectedbyaseriesofmirrors,focusedinthesamplechamber4.6(i)or4.6(j),andsenttothedetectorthroughanothersetofmirrors.Therestoftherecombinedbeamisreectedbacktothelightsource.Therecordedinterferogramhasamaximumamplitudewhenthetwoarmsoftheinterferometerhaveazeropathdierence(ZPD).Whenthedouble-sidedmirrormovesduetoascanningmechanismataconstantspeed,apathdierenceof=4t,wherethetimetismeasuredfromtheZPD,isintroduced.

PAGE 99

Figure4.6:SchematicdiagramofBrucker(IBM)113vFT-IRspectrometer. Figure4.7:Schematicdiagramofthereectancestage.

PAGE 100

Figure4.8:Schematicdiagramoftheliquidheliumcooleddetector.Inthissystemthemirrorisair-bearingandismovingcontinuously.Thecalibrationofthepositionofthedouble-sidedmirrorismonitoredbyaHe-NelaserFigure4.6(g)andawhite-lightreferenceinterferometer. Thedetectedsignalisampliedbyapreamplier,andsenttoa16-bitanalog-to-digitalconverter.Thedataarecollectedbyacomputersystem,wheretheinterferogramisFouriertransformedtoproducethesinglebeamspectrum.Thereectancespectrumisthencalculatedbytakingtheratioofthesinglebeamspectrumofthesampletothesinglebeamspectrumofthebackgroundreference,usuallyanaluminum(Al)mirror.Additionalcalculationsneedtobedoneinthiscase,inorderfortheAlmirror/referencereectancetobetakenintoaccount.4Forcaseoftransmittancemeasurements,the

PAGE 101

Table4.15providestheexperimentalparametersforthespectralregions,andthevarioustypesofbeam-splitters,opticallters,etc,foreachoneoftheregions. Table4.1:Brucker113vFT-IRinterferometerparameters. Spectral Beam Scanner Phase Optical Apod. Range Source Detector Splitter Velocity Correct. lter function (cm1) (m) (kHz) Mode 20-100 Mylar23 12;29.73 Black 35-200 Hgarc BoloMylar12 7;12.5 Mertz PolyNorton 150-650 meter Mylar3.5 7;12.5 ethylene Beer 400-5500 Globar DTGS Ge/KBr 7;12.5 Open (med) AsshowninTable4.1,theFT-IRspectrometeremploystwothermaldetectors,abolometerandadeuteratedtriglycinesulfateorDTGS.Duringanexperiment,thepartoftherecombinedbeamthatreachesoneofthesedetectors,getsabsorbedfromasensor,andcausesheatingeects.Thischangeofthetemperatureinducessomemeasurableparameterchange.Morespecically,thebolometersignal-detection-operationisbasedonthechangeoftheelectricalresistanceofasemiconductormaterial,whereastheDTGSdetectorreliesonthechangeofthespontaneouspolarizationofadielectricmaterialasafunctionofthetemperature[63]. Thebolometer,showninFigure4.8,isusedforsignaldetectioninthefarinfraredregion.ItisaHe-cooled4.2KSibolometerandconsistsofthreemainparts:liquid laser(cm1);(4.16) wheretheHe-Newavenumberislaser=15;798cm1.

PAGE 102

Inamonochromator,thelightfromthesourceformsanarrowimagethroughanentryslitandthebeamiscollimatedtowardsagratingwhichspatiallyseparatestheindividualfrequencycomponents.Anexitslitselectsanindividualfrequencywhichiseithertransmittedthrough,orreectedbythesample,dependingonthetypeofmeasurementinprogress,andnally,isfocussedonthedetector.Aspectrumisproducedafterscanningtheentireregionofinterestbyrotatingthedispersiveelementstepwise,bychangingtherotationangle.Therefore,thewavelengthisrelatedtotheelementorientation. Thegratingdiractionequationfornormalincidenceis

PAGE 103

wherem=0;1;2;3;:::istheorderofdiraction,Iistheincidentangle,Disthediractedangle,anddisthegratingconstant(innumberoflinespermm).Asthegratingrotates,theincidentanddiractedangleschangebuttheangle2betweenthemremainsconstant,becauseitdependsonthegeometryofthemonochromatorandhenceisxed,seeFigure4.10. AnimportantparameterofthissystemistheangulardispersionD=,whichgivesthechangeofdiractionanglethatcorrespondstoasmallchangeinwavelength.

PAGE 104

Figure4.10:Schematicdiagramofthediractiongrating,wheredisthegratingcon-stant.Theangulardispersionisobtainedbydierentiatingthegratingequationwithrespecttothewavelength,whilekeepingtheincidentangleIxed: =m dcosD:(4.18) Thisquantityisgreaterforsmallergroovespacingsd,largerordersm,andlargerdif-fractionanglesD[64]. Furthermore,onecancalculatethelineardispersionL=attheexitslitofthemonochromator,whichwillvarywiththeoutputfocallengthf,andthediractedangleD.Morespecically,thelineardispersionistheproductofthefocallengthandtheangulardispersion =fD =fm dcosD:(4.19) Anotherparameterthatshouldbementionedistheresolutionoftheinstrument,whichisdenedastheabilityofthemonochromatortoseparatetwospectrallinesthat

PAGE 105

min=Nd(sinI+sinD) whereministhesmallestwavelengthdierencethatcanberesolved,isthemeanwavelength,andNdisthegratingwidth. Themaindisadvantageofthistypeofspectrometeristheslowscanningprocess.Thecompletionofasinglescanistimeconsumingduetothefactthatinformationfromeachfrequencyiscollectedindividually.MicroscopePhotometer However,thereareafewotherimportantopticalcomponentsandfeaturesofthissystemthatshouldbementioned.Therstoneistheobjective,item(10)inFigure4.11,thebasicfunctionofwhichistogatherthelightthatisreectedfrom,ortransmittedthroughthespecimen,andproduceanaccurateimageofthespecimenintothebodyof

PAGE 106

Otheropticalcomponentsofthemicroscopeincludecollectorlenses,tubelenses,mirrors,variableirisandelddiaphragms,lightshutters,lightmodulator,prism,andobservationtubes.Moreover,amotorized8xlterturretwitheightapertures,item(3)inFigure4.11,isplacedclosedtothedetectoritem(1).Alter,selectedwiththesoftware,canbeaccommodatedthroughanopeningineachaperture.Finally,ascanningstage(4"x4")isbuiltintoastandbetweentheobjective,item(10),andthecondenser,item(12).Themountingframehasplateswithorwithoutanopeningforreectanceortransmittancemeasurementsrespectively.ThestagecanbetranslatedintheXandYdirections,andcanbepreciselymovedintheZdirectionwithacoarseandnefocusingmechanism. Formeasurements,thelightbeamspotsizeonthespecimenisselectedwiththevariableirisdiaphragmandhasaminimumspotsizeof1m.Independentofthespotsize,theentranceandexitslitsmaybesettothefollowingxedbandwidths:4nm,10nm,20nm,40nm,and80nmfortheNIRgrating,and1nm,2.5nm,5nm,10nm,and20nmfortheVis/UVgrating[67].Polarizersandanalyzersarealsoavailableinthissystemforpolarizedmeasurements.

PAGE 107

Figure4.11:SchematicdiagramofthereectedandtransmittedlightbeampathatZeissMPM800MicroscopePhotometer. Forreectancemeasurementsthelightbeamgeneratedbyoneofthesources,housedontheupperbacksideofthemicroscopeitem(15)inFigure4.11,passesthroughthecollectorlens,aperture,andelddiaphragmtothebeamsplitteritem(7).Thepartofthelightbeamthatisbeingreectedgoesthroughtheobjectiveitem(10),toilluminatethespecimenitem(11),whichisplacedonthestage.Lightfromthesurface

PAGE 108

whereQ(Quotient)isthereectancespectrumafterspectralcorrection,O(Object)isthesinglebeamspectraofthesample,S(Standard)isthesinglebeamspectraofthesourcelamp,8P(Parasitic)isthemeasuredspectrumofthestraylight,andR(Reference)isthereectanceofthestandard.ThedefaultfromthesoftwareissetR()=100%,sothequotientneedsanadditionalcorrectionforthereference,whichforthisworkisthealuminumreectance. Fortransmittancemeasurements,thelightbeamgeneratedbyoneofthesources,whicharehousedonthelowerbacksideofthemicroscopeitem(17)inFigure4.11,passesthroughaseriesoflters,theelddiaphragmitem(13),throughthecondenseritem(12),andthroughthespecimenitem(11),whichisplacedoveranopeningonthestage.Lightisthengatheredbytheobjectiveitem(10),andisdirectedeithertotheeyepiecesthroughthephotometertubeforobservation,ortothecameraportforphotomicrographyitem(5),orthroughanaperturediaphragm,ltersandagrating,tothedetectoritem(1)inorderforthesignaltoberecorded.Itshouldbenoticedthatthearm,whichhousesthetwosources,needstobeplacedaccordingtothetypeofmeasurementsrequired,upperpositionforreectancemeasurementsandlowerpositionfortransmittancemeasurements.Thebasicformulaforcalculatingthetransmittanceatwavelengthis

PAGE 109

whereQ(Quotient)isthetransmittancespectrum,O(Object)isthesinglebeamspectraofthesample,andS(Standard)isthesinglebeamspectraofthesourcelamp. Thesystemalsooerstheoptionforphotoluminescencemeasurementsbuttherewasnouseforthisoptionthroughoutthiswork.Moreinformationaboutitcanbefoundintheliterature[67,68].Perkin-ElmerMonochromator Duringtheexperimentalmeasurementsthelightbeamgeneratedbyoneoftheabovesourcespassesthroughachopper,asetofloworhighfrequencyband-passlters,andanarrowrectangularentranceslit,andentersthegratingmonochromator.Thechoppergeneratesasquarewavesignalforlock-indetectionandtheltersrejecttheunwantedhigherorderdiractionfromthegrating,whichoccursatthesameangleasthedesiredrst-ordercomponent.Thiscaneasilybeseenbythediractionequation:dsin=n,wheredisthegratingconstant,asshowninFigure4.10.Atanangle,therst-ordercomponentofwavelength,forn=1,thatsatisesthediractionequationisselected,whileallthehigherordercomponents,thatpassthroughtheslitareabsorbedbythelters.Theangleischangedatpredeterminedintervalsbyrotating

PAGE 110

Table4.2:Perkin-Elmergratingmonochromatorparameters. Frequency Grating SlitWidth Source Detector (cm1) (lines/mm) (m) 801-965 101 2000 905-1458 101 1200 1403-1752 101 1200 1644-2613 240 1200 Globar Thermocouple 2467-4191 240 1200 (SiC) (TC) 4015-5105 590 1200 4793-7977 590 1200 3893-5105 590 225 4793-7822 590 75 7511-10,234 590 75 QuartzTungsten LeadSulde 9191-13,545 1200 225 (W) (PbS) 12,904-20,144 1200 225 17,033-24924 2400 225 22,066-28,059 2400 700 SiPhoto 25,706-37,964 2400 700 Deuterium Conductance 36,386-45,333 2400 700 (D2) (576) thegrating.Therotationofthediractiongratingisachievedbytheuseofastepmotorwhichprovidesscanningofafrequencyrangesequentially.Thediractedbeamleavesthemonochromatorthroughanarrowrectangularexitslit.Theexitslitswidthdeterminestheresolutionofthemonochromator.Anincreaseintheslitswidthwillresultinanincreaseofintensityofthelightbeam,resultinginhighersignal-to-noiseratio(S/N)atthecostoflowerresolution.MirrorM1,showninFigure4.12,isareferencemirrorwhichforreectancemeasurementsisreplacedbythesample.Fortransmittancemeasurements,thesampleismountedatadierentplaceasshowninFigure4.12.Inthisposition,thetransmittancesampleandtheSidetectorarethetwofocalpointsoftheellipsoidalmirrorinbetweenthem.

PAGE 111

Furthermore,inthissystemtherearethreedetectorsavailabletocoverthefre-quencyrangefrom800cm1to45,000cm1,athermocouple(TC),aroomtemperatureleadsulde(PbS)photoconductor,andasilicon(Si)photodiodedetector.Thether-mocoupleisathermaldetectorthatconsistsofaseriesconnectionoftwodissimilarmaterials.Whenthelightbeamreachesthisdetector,itisproducinganelectricalvoltageasafunctionofthetemperaturedierencebetweenthetwojunctions.Thether-mocouplewasnotinstalledinthemonochromatorduringthiswork,sothisdetectorhasnotbeenused.Theleadsulde(PbS)detector'soperationisbasedonthemechanismofphotoconductivity.TheincidentphotonsareabsorbedinaPbSlayer,andcreateanelectron-holepairbyexcitinganelectronfromthevalencetotheconductionband.Theelectronandholemovefreelyandthus,theconductivityincreasestemporarilyuntiltheyrecombine.ThethirddetectorofthissystemisaSiphotodiodeorphotovoltaicdetector.Thisdetectorismadebybuildingap-njunctioninasemiconductor.Theincidentlightbeamisabsorbedatthejunctionofthetwomaterials,andproducesanexternalcurrenttoowordevelopsavoltage[63]. Finally,therecordedsignalfromthedetectorsisampliedbyaSR510lock-inamplier(StanfordResearchSystems),averagedovertime,andconvertintoadigitalsignal.ThecollecteddataarethentransferredtoaPCcomputerandstoredforanalysis.Itmustbenotedthatincaseofreectancemeasurements,thedataacquisitionprograminthissystemcorrectsautomaticallyforthereferencemirror.Moreinformationaboutthissystemcanbefoundintheliterature[69].

PAGE 112

Figure4.12:SchematicdiagramofthePerkinElmer16Uspectrometer.

PAGE 113

Electrochromic(EC)materialsexhibitreversiblechangesintheirelectronicstruc-tureandtherefore,intheiropticalpropertiesdrivenbyachangeintheelectrochemicalpotential.AmongallthegroupsofECmaterials,mentionedinChapter2,conjugatedpolymershavebeensingledoutoverthepastdecadeaspromisingECmaterialsduetotheireasyprocessability,lowcost,rapidswitchingtimes,andhighopticalcontrastratios.Anotherimportantaspectofthistypeofmaterialsistheabilityofnetuningoftheenergybandgapoverthewholevisiblespectrumthroughthederivatizationofthemonomerstructure,whichprovidesaccesstodierentcolorstates. Previouslymaterialswereconsidered\electrochromic"whentheycouldundergoavisiblecolorchangeuponapplicationofanexternaleld.Latelythedenitionofelectrochromismhasbeenextendedtoincludeamulti-spectralenergymodulationbyreectanceortransmittancethatmightcoverultraviolet(UV),nearinfrared(NIR),mid-infrared(MIR),andmicrowaveregion(MW),with\color"correspondingtotheresponseofthedetectorsatthesewavelengths[70{74]. Reectiveelectrochromicdevices(ECDs)aredesignedtomodulatethereectanceofincidentelectromagneticradiationuponapplicationofvoltage.Themostcommonapplicationsofthesetypeofdevicesincludemirrors,opticaldisplays,spacecraftther-malcontrol,andcamouage[33,75].Inorderforconjugatedpolymerstobeusefulaselectrochromic(EC)materialsforcommercialapplications,theyshouldexhibitlargechangesinreectanceortransmittancebetweenneutralanddopedstates(largeopticalcontrast%R),rapidredoxswitchingtimes,andlongtermstability[76]. Ourgroupandcollaborators,inProf.Reynolds'group,havestudiedexten-sively[25,41,77,78]theopticalpropertiesofdioxythiophene-basedconjugatedpolymers(PXDOT),andhaveincorporatedtheminelectrochromicdevices.Inthischapterwe104

PAGE 114

Furthermore,weaddresstheoptimizationofthistypeofdevicesthroughthede-velopmentofafabricationmethodfortheremovalofunwantedabsorptionbandsinthemid-infraredregion.ThesebandshindertheperformanceoftheECdevicesbydrasticallyreducingthelifetime(longtermswitchingstability),andtheelectrochromiccontrast%R.Furtherimprovementsthatneedtobeaddressed,orhavealreadybeenconsidered,arealsobeingdiscussed.5.1ECDsFabrication

PAGE 115

Figure5.1:Side-viewschematicdiagramofadualpolymerreectiveelectrochromicdeviceinacircuit[77]. Indetail,theelectrochemicalcelldepictedinFigure5.1consistsoftwelvelay-ers.Thedeviceisexibleifthepolyethylenewindowisused.Thelayersare,fromlefttoright:opticalwindow,gelelectrolytelayer,anupper,initiallyp-doped,poly-merlm(WEpolymer)ongold-coatedMylarreectiveconductingsubstrate(poly-mer/gold/Mylar),apolypropyleneporousseparatorsoakedinelectrolytegel(electrolytegel/separator/electrolytegel),alower,initiallyneutral,polymerlm(CEpolymer)ongold-coatedMylarreectiveconductingsubstrate(polymer/gold/Mylar),andapoly-ethylenesheetforbacksupport.Animportantissuethatshouldbenotedisthatthe

PAGE 116

Fortheencapsulationofthecell,twomethodscanbeemployeddependingontheopticalwindowthatisbeingused.Inthecaseofpolyethylene,twobigpiecesofpolyethyleneareusedasthebacksupportandthefrontwindow,andthethreesidesaroundtheelectrodesareheat-sealed,whereasthefourthsideisusedtoevacuatethecell,andthenisheat-sealedtoo.Copperwiresthathavebeenattachedtothegoldelectrodeswithconductivesilverpaste,comeoutofthe\polyethylenebag"throughpunchedholesthathavebeensealedusinganepoxy-basedglue.Theadvantageofthismethodisthatitpermitstheuseofliquidelectrolytes.InthecasewheretheopticalwindowusediseitherZnSeorglass,theencapsulationisprovidedbytheelectrolytegel.Theelectrochromiccellisself-sealedalongtheedgesastheACNintheelectrolyteevaporates,andthePMMAbecomesinsoluble.Thus,furtherevaporationisminimizedandleaksareprevented. Finally,thecellisconnectedtoanelectricalsupply,asshowninFigure5.1.Asdif-ferentvoltagesareapplied,thepotentialdierencebetweenthetwoelectrodeschanges,andcausesthedopinglevelsoftheconjugatedpolymerlmstochangereversibly.Figure5.2showsaphotographofthetopviewofarealelectrochromicdeviceinitstwoextremastates.Ontheleftside,theactivepolymerisintheneutral,colored,stateandontherightsideisswitchedtoitsoxidized,transparent(bleached),state.Thiscolorchangeisattributedtothecreationofnewmid-gapstates,upondoping.Thedopingmodiestheelectronicstructureofthepolymerscausingthedepletionofthe-transitionwhile

PAGE 117

Figure5.2:Top-viewphotographofanECdevice,usingPProDOT-Me2polymerastheworkingelectrodepolymer,initstwoextremastates.Leftside:PProDOT-Me2inneutralstate,coloredstate(Vcell=1:1V).Rightside:PProDOT-Me2inp-doped,oxidized,state(Vcell=+1:1V)[41,78]. Figure5.3:Monomerstructures:a)3,4-(2,2-dimethylpropylenedioxy)-thiophene,ProDOT-Me2,andb)3,6-bis(2-(3,4-ethylenedioxy)thienyl)-N-methylcarbazole,BEDOT-NMeCz. Theworkingelectrodepolymerusedinthedeviceshownabove,ispoly[3,4-(2,2-dimethyl-propylenedioxy)-thiophene]orPProDOT-Me2,andthecounterelectrodepoly-mer,usedforchargeandcolorationbalance,ispoly[3,6-bis(2-(3,4-ethylenedioxy)thienyl)-N-methylcarbazole]orPBEDOT-NMeCz.Bothpolymershavebeenelectrochemicallydepositedfromsolution(10mMmonomer,ProDOT-Me2orBEDOT-NMeCz,in0.1M

PAGE 118

Inspectralregionswherethepolymerhaslowabsorption(infraredregionwhenthepolymerisinitsneutralstateandeitherfarinfrared,orvisibleregionwhenthepolymerisinitsp-dopedstate),thelightpassesthroughthepolymer,reectsfromthegoldsurfaceandpassesagainthroughthepolymerlm,theelectrolytegellayer,andtheopticalwindowandtravelstowardsthedetectorwhereitwillbecollected.Thus,thelightpassestwicethroughtheupperactivepolymerlayer.Forthiscase,iftherewerenoabsorptionsfromtheupperthreelayers,opticalwindow,electrolytegel,andpolymerlm,wewouldhavehadthesamereectivityasgold(99%fromfarinfraredtomid-visible,andaftertheplasmaabsorptionedge,at18,500cm1or540nm,thereectivitywouldhavedroppedat40%[41]).However,thereisalwayssomeabsorptionandhence,thereectancespectraaresimilartothetransmittancespectra,ofatransmissivedevicewithapolymerlmtwiceasthickastheoneusedinthereectivedevice. Inspectralregionswheretheabsorptionislarge(visiblewhenthepolymerisintheneutralstateandmid-infraredwhenthepolymerisinthep-dopedstate)mostofthelightentersthepolymerlayer,whereitisabsorbedeitheronthewaytowardsthegoldsurface(WE),oronthewayout,afterithasbeenreectedfromthegold.Although

PAGE 119

Asoneshouldhavealreadynoticed,theopticsoftheelectrochromiccellisquitecomplicated.EventhoughweprobetheopticalpropertiesoftheECmaterialinre-ectancemode,wedonotmeasuretherealreectanceofthepolymer.Themeasuredreectancecontainscontributionsfromallfourupperlayersofthecell.Inorderfortheopticalpropertiesofthepolymerlmtobeextracted,theDrude-Lorentzmodelformulti-layeredsystemsneedstobeusedtotthereectancedata.Eachofthefourlayersisttedseparately,usingaparameterlethatcontainsinformationaboutthecenterfrequency,plasmafrequency,andoscillatorstrengthofeachoscillator,aswellasthedielectricfunctionatinnity,andthicknessofthettedlm.Fromtheparametersobtained,wecomputeopticalconstantswhichyieldinformationabouttheelectronicstructureoftheneutralanddopedstatesofthepolymerlm.Fortheinterestedreader,thispartoftheanalysisandtherelevantparameterlescanbefoundinreference[25].5.3ResultsandDiscussion Inordertostudytheelectrochromiceectoveralargespectralrange,twodier-entwindowshadtobeused.ForthereectancemeasurementsshowninFigure5.4,twoelectrochromiccellswereconstructed,onewithazincselenide(ZnSe)window2mmthickformeasurementsinthemid-infrared(MIR)regionandonewithaglasswindow

PAGE 121

Forthesewiderangemeasurements,twodierentinstrumentshadtobeemployed;aBrucker113vwasusedformid-infrared(MIR)regionmeasurements,andaZeissMPM800microscopephotometerfornearinfrared(NIR)andvisible(VIS)regionmeasure-ments.InFigure5.4thereisadierencebetweenthedatatakenforthemid-infraredregionandthedatatakenforthenearinfraredregion.Thismismatchisduetothedierentthicknessesoftheupperactivepolymers,aswellasthedierentthicknessesoftheopticalwindowsbetweenthetwocells,thatwereusedfortheseparticularmea-surements.However,theabovedierencescanbetakenintoaccountandthedatacanbecorrectedaccordingly.Inthedatapresentedhere,theparasiticreectivityofthewindowshasbeensubtracted.Forcommercialapplicationsinorderfortheparasiticreectivitytobeminimized,thewindowscanbecoatedwithananti-reectivematerial,thiswasnotdonefortheopticalwindowsusedinthesemeasurements. InFigure5.4theelectronictransitionscharacteristicoftheneutral,intermediatedoped,andheavilydopedstatescanbeclearlyobserved.Forthesemeasurements,positivevoltagerepresentstheneutralpolymer,andnegativevoltagerepresentsthep-dopedpolymer.Thereectanceshowslargeelectrochromiccontrast,%R,betweentheneutral(Vcell=1:5V,purpleline),andthefullydoped(Vcell=+1:0V,blueline)statesupto5m.Atlongerwavelengths,above5m,therearestrongabsorptionsduetotheelectrolytegel[77].Thehighestreectancecontrast,%R,isdetectedat1.8m,where%Risgreaterthan90%.Atlongerwavelengths,between3.5-5.0m,thecontrastisreducedtoavalueofaround60%,andatshorterwavelengths,around0.6m,%Ris55%.Attherange3.5-5.0m,thereectanceoftheneutralpolymerdecreasesandthatofthefullyp-dopedpolymerincreases,thusresultinginlowercontrastvalue,comparedto%Rat1.8m.Atthevisiblerange,thesituationisreversed;thepolymerinitsp-dopedstateexhibitshighreectance,andinitsneutralstateexhibitslowreectance,butthecontrastisstilllessthanthemaximumvaluereachedwithinthenearinfraredregion.

PAGE 122

Inmoredetail,thepolymerinitsneutral,insulating,state(Vcell=1:5V)isstronglyabsorbinginthevisibleregion,wherethe-transitionis(0.4-0.65m).Therefore,thereectanceofthedeviceinthisregionisverylow.Atwavelengthslongerthan0.9m,thepolymerlmbecomesquitetransparentandthus,thegoldlayerun-derneathdominatesthereectanceandhence,thedevicereectivityisveryhigh.Atevenlongerwavelengths,longerthan5.5m,thereectanceofthedeviceisdiminishedduetostrongvibrations[77].Whentheupperpolymerisatitsfullyp-dopedstate,(Vcell=+1:0V),thevisibleabsorptionisabsent,whileastronginfraredabsorptionbandduetobipolarons,thathidestheunderlyinggoldelectrode,appears.Attheinter-mediatedopedstates(Vcell=0:0V,-0.5V,and-1.0V),thevisibleabsorptiondecreasesasthedopinglevelincreases,whileatthesametimestrongabsorptionbandsappearatlongerwavelengthsduetopolaronicstates.Ingeneral,attheseintermediatedopedstates,thereectanceoftheECcellliesbetweenthereectanceofthetwoextremastates;fullyoxidized(p-doped)andfullyreduced(neutral)states.Thisistrueforthewholespectrumrangethatwasmeasuredexceptfromtheregionbetween0.4mand1.1m,wheretwooftheintermediatestateshavelowerreectance,strongerabsorptionbands,thananyoftheextremastatesshowninFigure5.4. Inconclusion,theneutral(insulating)polymerhasits-transitioninthevisibleregion,anditistransparentintheinfraredregion,exceptfromvibrationalabsorptions.Atthelightlydopedstates,twosymmetricsub-gapstatesareintroducedupondopingandthus,absorptionbandsappearinthespectrumatlowerenergiesduetothepresenceofpolaronstates;the-transitionisstillpresentbutdiminished.Atthefullydopedstate,asinglebroadbipolaronabsorptionbandisproduced,whereasthepolaronand-transitionsareabsent. Othercharacteristicsthantheelectrochromiccontrastofthesetypeofelectrochromiccellsthatareimportantforapplications,suchasdisplaysandthermalcontrol,aretheswitchingtime,andthelifetime,orlongtermstability,ofthecell.Theswitchingtimeisthetimethatisneededfortheactivepolymertoswitchbetweenthetwoextrema

PAGE 123

Thesecondcharacteristic,thelifetime,orlongtermstability,ofthesedevicesisthenumberofswitchingcyclesacellcanundergobeforedegradationresultsinsignicantlossoftheelectrochromiccontrast%R.Inorderforthelifetimetimeofthecelltobedetermined,thesamesetupasfortheswitchingtimeisbeingused;doublepotentialstepsareapplied,andthereectanceismonitoredataxedwavelengthmaxwherethecontrast%Rhasahighvalue.ForthesePProDOT-Me2baseddeviceswithLiClO4basedelectrolytegel,thedegradationsetsinafterabout1500cycles.Thereectivityofthepolymerinitsp-dopedstateisrelativelyconstantbutthereectivityoftheneutralformofthepolymergraduallydecreases.PProDOT-Me2baseddevicesthatusedalithiumbis(triuoromethyl-sulfonyl)imide,orLi[N(CF3SO2)2],basedelectrolytegelsurvived10,000cycleswithreectancecontrast%R80%measuredat1.3m(orequivalently7692cm1)[77].5.4EnhancingthePerformanceofECDs

PAGE 124

Inaddition,thepresenceofwaterintheelectrochromiccellcancauseover-oxidationoftheoxygen-sensitivepolymerlm,whichisanirreversibleprocessandthus,reducesignicantlythelifetime,orlongtermstability,ofthedevice.Theseab-sorptionbands,dipsinthereectancespectrainFigure5.4areattributedmainlytotheelectrolytegel,sincethethicknessoftheupperactivepolymerisonly150-200nmandthus,thecontributionofthepolymerlmtotheintensityoftheC-Hmodeisverysmall.Ourworkfocusesonthereduction,orremoval,oftheabsorptionsduetotheabovevibrationalstretchesandtherefore,theenhancingoftheperformanceoftheelec-trochromiccell.Therststeptowardsthisdirectionwastocarryoutthepreparationoftheelectrodes,theelectrolytegel,andtheassemblyofthedeviceinadryenvironment.Thepurposeofthisprecautionistoavoidanyexposureofthematerialstotheambienthumidityandthereforetoexcludethewaterfromthecell. Thefourchemicalcomponentsusedfortheelectrolytearebroughtinthedrybox,andtheelectrolytegel(ACN:PC:PMMA:Li[N(CF3SO2)2]withweightratioof70:20:7:3)ispreparedinaninert(Ar)environment.Indetail,Li[N(CF3SO2)2]pur-chasedfrom3Mwasdriedinavacuumovenfor24hat60C,whilethePCpurchasedfromAldrichinSureSealRwaspercolatedthroughtype3Amolecularsievesfollowedbyvacuumdistillation(10mm).Moreover,theACNpurchasedfromAldrichinSureSealRisdistilledatatmosphericpressureunderArovercalciumhydride(CaH2).Fi-nally,activatedmolecularsieveswereaddedtothesolventsPCandACNtoremoveanyresidualwater,andthentheywerestoredinargonatmosphere. Themixingprocedureoftheabovechemicalcomponentsisthefollowing:theacetonitrile(ACN)isputinaglasscontaineronaheatedplateandstirredvigorously.Afterabouttenminutes,thesaltLi[N(CF3SO2)2]isadded.Upondissolutionofthesalt,

PAGE 125

Inordertotestiftheprecautionsthatweretakenfortheassemblingoftheelec-trochromiccellweresucienttoeliminatethewatersignature(O-Hstretchingmode)fromthereectancespectraoftheECdevice,thetransmittanceoftheanhydrouselec-trolytegelismeasured.Thecongurationthatwasusedisthefollowing:twoZnSewindowsfortransmittancemeasurementsinthemid-infraredregion,wherethewaterabsorptionlies,sandwichedtogether,andinbetween,alayerofelectrolytegelthethick-nessofwhichiscontrolledbya25mspacer.Thecomponentswereputtogetherinsidethedrybox,andthetransmittanceofthecongurationwasmeasured.Then,theZnSewindowsweretakenapart,theelectrolytegelwasexposedintheambientenvironmentforoneminute,andthenthetransmittanceoftheabovecongurationwasmeasuredagain.Thetransmittancedataoftheanhydrouselectrolytegel,andtheelectrolytegelafteroneminuteexposuretoambientenvironmentareshowninFigure5.5. AscanbeclearlyseeninFigure5.5,theabsorbingO-Hmodeinthe\waterprotected"electrolytegel(solidline)isalmostnon-existing,whilethesameabsorptionmodebecomesveryprominentevenafteronlyaveryshorttime,oneminute,ofexposureintheambientenvironment(dottedline).Theseresultsdemonstratethatourprocesssuccessfullypreventsmoisturefromenteringtheelectrolytegelandtherefore,eliminatesthewatersignaturefromthereectancespectraofthecell.

PAGE 126

Figure5.5:TransmittancedataofZnSe/GelElectrolyte/ZnSeconguration;a)gelelec-trolyteexposedinairforaperiodof1minute,andb)gelelectrolytepreparedinargon,(Ar),environment. However,theremovalofthesecondabsorptionband,theC-Hstretchingmode,wasnotquiteassuccessful.Althoughitcouldbeachievedthroughthesubstitutionofsomeofthecomponentstheelectrolytewasmadeof,withcomponentsthathave,forexample,C-Fbondstheabsorptionbandofwhichis8-9m,insteadofC-Hbonds.However,weusedtheelectrolytedescribedaboveandworkedtowardsthereductionoftheintensityoftheunwantedbandthroughcarefulcontrolofthethicknessoftheelectrolytegellayerbetweenthetransparentwindowandtheupperactivepolymerlm.Forthepurposeofcontrollingthethicknessoftheelectrolytegellayer,anewsampleholderwasdesigned.Thissampleholder,whichisshowninFigure5.6,tsperfectlyinsidetheshroudofourreectancestage,andmeetsalltherequirementsforreectancemeasurementsbyhavingvedegreesoffreedom;tworotationalandthreetranslational,inordertoachieveagoodalignmentformaximumdetectorsignal.Thenewsample

PAGE 127

Figure5.6:Reectancesampleholderformeasurementsoftheelectrochromiccell,shownintheside,front,andtopviews.Thisdesignprovidescontrolofthethicknessoftheupperelectrolytegellayer[81].) Usingtheabovetechnique,anelectrochromicdevicehadbeenassembledthatconsistedofa2mmZnSewindow,a125mMylarspacer,a200nmPProDOT-Me2lmonslittedgold/Mylar,astheworkingelectrode,aporousseparatorsoakedinelectrolytegel,andaPBEDOT-NMeCzlmongold/Mylarasthecounterelectrode.Thedevicewasmountedonthenewsampleholder,inwhichthethicknessoftheelectrolytegel

PAGE 128

Figure5.7:ReectancemeasurementsofPProDOT-Me2/PBEDOT-NMeCzelec-trochromicdevicefromvisibletomid-IRregionsoftheelectromagneticspectrum,atvariouspotentials. Onethingthatshouldbementionedatthispointisthatalthoughtheoreticallythefullyp-dopedstate(Vcell=+1:0V)isexpectedtobelessreectivethantheintermediate

PAGE 129

Theuseofthenewsampleholderenablesustocontrolthethicknessoftheelec-trolytegel,whichresultstothereductionoftheC-Habsorptionbandinthereectancespectra,butithasamajordisadvantage:duetotheappliedpressuretheslitsontheworkingelectrodearesqueezedtogetherandthus,thediusionoftheelectrolytegelthroughtheslitsisdicult.Asaconsequencetheiontransportisveryslow.TheswitchingtimeforanECdevice,mountedinthenewsampleholderincreasedbyafac-torof103,comparedtotheswitchingtimeofthesamedeviceundernopressure.Hence,furtherstudyisrequiredtomodifytheexistingsampleholderinordertoavoidthisproblem,ortondadierentwaytocontrolthethicknessoftheelectrolytegel.More-over,theoptionofusingdierentchemicalcomponentsforthecreationofanelectrolytegel,whichwilllackC-Hbondsshouldalsobeexplored.5.5Conclusions

PAGE 130

AdditionalworkneedstobedonetowardstheeliminationoftheC-Hstretchingmodethatappearsinthemid-infraredregion.Usingthecurrentchemicalcomponentsfortheelectrolytegel,weneedtoeithermodifythenewsampleholderinordertoavoidthelongswitchingtimescausedbythelargeappliedpressure,ortonddierentwaystocontroltheuppergellayerthickness.Anotherpossibilitythatcouldleadtoaresolutionofthisproblem,istheuseofdierentchemicalcomponentsfortheelectrolytegelthatlackC-Hbondsbuthaveinstead,forexample,C-Fbondstheabsorptionbandofwhichisfarfromtheregionofinterest. Inconclusion,basedonreectancemeasurementsandextensivestudythatleadtodeepunderstandingoftheopticalandelectrochemicalpropertiesofconjugatedpolymers,wewereabletoconstructelectrochromicdeviceswithenhancedperformance.Vari-ablereectiveelectrochromicdevicesbasedonPProDOT-Me2andPBEDOT-NMeCz,wereinitiallyoptimizedtoexhibitcontrastratiosof60-70%inthemid-infraredregionwheretwostrongabsorptionpeaks,waterabsorption(O-Hstretchingmode),andC-Hstretchingmodewerehinderingtheperformanceofthedevices.Anoveltechniquewasdevelopedinordertobuildwater-freedevices,andanewsampleholderwasdesigned,thatallowsforcontroloftheelectrolytegelthicknessandthus,signicantlyreducestheC-Hsignature. Aneorttofurtheroptimizethistypeofdevicesledtothereplacementoftheslittedgold-coatedMylarbyametallizedporousmembrane(polycarbonate,10mporesize).Thisreplacementprovidedmorehomogeneousiondiusion,fasterswitchingtimes(oftheorderofsub-seconds)betweentheextremastatesofthepolymerlm,andlongerlifetimestability.Thisnewgenerationofreectiveelectrochromicdevicesprovedtobebetterlightmodulatorsthanthepreviousoneusingtheslittedgold-coatedMylar.However,severaldicultieswereencounteredwiththistypeofdevicestoo.MoredetailscanbefoundinAppendixA.

PAGE 131

Duringthepasttwodecades,variousconjugatedpolymershavebeensynthesizedwithuniqueoptical,electrical,andmagneticpropertiesduetotheirextended-electrondelocalizationalongtheirbackbone.Carbonnanotubesareconjugated,onedimensional,all-carbonsystemsthatalsopossessremarkableelectronicandmechanicalproperties.Theyconsistofone,ormore,graphenesheetsrolled-uptoformaseamlessnanometer-diametercylinder,orconcentriccylindersifmorethanone,cappedattheendswithhalf-fullerenemolecules.Animportantcharacteristicofthissystemisthefactthatcarbonnanotubescanbeeithermetallicorsemiconductingdepending,solelyforthecaseofisolatedtubes,ontheirgeometry,i.e.,diameterandchirality.TheirdiscoverybyS.Iijimain1991[82]stimulatedtheinterestofmanyresearcherswhoenvisionednumerouspossibleapplicationsfortheseexoticmaterials.Carbonnanotubeshavebeenstudiedintensivelyeversince,andtheyareconstantlyattractingmoreattention.6.1CarbonNanotubes

PAGE 132

Althoughcarbonnanotubessharemanystructuralcharacteristicswiththesinglegraphenesheet,whichisasemiconductorwithazeroenergybandgap,carbonnanotubeswithnodopingimpuritiespresentcanbeeithermetallicorsemiconductingdependingsensitivelyonthetubediameterandwrappingangle.Therefore,smalldierencesinparameterscanprovidesimilarlyshapedtubesconsistingofonlyoneelement,carbon,withverydierentelectronicproperties. Thecarbonnanotubestructureisdenedintermsofthetubediameterdt,andthewrapping,orchiral,angleasshowninFigure6.1.ThechiralvectorChdeterminestherollingdirectionofthegraphenesheetbyconnectingtwocrystallographicallyequivalentsitesOandAonatwo-dimensional(2D)graphenesheet,whereacarbonatomislocatedateachvertexofthehexagonalstructure.Thisvectorisexpressedasfollows intermsofthebasis,orunit,vectorsofthehoneycomblattice1,2,andintermsofapairofintegers(n;m),whichrepresentsChcoordinatesinthelattice. InthegraphenesheetshowninFigure6.1,OBB'Adenestheunitcellofa1Dnanotube.!OBistheshortestrepeatdistancealongthecarbonnanotubeaxis,andthereforedenesthetranslationvectorT,whichcanalsobeexpressedintermsoftheunitvectors1,2,andanewpairofintegers(t1;t2)asfollows ThetranslationvectorTisperpendiculartothechiralvectorCh.Therefore,usingtheconditionfororthogonality,ChT=0,arelationshipbetweenthecoecients(t1;t2)and(n;m)canbederived,andtheresultisthefollowing dR(6.3)

PAGE 133

dR;(6.4) wheredRisthegreatestcommondivisorof(2m+n)and(m+2n). Themagnitudesofthechiralandtranslationvectorsare dR;(6.6) where

PAGE 134

isthelatticeconstantofthetwo-dimensionalgraphenesheet,andListhelengthofthechiralvector.Inotherwords,Listhecircumferenceofthenanotube,sinceCh=dt,wheredtisthediameterofthenanotubeuniquelydeterminedbythe(n;m): withCCbeingthenearest-neighborcarbon-carbondistance(CC==p Furthermore,thechiralangle,whichcanbeexpressedas: jChjj1j=2n+m istheanglebetweenthechiralvectorChandtheunitvector1. Theareadelineatedbythetranslationandchiralvectorsdenestheunitcellareaofthenanotube.ThenumberofhexagonsNcontainedinthisunitcellareacanbedeterminedbytheintegercoecientsofthechiralvector(n;m)asfollows j12j=2(n2+m2+nm) Everyhexagonhasonecarbonatomineachvertexthatisbeingsharedwithtwoneighborhexagons,thereforeonlytwocarbonatomsbelongtoeachhexagon,andhencethereare2Ncarbonatomsperunitcell.Forexample,foranarmchairinwhichn=m=10,usingequations(6.8)and(6.10)weobtainedN=20,andthusthereare20carbonatomsperunitcell.6.1.1ElectronicStructureofCarbonNanotubes

PAGE 135

Graphiteissemi-metallic,whereasasinglegraphenesheetisazerogapsemi-conductorduetothedegeneracyofthebondingandanti-bondingenergybandsatthecornersoftherstBrillouinzone(BZ).ThesecornerscorrespondtoKpoints.Therefore,metallicconductionoccursfortheSWNTsonlywhenoneoftheallowedwavevectorskgothroughtheseKpointsinthetwodimensionalBZ.Atthesepoints,thevalenceandconductionbandsaredegenerateduethethespecialsymmetryofthegraphenelattice[84].Inanyothercaseof(n;m),theSWNTsaresemiconducting. ThesepredictionsonwhetheraSWNTismetallicorsemiconductordonotincludetheeectofthecurvatureofacarbonnanotube.Tight-bindingcalculationswhichhavetakenintoaccountthecurvatureeect,haveshownthatitleadstoanorbitalhybridizationwhichresultsinasmallbandgap,oftheorderofmeV,forthechiral

PAGE 136

Thetwo-dimensionalenergydispersionrelationforbandsofthegraphitelayer,basedonthetightbindingmodelinwhichonlythenearestneighborinteractionsareconsidered,canbewrittenasfollows

PAGE 137

wheretisthenearestneighboroverlapenergy,alsoknownastransferintegral.Inthecaseofnanotubes,thereissomemixingofthe(2pz)and(2sand2px;y)carbonorbitals,thatoccursduetotheinducedcurvature,butsinceitisverysmallitcanbeneglectedaroundtheFermilevel[86]andthus,onlytheorbitalsareconsidered.Byimposingperiodicboundaryconditionalongthenanotubecircumference: whereChisthechiralvector,Kisthereciprocalvector,andqisaninteger(q=1;2;:::;NwithNbeingthenumberofhexagonsinthetranslationalunitcell),andbyusingarotationtransformation: where cos=p 2n+m 2nm theenergydispersionrelationfora(n;m)carbonnanotubecanbeexpressedasafunc-tionofthewavevectorkk,describingtheelectronmotionparalleltothenanotubeaxis. Usingthismethod,wecanplottheenergydispersionrelationandthedensityofstatesfortwospecialcases.First,foratransformationangleofzero=0,theSWNTsareofarmchairtype(n;n)withanenergydispersionrelationshowninFigure6.3andgivenby:

PAGE 138

ncosk where withq=1;2;:::;N,and Thesecondspecialcaseisforatransformationangleof=90,forwhichtheSWNTsareofzigzagtype(n;0)withanenergydispersionrelationshowninFigure6.4andgivenby: n+4cos2q n(6.20) where n(6.21)

PAGE 139

withq=1;2;:::;N,and AscanbeseeninFigure6.3,the1Delectronicdensityofstatesforthecaseofmetallicnanotubeshasanon-zerovalueattheFermilevelEF,andthisnon-vanishingdensityofstatesisindependentofenergy,untiltheenergyisequaltotheenergydier-encebetweenthesub-bandedgesofthevalenceandconductionbands.Thesesub-bandedgesorsingularities,whicharecharacteristicof1Dsystems,correspondtoextremaoftheenergydispersionrelation.Incomparison,forthecaseofagraphenesheet,the2DdensityofstatesvanishesattheFermilevel,andvarieslinearlywithenergymovingawayfromEF(Figureforthe2Ddensityofstatesforgraphenesheetcanbefoundinliterature,forexampleinreferences[84,86]). Furthermore,ascanbeseeninFigure6.4,thedensityofstatesofsemiconductingnanotubesiszerowithintheenergybandgapEg,whichisequaltotheenergydierencebetweenthersttwovanHovesingularitiesinthe1Ddensityofstates.Thesesingu-larities,alsocorrespondtoextremaoftheenergydispersionrelation.Whenthephotonenergyisequaltotheenergyseparationbetweenanoccupiedandanunoccupiedpeak

PAGE 140

Typically,thecarbonnanotubesinallthesetechniquesareformedonthesurfaceofcatalysts,e.g.,Co,Ni,Fe,orY.Theproducedoras-preparedcarbonnanotubesarenotpureandhence,furtherpuricationisneeded.Theimpuritiesthataregenerallyfoundintheas-preparednanotubesaremetalcatalystparticlesandamorphouscar-bon.Duetothefactthathighpuritycarbonnanotubesarerequiredforapplications,severalpuricationtechniqueshavebeendeveloped.Somemajorcategoriesofthesetechniquesareoxidation,ltration,ultra-sonication,andchromatography.Thesepuri-cationmethodsattempttoremovethemetalcatalystparticlesandunwantedamorphouscarbon,withoutaectingthecarbonnanotubes.Moredetaileddescriptionconcerningthedierenttechniquesofcarbonnanotubegrowthandpuricationcanbefoundintheliterature[92,93].

PAGE 141

CNsoeracandidatetoreplacesiliconinelectronicdevices.Silicondevicesaremovingtowardstheirscalinglimitsastheycontinuouslyshrink,inaneorttoachievehigherspeedandhigherdensities.CNsprovideanalternativesolutiontothislimitationproblem.Theiruniqueelectronicandmechanicalpropertiesthoughmakethemgoodcandidatesforanumberofotherdevicestoo,e.g.,CNlogicgates(nanotube-basedeldeecttransistorsCNFET),energystoragedevices(hydrogenstorage),electrochemicaldevices(supercapacitors),nanoprobes(AFM,STM,SFMprobetips),nanotweezers(ac-tuatorstograbandmanipulatesub-micronclustersandnanowires),sensors(chemicalsensors,biosensors,infraredsensors),nanothermometers,andnanotubecompositema-terials(incorporationofnanotubesintoplastics,toobtainultrahighstrengthmaterials). Althoughthispictureprovidesaveryoptimisticviewforthefutureofthesema-terials,therearestillenormouschallengesthatneedtobedealtwithbeforecarbonnanotubesaresuccessfullyincorporatedintousefuldevicetechnology.Someoftheis-suesthatneedtobeaddressedareinitiallythesmallscalegrowthtechniquesthatarecurrentlyavailableandthemanufacturingcost,sincehighlypuriedCNsareveryex-pensive.Moreover,CNlmsareproducedasamixtureofmetallicandsemiconducting

PAGE 142

Nevertheless,inthelastfewyearstherehavebeenanexponentialincreaseinpub-licationsandpatents,thatindicatesthatnotonlytheacademicbutalsotheindustrialinterestincarbonnanotubesisgrowingfast[83].Independentofwhethercarbonnano-tubeswillbesuccessfullyincorporatedintocommercialproductiontechnologyornot,thesesystemshavepavedthewayfortheadvancesandthefutureofnanotechnology.6.2Free-StandingSWNTFilms: Transparency,andOpticalConductivity

PAGE 143

Forthetransmittancemeasurementsdescribedinthischapter,theSWNTlmsweretransferredtoasubstratewithahole,overwhichthelmwaslaidbeforethemembranedissolution.Thisallowedustoperformtransmittancemeasurementsoverawidepartoftheelectromagneticspectrum,fromfarinfraredtovisibleregion.Thesubstratewaseitherapyrolyticgraphitetemplatewithapproximatelya5mmholewhichwasthenmountedonametalframewithasmallersizeaperture,orthemetalframeitself.Thegraphitetemplatewasusedfortemperaturedependentstudiesduetothefactthatthethermalpropertiesofthissubstrateareclosertothethermalpropertiesofcarbonnanotubes,andthusthestretchingofthenanotubelmduringthechangeofthetemperatureisminimized. Themetalframe,showninFigure6.5bottom,isacopperplatewhichispartofourtransmittancesampleholderconguration.InFigure6.5top,atopviewofthissampleholderisalsoshown,inwhichtwoidenticalcopperplates(thesamesizeaperture)aremountedina90anglerelevanttoeachother.Thesecondcopperplatedoesnothaveanythingovertheholeandisusedforthe\background"measurements.Aftertheplatesaremountedinthesampleholder,theholderisattachedtoaliquidheliumcryostatwhichwillpermitustoperformtemperaturedependenttransmittancemeasurements.Theshiftduetothermalcontractionofthecryostatcoldngeriscompensatedbyaremote-operatedz-stage,whichcanrepositionthesampletoitsoriginalplacewithgreataccuracy.

PAGE 144

Figure6.6showsanatomicforcemicroscopy(AFM)imageofa,150nmthick,transparentSWNTlmsurfacepreparedwiththemethoddescribedabove.ThelmthicknessisobtainedbyAFMatthelmedge.Ascanbeseen,thenanotubesarerandomlydistributedoverthespace,andtheyformintobundlesduetovanderWaalsforces.Itshouldbenoted,thatthemorphologyoftheselmsishighlyuniformandtheypossessnanoscaleporosity.Thesucientnanotube-nanotubeoverlapensuresthattheselmshavegoodelectricalconductivity. Thedensityofthismaterialdeterminedbyweighingalmofknownthicknessis0.4g/cm3.Thisvalueisverylow,muchlowerthanthegraphitedensity,andasaconsequencetheselmstendtooatonthesurfaceofwaterandorganicliquids[F.Borondicsetal.unpublishedresults2005].

PAGE 145

Freecarriersareintrinsicinmetallicnanotubes,andcanbeintroducedbydopingwithelectrons(alkalimetals)orholes(halogensoroxidizingagents)insemiconductingtubes.Someholedopingiscommonlythecaseforpuriedsamples.Puricationmeth-ods,whichservetoremovethemetalcatalystparticlesandtheunwantedamorphouscarbonleavethecarbonnanotubeshole-doped.Annealingofthepuriedsamplesun-dervacuumrevertsthemtotheirun-dopedstate,andthischangecanbefollowedbyinfraredspectroscopyduetotheaccompanyingchangeintheoscillatorstrengthoftheopticaltransitions[95].

PAGE 146

Althoughtheoreticalmodelshavepredictedmanyinterestingphenomenaforthefarinfraredpartoftheelectromagneticspectrum,littleattentionhasbeenexperimen-tallydevotedtothisregionsofar[95{97].Farinfraredspectroscopyprovidesapowerfultoolforpristinecarbonnanotubes,andcancomplimentuorescencemeasurements[98]becauseitgivesustheopportunitytostudyonlythemetallicandthelowbandgapcarbonnanotubes.Thereareexisttwokindsoftheoreticalpredictionsfortheori-ginofthefarinfraredopticaltransitionduetolowbandgaps:oneisthesecondarybandgapinducedbythecurvatureforsemiconductingnanotubeswithchiralityindexof2n+m=3k,wherekisanon-zerointeger,andtheotherone,atlowerenergies,isapseudogapcausedbythesymmetry-breakingeectofneighborsinabundleofmetallictubes. Inthiswork,transmittancemeasurementshavebeenperformedontransparentfree-standingSWNTlmsofdierentthicknesses.Thefree-standinglmsgiveusthecapabilitytocoverawiderangeoftheelectromagneticspectrum,fromfarinfraredtoultraviolet.Underothercircumstancesseveraldierentsubstrateswouldhavetobeused,andadicultnormalizationprocedurewouldberequiredinorderforthedataofallregionstobemerged.Furthermore,thiscongurationalsosimpliesthetemperaturestudiesandthevacuumannealingprocessofthepuriedsamples,wherethethermalpropertiesofthelmwouldhavetobematchedwiththethermalpropertiesofeachsubstrate. Transmittancespectraofas-prepared(puried)free-standingSWNT,of1.4nmmeandiameter,lmsofdierentthicknesseswillbepresentedfromthefarinfraredthroughtheultravioletregion,atdierenttemperatures.Thesesamplesexhibitin-creasedmetallicabsorptioninthefarinfraredregion,anddecreasedinterbandtran-sitions,inthenearinfraredandvisibleregions,duetonitratedoping.Acidtreat-ment(HNO3)duringthepuricationprocessleavesthesemiconductingnanotubelmshole-doped(p-doped)andtherefore,shiftsthespectralweightfromtherstvanHovetransitionintotheinfraredregion.Transmittancespectraofthesesamplesafterheat

PAGE 147

Thetransmittancemeasurementsofthefree-standingcarbonnanotubelmswereperformedinthespectralrangefrom30-30,000cm1(4meV-4eV).Forthefarinfraredandmid-infraredregionsaBrucker113vinterferometerwasusedattemperaturesbe-tween50Kand300K,whileforthenearinfraredthroughultravioletaPerkin-Elmerdualbeamspectrometer(Lambda900)oramodiedPerkin-Elmer(PE)16Umonochro-matorwereusedatthesametemperatures.Thespectrometershavesucientspectraloverlaptoallowunambiguousmergingofthetransmittancespectrabetweenthedierentpartoftheelectromagneticspectrum. Theanalysisoftheobtaineddatawasbasedonthefactthatthetransmittanceofalmissubjecttothesamecausalityrestrictionsasthereectance.ThismeansthatonecanuseKramers-Kroniganalysisoftransmittanceinordertoestimatethephaseshiftontransmittance,justasitisusedforreectancemeasurements.Oncethephaseshiftisknown,theopticalconstantscanthenbecalculatedbynumericalinversionof Tei=4N where c;(6.24) isthecomplexphasethroughalm,ofthicknessd,and

PAGE 148

isthecomplexrefractiveindexofthelm.AnimportantcaveatthatshouldbemadeatthispointisthatthephasetheradiationgainspassingathicknessdofvacuummustbeaddedtobeforeoneproceedstocalculateN.RoomTemperatureSpectra ThepuricationprocessleavesthesemiconductingSWNTshole-doped,whichleadstofree-carrierabsorptioninthefarinfraredspectralrange.ThedepletionofelectronsfromthevanHovesingularitiesinthevalencebanddecreasestherateoftheinterbandelectronictransitions.Thermalprocessingininertgasdesorbsthedopantstransferringthespectralweightfromthefarinfraredregiontotheinterbandtransitions.Thisresultsinmaximizingthestrengthinabsorptionbands,whichcorrespondtothersttwointerbandtransitionsinthesemiconductingnanotubesS1andS2,andthersttransitioninthemetallicnanotubesM1,asshowninFigure6.7.Furthermore,insomecasesinthetransmissionspectraofthede-dopedSWNTlmsapeakisapparentinthefarinfraredregionat100cm1.Thispeakisprobablyduetothecurvaturein-ducedenergygapintheDOSofthesemiconductingnanotubeswithchiralityindexof2n+m=3k,wherekisanon-zerointeger[95]. ExperimentaldetailsoftheprocessingforthemeasuredsamplesareshowninTable6.1.SamplesAandBaretwosamplesof250nmthicknessfromthesamebatch,andsamplesC,C'of150nmthickness,areidentical,andcomefromadierentbatchthanA,Bsamples.SampleAhasbeenmeasuredas-prepared(puried),unbaked,whereas

PAGE 149

Table6.1:Processingdetailsofthedierentsamplesstudied.SamplesCandC'wereidentical,measuredbeforeandafterthermaltreatment,respectively. SampleThicknessProcessingtimeID(nm)temperature(C)(hours) A250as-prepared(puried)B250heat-treated(1000C)12C150as-prepared(puried)C'150heat-treated(1000C)12 ThetransmittancespectraofsamplesA(blackcurve)andB(redcurve)atroomtemperatureareshowninFigure6.8,wherelogarithmicscaleshavebeenusedforboth

PAGE 150

Inthefarinfraredregion,thetransmittanceisrelatedtotheopticalconductivitybythewellknown\Glover-Tinkham"formula: (1+21d c)2+(22d c)2;(6.26) where1and2representtherealandimaginarypartsoftheopticalconductivityanddisthelmthickness.Therefore,thelowertransmittanceoftheas-prepared(puried)sampleresultsfromahigherlow-frequency(anddc)conductivitycomparedtotheheat-treatedsample.Thisisconsistentwithahighermobilechargedensityinthissample,andhencesupportstheaboveassumption. Athigherfrequencies,inthenearinfraredandvisibleregions,thetransmittancedipsareduetotheinterbandelectronictransitions.ThersttwodipscorrespondtoS1andS2transitionsinthesemiconductingnanotubes,andthethirdtoM1transitioninthemetallicnanotubes,showninFigure6.7.Theseinterbandtransitionsareveryprominentinthetransmittancespectrumoftheheat-treatedsample,whereasintheas-prepared(puried)sampleduetothetransferringofthespectralweighttolowerfrequenciesthedips,althoughstillevidentarediminished.Thebreadthofthesedipsresultsfromthedistributionofthenanotubediameterswithinthelms,andalsofrombundling. Atevenhigherfrequenciesthetransmittanceofbothsamplesisverysimilar.Thedipwhichappearsattheregionabove30,000cm1isattributedtotheplasmonex-citationabsorption,whichextendstomuchhigher,upto50,000cm1,frequenciesandtherefore,noattempthasbeendonetoanalyzethispartofthespectrumsinceourdatadonotextendasfar.

PAGE 152

Animportantcharacteristicinthetransmittancespectraoftheheattreatedsam-ple,showninFigure6.8,isthetransmittancedipbelow100cm1.ThisdipisascribedtoacurvatureinducedenergygapintheDOSofthesemiconductingnanotubeswithchiralityindexof2n+m=3k,wherekisanon-zerointeger.Thishasbeenobservedbyothergroupstoo[95]. Figure6.9showsthetransmittancespectraofthe150nmSWNTlmbeforeandafterheattreatment.Herealso,thetransmittanceandthefrequencyaxesareshownonlogarithmicscales.Themaincharacteristicsofthetransmissionspectraofthissamplearethesameasthosedescribedforthethickerones.AnimportantdierenceisinthespectrumofthesampleaftertheheattreatmentC',andistheabsenceofthedipbelow100cm1,whichisobservedinthethickersample.Thismightbeduetoaccidentaldopingbyatmosphericoxygen,sinceithasbeenreportedthatexposureofcarbonnanotubestoairoroxygencanaltertheirelectronicproperties[99].Evenalowdopingcouldresulttosmearouttheenergybandgapsofsomenumberofnanotubesandconsequentlychangethedistributionofgaps,sothatthemedianshiftstolowerfrequencieswhereitcannotbeobservedfromthesemeasurements. Finally,Figure6.10showsthetransmittancespectraofallfourmeasurementsforcomparison.Ingeneral,theas-prepared(puried)sampleshavelowertransmittanceinthefarinfraredregionduetothefreecarrierabsorption,andhighertransmittanceinthenearinfraredandvisibleregionscomparedwiththesamethicknessheat-treatedsamples.SamplesCandC'are,ofcourse,moretransparentthroughoutthemeasuredrangecomparedtosamplesA,Brespectivelyduetothedierenceinthickness.Ithasbeenreported[90]thataSWNTlmof50nmthicknessexhibitstransmittancegreaterthan70%overthevisibleregionoftheelectromagneticspectrum,higherthan90%inthenearinfraredpart,andisexpectedtoretainthisvalueoverthemid-infraredregiontoo.Thesheetresistanceforthis50nmlmintheas-prepared(puried)statewasmeasuredtobe30/S.Therefore,SWNTscanbeexcellenttransparentelectrodespossessinghightransmittanceforlowsheetresistance.Itshouldalsobenotedthat,

PAGE 153

Thetemperature-dependenceofthetransmittancespectrumofthe250nmSWNTlmaftervacuumannealingisshowninFigure6.13.Inthisspectra,theimportantfeatureistheweakminimuminthefarinfraredregion,whichisinconsistentwitha

PAGE 155

Theeectofnitratedopingofthesemiconductingcarbonnanotubesduringthepuricationprocessisquiteobviousintheopticalproperties.AscanbeseeninFigure6.14,thereisasignicantincreaseintheconductivityinthefarinfraredregionbetweentheas-prepared(puried)andheat-treatedlms.Inthenearinfraredregion,adecrease

PAGE 156

Theconductivityvaluesoftheheat-treatedsamplesatlowfrequenciesare8001cm1and10001cm1forthe250nmand150nmlmsrespectively.Thesevaluesareinagreementwiththevaluesdeterminedfromthesheetresistanceoftheselms[90].TheDCconductivityoftheas-prepared(puried)samplesisaround26001cm1.Thisvalueislowerthanthevaluedeterminedfromthesheetresistanceoftheselms[90],butveryclosetootherexperimentalresultsofhigh-resolutionelectronenergy-lossspectroscopy(EELS)[100]. ThetemperaturedependenceoftheopticalconductivityisshowninFigure6.15.Thedependenceofconductivityfromtemperatureisbasicallyconnedinthefarinfraredregion,wheretheDrudefeaturebecomesnarrow,whileinthemid-infraredrangetheconductivityisdecreasingasthetemperaturedecreasesfrom300Ktolowervalues.Morespecically,theconductivityatlowfrequenciesisdecreasingasthetemperature

PAGE 157

Figure6.15:Temperature-dependenceofopticalconductivityforthe150nmas-prepared(upperpanel)andheat-treated(middlepanel),andthe250nmheat-treated(lowerpanel)free-standingSWNTlm.

PAGE 158

Amulti-oscillatortwasappliedtothetransmittancedataforthinfree-standinglms.Thetransmittancewastusingthefollowing,Drude-Lorentz,model: ~(!)=1+Xj!2pj whichyieldsestimatesofthehighfrequencydielectricconstant1,theDrudeplasmafrequency!pD,theDrudelifetimeD,aswellastheplasmafrequencyoroscillatorstrength!pj,centerfrequency!j,anddampingjforeachmode.ThettingparametersthatwereobtainedfromthismodelforallsamplesareshowninTables6.2and6.3. Table6.2:Drude-Lorentzttingparametersfortheopticalconductivityofeachsamplestudiedatseveraltemperatures.Theparametersforhigh-frequencyoscillatorsaregiveninTable6.3. 300K20577624054351283375618414065179106173859B:de-doped 50K18482424325557152465459331287455310,4632899100K19592591325456157465559291299453610,4672881200K19152120324465164463259341264456710,4572922300K19732155323066186466059301298454510,4672890C:mostdoped 100K2426385244250135247862601801479810,8804207200K2411336344080140250062531825479010,8734205300K2368285843650146250262541792479110,8574203C':de-doped 100K20512223291820120411359641190503310,4773502200K22862373293120128411459481177508010,4443552300K263525612936201394269596512345078104603520 Thepeakoftheopticalconductivityatlowfrequencies(100cm1or10meV)inthethickerheat-treatedsamplecannotbewellttedbyaDrudefreecarrierresponsealone;hence,aLorentzoscillatorisalsorequiredatthisfrequencyrange.TheDrude

PAGE 159

Table6.3:Drude-Lorentztparametersfortheopticalconductivityofeachsampleforthehigh-frequencyoscillators(where1=1:04). SampleState!p4!44 de-doped33,30035,00053,400 150nm:doped22,00034,00040,000 de-doped39,80047,00064,000 componentoriginatesfromthetrulymetalliccarbonnanotubesinthelms,whiletheLorentzcomponentisattributedtothecurvatureinducedenergygap,orameanvalueofenergygapsofsemi-metallictubesofdierentdiameterandchirality.Forthethinnerheat-treatedsample,thebesttwasaccomplishedbytheuseoftwoDrudecomponentsofdierentdampingconstants.However,ifwehaddatabelowthemeasuredrange,i.e.,30cm1andtheconductivitypeakwasindeedredshifted,thenmaybeoneDrudeandoneLorentzcomponentwouldprovideabettertforthissampletoo.Becausewehavenodataatlowerfrequenciestosupportthisassumption,wekeepthetwoDrudecontributionsforsampleC'.Finally,noevidenceofabundleinducedpseudogapinmetalliccarbonnanotubeswasseeninoursamples(800cm1or100meV),asthiseectwasobservedelsewhere[87]. Theeectofdopingcanalsobeseenifwefollowthechangesinthettingpa-rametersbetweentheas-prepared(puried)andtheheat-treatedsamples.Uponholedopingofthesamplestheintensityofthersttwopeaks,DrudeandLorentzforsam-plesAandBandtwoDrudecomponentsforCandC'areincreasing,whileatthesametimetheintensityofthenexttwopeaks,whichcorrespondtotheS1andS2electronictransitionsforsemiconductingtubesaredecreasing.Thischangeisconsistentwiththetheory,wherearedistributionoftheoscillatorstrengthinthenanotubesisexpectedupondopingofthesemiconductingnanotubesinthelms.Hence,thefreecarrierdensityincreasesandconsequentlythelowfrequencyabsorptionandtheconductivity

PAGE 160

Figure6.16showstheroomtemperatureabsorptioncoecientofthelmsintheas-preparedandheattreatedstates.Atlowfrequencies,theabsorptionoftheas-prepared(puried)samplesisconsiderablyhigherthantheheat-treatedsamplesandverybroad,whileathigherfrequenciesthesituationreverses,asisexpected.TherstvanHovesingularitypeak,whichcorrespondstotheS1transitioninthesemiconductingnanotubesismuchstrongerintheheat-treatedthantheas-prepared(puried)samples.ThedierenceintheabsorptionofthepeaksthatcorrespondtotheS2andM1transi-tionsinthesemiconductingandmetallicnanotubesrespectively,betweenthedopedandun-dopedsamplesissmaller.Athighfrequencies,above20,000cm1,theabsorptiondueto-electronexcitationbeginstoshow.Thisabsorptionextendstomuchhigherfrequenciesthanthemeasuredrangepresentedhere.

PAGE 161

m=2mVcell wheremisthecarriereectivemassandVcellisthevolumeoccupiedbyoneformulaunit.Here,basedontheexperimentaldensityresultswherethedensityofthismaterialwasdeterminedtobe0.4g/cm3[F.Borondicsetal.unpublishedresults2005],thevolumeVcellistakentobe50A. Theresultsforthepartialsumrulefor1foroursamplesareshowninFigures6.17and6.18.Bothgroupsofsamples,as-prepared(puried)andheat-treatedshowthesametrends.Theplateauvaluearound1000cm1providesagoodestimateofthelowenergyspectralweightofthemetalliccarriers.Thelargernumberoffreecarriersintheas-prepared(puried)samplesisquiteobvious.Thenearagreementofthetwocurvesinbothguresathigherdensitiesindicatesthatthedopingreducestheintensityoftheenergybandgaptransitionofthesemiconductingnanotubesandtransfersweighttothefarinfraredregion.Thiseectwasseenaboveasaredistributionoftheoscillatorstrengthinthesemiconductingnanotubes.UponholedopingtheoscillatorstrengthcorrespondingtotheenergybandgaptransitionS1isdecreasingduetothedecreasingelectronconcentrationinthevalenceband,whileatthesamethereisanincreaseofthefreecarrieroscillatorstrength.6.2.5Conclusions

PAGE 162

Figure6.17:Roomtemperaturesumruleforas-prepared(puried)andheat-treatedsamplesof150nmthickness. Figure6.18:Roomtemperaturesumruleforas-prepared(puried)andheat-treatedsamplesof250nmthickness.

PAGE 163

Thevacuumannealedlmsareconsideredtobeclosesttopristinecarbonnano-tubesandthus,provideustheopportunitytostudythetrulymetallicandsmallbandgapsemiconductingcarbonnanotubes.Therefore,thefarinfraredspectroscopycanserveasacomplimentarymethod,forexample,touorescenceexperiments,helpingresearcherstoacquireamorecompleteunderstandingofthisuniquesystem. ByKramers-Kroniganalysisofthetransmittancedata,theopticalconstants,op-ticalconductivityandabsorbance,aswellasthepartialsumrulefortheconductivityofthefree-standingcarbonnanotubeswereobtained.Theresultsofthisanalysisforsamplesofdierentthicknessesandindierentstates,dopedandun-doped,supportthetheoreticalpredictionsverywell.Inaddition,agapfeaturebelow10meVthatwasobservedinoneofthesampleswasattributedtothecurvatureinducedbandgapofsemi-conductingnanotubes.Finally,noevidencewerefoundofabundleinducedpseudogapinmetalliccarbonnanotubesinoursamples,asthiseectistheoreticallypredictedandhasbeenobservedelsewhereat800cm1or100meV[87].

PAGE 164

Atransmissive/absorptiveelectrochromic(EC)deviceexhibitsreversibleswitchingofanECmaterialbetweenatransmissive(bleached)andacolored(absorptive)stateonatransparentconductingsubstrate.Theconstructionofatransmissive/absor-ptivedevicethatoperatesnotonlythroughoutthevisiblebutintheinfraredregionalsohadbeenimpossibleuntilnowduetotheabsenceofasuitable,infraredtransparent,conductingmaterial.Webuiltthersttransmissive/absorptivedeviceusingSWNTlmsastheconductingsurfaces.Theextensivestudyoncarbonnanotubesasfree-standinglmsthatwasdescribedinChapter6showedthattheselmshavethehighesttransmittance(around90%forthicknessesbetween50-350nm)ofanytransparentconductorinthe2-5mspectralrangeforverylowsheetresistance(resistivityof1.5x104-cm[90]).7.1SamplePreparation7.1.1PuricationofLaboratoryChemicalsandMaterials

PAGE 165

Chemicalsandelectrodeswereplacedintheantechamberofthedryboxandpumpedforthreecyclesof45minuteseachwithanexceptionofthelithiumbasedsalt,whichispumpedovernight.

PAGE 166

Indium-Tin-Oxide(ITO)coatedglassslideshavealsobeenusedinsomecases,onlyforcomparisonreasonsinordertoenhanceourunderstandingofthenewelectrodesintheelectrochromicdevices.Anotherelectrodethatwasalsousedconsistsofgold(100nm)coatedkaptonsheetthatiscommerciallyavailable.Inthiselectrodeapunchedholeof3-4mmdiameterwasmadeonit,sothatthelightbeamcouldgothroughthedeviceandthereby,transmittancemeasurementstobeobtained.Theholecanbesealedwithaninfraredtransmissivewindow,whichcanbeplastic(PET)orZincSelenide(ZnSe).7.1.3ElectrochemicalPolymerization

PAGE 167

Electrochemicalpolymerizationiscarriedoutinamonomerelectrolytesolution,initiallycontainingthemonomerEDOT(10mM),inTBAP(0.2M)andPC,atconstantpotential,+1.26VvsFc/Fc+.ACNcanalsobeusedasasolventfortheSWNTs/PETelectrode,butnotforthegoldonkaptonelectrodebecauseinACNthekaptonsupportswellsandcurlsduringpolymerization.Thepolymerlmsdepositedusingthismethodwerewithin150-200nmrange.Thethicknessofthelmswascontrolledbymonitor-ingthechargeddensitypassedduringelectrosynthesis.Afterthepolymergrowth,thepolymer-coatedelectrodeswereremovedfromthemonomersolutionandrinsedwithsolvent,inthiscasePC.Thelmsatthisstagearep-dopedandtheycanbeneutralizedbytheapplicationofapropervoltageinmonomerfreesolution;TBAP(0.2M)andPC. Sametypeofelectrochemicalpolymerizationhastakenplaceforthedepositionofanothermonomer,theBEDOT-Hx2-Pyr-Pyr,thathasalsobeenusedinthetrans-missive/absorptivedevices.ThemonomerelectrolytesolutioncontainsthemonomerBEDOT-Hx2-Pyr-Pyr(10mM),inTBAP(0.1M),ACNandDCM(dichloromethaneorCH2Cl2)withweightratioof1:1,atconstantpotential+1.21VvsFc/Fc+.Althoughmostofconjugatedpolymersarestableonlyintheirp-dopedstate,thisparticularpoly-mer,PBEDOT-Hx2-Pyr-Pyr,hastheabilitytoformastablen-dopedstateinadditiontothep-dopedstatealongthepolymerbackbone.ThisoersusthepossibilityofusingthesamepolymeronbothelectrodesinaECtransmissive/absorptivedevice,inwhichbothlmswillbeintheirneutralstateatthesametimeandupontheapplicationofavoltage,onewillforman-dopedstateandtheotheroneap-dopedstate.Theresulting nFlnoxid

PAGE 168

Figure7.1depictsthechemicalstructuresofthetwopolymersusedintheECdevicesinthischapter,andtheircolorsuponoxidationandreduction. Thepreparationofthegelhastakenplaceinadryenvironmentinordertoyieldawater-freeelectrolyte.Theelectrolytegelthathasbeenusedinourdevicesconsisted

PAGE 169

Themixingprocedureofthesefourchemicalcomponentsisthefollowing.Theacetonitrile(ACN)wasputinaglasscontaineronaheatedplateandstirredvigorously.Afterabouttenminutes,thesaltLi[N(CF3SO2)2]wasadded.Upondissolutionofthesalt,theadditionofpoly(methylmethacrylate)(PMMA)followed.Thelattercompo-nentisnoteasilydissolved,soitwasstirredandheatedat60Contheplateforabouttwohours.DuringthattimetheACNevaporated.AssoonasthePMMAwasdissolved,propylenecarbonate(PC)wasintroducedtothemediumandwhenthemixturehadaviscousconsistency,afterabouttwomorehoursofstirringandheating,itwaseitherstoredunderargon(Ar)environment,orusedinadevice.Thistypeofgelelectrolyteinadditionofbeingtheiontransportmedium,italsoencapsulatestheECdevice.Attheedgesofthedevice,theACNevaporatesleavingbehindthePMMAwhichbecomesinsoluble.Thisprocessminimizesfurthersolventevaporationandpreventsleaking.7.2PEDOTonSWNTElectrodes7.2.1SpectroelectrochemistryofPEDOT

PAGE 170

BeforethecarbonnanotubelmswereusedtobuildECdevices,wewantedtostudytheirsuitabilityaselectrodesforthisspecictypeofdevices.Forthisreason,

PAGE 172

TwolmsofPEDOTwerepreparedelectrochemicallyonthetwodierentsub-strates,andtheirabsorptionwasmonitoredatdierentvoltages.Figures7.2through7.5showthetransmission/absorptionspectrafornearinfraredandvisibleregionsofthePEDOTonITO/glasselectrodeandthePEDOTonthenewcarbonnanotube/plastic(SWNTs/PET)electrodeatavoltagerangefrom-0.7Vto+1.0V. TheseguresshowthatthebehaviorofthecarbonnanotubelmisverysimilartothatofthewidelyusedITOlm.Therefore,thecarbonnanotubelmscanbeusedastheconductivelayerinECdevices.ThedierenceintheabsorptionbetweenthelmonthedierentelectrodescouldbeeitherduetothefactthatthePEDOTlmontheITO/glasselectrodeisalittlebitthickerthanthePEDOTlmonSWNTs/PETelectrode,orduetothedierentvaluesoftheworkfunctionofITOandcarbonnanotubes.Inorderto

PAGE 173

Thedipsinthenearinfraredregion,at1170nmand1370nmareduetoACNabsorption,ascanbeseenfromFigure7.6,andtheywereconsistentinalltheabsorptionmeasurementsperformedinsolution.Thesedips,whichinsomeothermeasurementsappearaspeaksareduetothesystemsdeciencytosubtractthebackgroundabsorbancecorrectly.7.2.2ThicknessDeterminationofPEDOTonSWNTElectrodes Duetotheporousmorphologyofthecarbonnanotubelms(biggersurfacearea),thedepositiontimeforthesameamountofchargewasmuchlessforthemwhencom-paredwiththedepositiontimeonITObecausethedepositedpolymerdoesnotgrowonlyontopoftheelectrodesurfacebutinbetweenthenanotubebundlestooduetothepresenceofvoids.Howevertheoretically,sincethesameamountofchargeispassedduringthepolymergrowthonbothelectrodes,theresultedlmthicknessshouldbethesame.Experimentallythough,sinceasimpleprolometrymeasurementthatwouldverifythisassumptioncouldnotbeperformedduetotheveryroughsurfaceofthenan-otubelmsandtheirporousnature,anothermethodhadtobefound.Therefore,thedeterminationofthelmthicknesswasobtainedthroughspectroscopicmeasurementsand,theanalysisandtofthedatausingtheDrude-Lorentzmodel.

PAGE 174

Forthesemeasurements,threePEDOTlmsofdierentthicknessesweregrownpotentiostaticallyonSWNTs/PETunderthesameconditionsexcepttheamountofthe

PAGE 175

Theresultsofthetransmittancemeasurementsthatwereperformedonthesesam-ples,p-dopedPEDOTlmsofdierentthicknesses,fromthenearinfraredtotheultra-violetregionandtheDrude-LorentzbasedmodelingthatwasperformedareshowninFigure7.8.TheparametersusedtotallthreesamplesaregiveninAppendixBandareallthesameexceptthethicknessparametersthatwereleftfreetoadjust.Theresultsthatweobtainedfromthisstudyledustotheconclusionthatforthesameamountofchargethe\optical"thicknessoftheresultinglmisthesameforbothelectrodes,SWNTsandITO.Therefore,thesamecalibrationplotcouldbeused.7.2.3AtomicForceMicroscopy(AFM)Images Inmoredetail,thelefthandsideimageofFigure7.9showsa50nmSWNTsonPETlm.Thecarbonnanotubes,whichcanbeclearlyseeninthisimage,formintobundlesandarerandomlydistributedoverthelmsurface.Althoughtherearealotofvoids,whichrevealthemicroporousnatureofthecarbonnanotubelmsthereisenoughoverlappingtoensuregoodconductivityoverthewholesurfaceofthelm.Therighthandsideimageinthisgureshowsa200nmPEDOTlmgrownonthe50nmSWNTsonPETsubstrate.Onethingthatshouldbenoticedisthatthecarbonnanotubesarenotvisibleanymoreafterthepolymergrowth,andthereforea200nmpolymerlmthicknesscoversadequatelythevoidsbetweenthenanotubebundles,andtheelectrodesurface.Thepolymersurfaceofthe200nmlmthicknessappearsnottobeverysmoothduetheformationofislands. Figure7.10showsa50nmSWNTsonPETlminadierentscaleincomparisonwithFigure7.9.Thecarbonnanotubelmofthisgurehasthesamecharacteristicsasdescribedabove.Therighthandsideimageinthisgureshowsamuchthicker,500nm,

PAGE 176

Figure7.9:AFMimagesofa50nmSWNTs/PETelectrode(lefthandside)anda200nmPEDOTlmgrownelectrochemicallyona50nmSWNTs/PETelectrode. Figure7.10:AFMimagesofa50nmSWNTs/PETelectrode(lefthandside)anda500nmPEDOTlmgrownelectrochemicallyona50nmSWNTs/PETelectrode(righthandside).Noticethatthescaleisdierentcomparedtothepreviousimage.PEDOTlmgrownonthe50nmSWNTs/PETsubstrate.Aftertheelectrochemicalgrowthofthepolymerlm,thecarbonnanotubesarenotvisibleanymore,andthepolymersurfaceofthe500nmlmthicknessappearstobealotsmoothercomparedtothesurfaceofthe200nmPEDOTlm.Therefore,asthepolymerthicknessincreasesthesurfacebecomessmoother.

PAGE 177

Figure7.11showsthecyclicvoltammogramoftheoxidationandtherstreductionofPBEDOT-Hx2-Pyr-PyronITO/Glassinsolution.PBEDOT-Hx2-Pyr-Pyrisoneofthefewpolymerscapableofformingap-dopedstateaswellasastablen-dopedstate.Then-dopedstateofelectroactivepolymersisusuallynotstableduetothehigh

PAGE 178

Figure7.12showsthecyclicvoltammogramoftheoxidationofPBEDOT-Hx2-Pyr-PyronSWNTs/PETinsolution.Thetwoplotsfollowthesametrends,withthecurrentbeingslightlysmallerforthecarbonnanotubeelectrode.Thisispossiblyduetothedierenceinthethicknessofthepolymerlmsgrownonthetwoelectrodes,since,althoughthesameamountofchargehasbeenusedfortheelectropolymerizationtheareaoftheelectrodeshasnotbeenmatchedandtherefore,isdierent.37.3.2SpectroelectrochemistryofPBEDOT-Hx2-Pyr-Pyr

PAGE 182

Onethingthatshouldbepointedoutisthatduringthespectroscopicmeasure-ments,thelightbeamgoingthroughthedevicepassesonlythroughtheworkingelectrodepolymerandthus,thedataobtainedfromthisdevicesdepicttheelectrochromicchangesofonepolymerlm.Otherwise,incasetheCEelectrodewastransparenttoo,wewouldnotbeabletousePEDOTasalmonbothelectrodessincethispolymercanaccessonlyaneutralandanoxidizedstate.Thismeansthatbychangingtheappliedvoltage,onelmwillbeoxidized(bleached)andtheotherwillbereduced(neutral-colored),andthusinadualdevicenochangeintransmittanceorabsorbancewouldberecorded.TransmittancemeasurementswerealsoobtainedfromadeviceofthistypethatusesPBEDOT-Hx2-Pyr-Pyronbothelectrodeswiththelightbeamgoingthroughonlyoneofthelms.

PAGE 183

ThesecondtypeofdevicesthatweconstructedisdepictedinFigures7.18and7.19.ThisdeviceconsistsoftwoPBEDOT-Hx2-Pyr-Pyr(300nm)coatednanotubeslmsonPETelectrodes,facingeachother.Theelectrodesareseparatedbygelelec-trolyte.Inthiscase,bothelectrodesaretransparent,fromthemid-infraredthroughthevisibleregionandtherefore,noholesarenecessary.ThisparticularpolymerdefersfromPEDOTbecauseitcanaccessthreedierentstates:p-doped,neutral,andn-doped.Therefore,thetwoPBEDOT-Hx2-Pyr-Pyrlmsareinitiallyintheirneutralstatefortheassemblingofthedeviceandupontheapplicationofvoltageonelmisoxidized(p-dopedstate)whiletheotherisreduced(n-dopedstate).Inthiscase,theresultingcolorisacombinationofthecolorsofp-andn-dopedlmsandthespectralchangeswillalsobeacombinationofthoseseenintheseparatep-andn-dopingexperiments.7.5In-SituTransmittanceMeasurementsofECDevices

PAGE 184

Figures7.20through7.23depictthetransmittancedataalongwiththettingresultsfrommid-infraredtothevisibleregion.ThePETsubstrateshowninFigure7.20isttedrstasasinglelayerandthentheelectrolytegelconnedinbetweentwoPETlayersistted,theresultisshowninFigure7.21.Forthettingofthegellayer,thettingparametersofthePETlayeraretakenintoaccount,andthiswayattheendoftheanalysisofalllayers,wehaveobtainedaparameterleforeachlayerthat,ideally,describessolelytheopticalpropertiesofthespeciclayer.Figure7.23showsthetransmittanceandttingresultsofadevice(withthehole),inwhichthepolymerunderstudy,whichisthePEDOTinthiscaseisinitsneutralstate.Thetransmittanceofareferencedevicehasbeensubtractedfromthedataandthetting.

PAGE 187

InFigure7.24canbeseenthatthepolymerinitsneutral,un-doped,state(Vcell=1:8V)isstronglyabsorbinginthevisibleregionwherethetransitionis.There-fore,thetransmittanceofthedeviceinthisregionishighestcomparedtoallotherstates.Atfrequencieslessthan3600cm1,thetransmittanceofthedeviceisdiminishedduetostrongabsorptionsmainlyfromtheelectrolytegelandtherefore,noconclusionscanbedrawnforthispartofthespectrum.Whenthepolymerisinitsfullyp-dopedstate(Vcell=+1:0V)thevisibleabsorptionisreduced.Attheintermediatedopedstatesthevisibleabsorptiondecreasesasthedopinglevelincreases.Ingeneral,attheseinterme-diatedopedstatesthetransmittanceoftheECcellliesbetweenthetransmittanceofthetwoextremastates;fullyoxidized(p-doped)andfullyreduced(neutral)states.Thisholdstrueforthewholespectrumrangethatwasmeasured. Figure7.25showsthetransmittancedataforthePEDOTdevice(withhole)di-videdbythereferencedeviceinthreestates;fullyp-doped,slightlyp-dopedandneutral.InthesamegureisalsoshownthetofthedatabasedontheDrude-Lorentzmodel.Theconclusionsdrawnfromthisspectrumandtheobtainedparameters(AppendixB)aredepictedmoreclearlythroughtheabsorptionspectradescribedinthenextsubsec-tion.OnethingthatcanbenoticedfromthespectrashowninFigure7.25isthebig

PAGE 189

PreliminarytransmittanceresultsofabetteroptimizedPEDOTECdevice(withhole)areshowninFigure7.26.ThespectrainthisgurecorrespondtothetransmittanceoftheactualECdevice,andnosubstractionofareferencedevicehastakenplace.Thisisobviousfromthestrongpeakat6000cm1,whichappearsinthereferencedeviceascanbeseeninFigure7.24.Dataanalysishasnotbeenperformedforthisdevicebutthetransmittanceissubstantiallyhigherthatthepreviousdevicewithvaluesrangingfrom5%toalmost80%,andanelectrochromiccontrast%Tof71%atmax4891cm1.Theseresultsareverypromisingandindicatethatwithamorecarefuloptimizationverygoodperformancecanbeachieved.AbsorptionCoecientofPEDOT

PAGE 190

Tosummarize,theneutral(un-doped)PEDOThasitstransitioninthevisibleregion,anditistransparentintheinfraredregion,exceptfromvibrationalabsorp-tions.Atthelightlyp-dopedstates,twosymmetricsub-gapstatesareintroducedupondopingandthus,absorptionbandsappearinthespectrumatlowerfrequencies/energiesduetothepresenceofpolaronstates.Thetransitionintheseintermediatestatesisstillpresentbutdiminished.Atthefullyp-dopedstate,asinglebroadbipolaronabsorptionbandisproduced,whereasthepolaronandtransitionsareabsent.

PAGE 191

Inconclusion,wewereabletoobserveandfollowthedopinginducedstatesupontheapplicationofavoltageonaPEDOTECtransmissive/absorptivedevice(withhole).TheseresultsagreewithwhatisexpectedfromtheoreticalstudiesofthissystemsaswellasexperimentalresultsfromreectivedevicesliketheonesstudiedinChapter5ofthisdissertation.LongTermSwitchingStabilityofPEDOTECDevice(withhole) Ascanbeseenfromtheabsorbancespectrum,thereissomedegradationintheperformanceofthePEDOTECdevice(withhole)duringtherst200cyclesbutafter

PAGE 193

Figure7.31showsthetransmittancechangesoftheactualdevice(withhole)upontheapplicationofvoltagewithintherangefrom-0.5Vto+0.9V.InthisvoltagerangePBEDOT-Hx2-Pyr-Pyrchangesfromaneutralstatetoanoxidizedstate.Theelec-trochromiccontrast%Tthroughoutthevisibleregionissmall10%,whileitbecomesquitebiggerinthenearinfraredregionwhereat2000nmor5000cm1reachesavalueofalmost45%. Figure7.32showsthetransmittancechangesoftheactualdevice(withhole)upontheapplicationofvoltagewithintherangefrom-0.5Vto-2.6V.InthisvoltagerangePBEDOT-Hx2-Pyr-Pyrchangesfromaneutralstatetoareducedstate.Duringthesemeasurements,afterreachingtherstreductionofthepolymerandobtainedallthe

PAGE 194

Theelectrochromiccontrast%TofthePBEDOT-Hx2-Pyr-PyrECdevice(withhole)betweentheneutralandthereductionstatesisbiggercomparedtothecontrastobtainedbetweentheneutralandoxidizedstatesthroughoutthewholespectralre-gion.Furthermore,inthiscase,themaximumcontrastisobtainedat880nm(11,364cm1)anditreachesavalueofalmost50%.

PAGE 195

Thetransmittanceofthereferencedevicehasnotbesubtractedfromtheaboveresults,whichdepictthetransmittanceoftheactualdeviceitself.Therefore,thepeakat1666nmor6000cm1ispresentinthesedataalso. Figures7.33and7.34showtheabsorbancespectraofthePBEDOT-Hx2-Pyr-PyrECdevice(withhole)atdierentappliedvoltages,fromneutraltooxidizedandneutraltoreducedstatesrespectively.InFigure7.33theabsorptionbandhasnotbeenfullydepletedandthismaybeduetothefactthathigherappliedvoltageswereneeded.Abiggerelectrochromiccontrastisobservedinthenearinfraredregionforthesamevoltagerange.InFigure7.34theabsorptionbandhasbeenfullydepletedwithintheappliedvoltagerangeexhibitinghighelectrochromiccontrastinthisregion,whereasatthenearinfraredregiontheexhibitedcontrastissmall.LongTermSwitchingStabilityofPBEDOT-Hx2-Pyr-PyrECDevice(withhole)

PAGE 196

TheabsorbanceofPBEDOT-Hx2-Pyr-PyrECdevice(withhole)whenthepoly-merwasinitsoxidizedstatewasincreasingduringtherst200cyclesandthenitstabilized,whereaswhenthepolymerwasinitsneutralstatetheabsorbancewasstablethroughoutthewholetimeoftheexperiment.ThisECdevicebetweentheneutralandoxidizedstatesshowadierenceintransmittanceof32%.Itshouldbenotedthatthisisnotanoptimizeddeviceeitherandtherefore,betterperformancecouldcertainlybeachieved.Furtherinvestigationisneededinordertodetermineifthedierencein%TbetweenthePEDOTandPBEDOT-Hx2-Pyr-PyrECdevices(withhole)isdueto

PAGE 198

Weperformedthesemeasurementsinordertogenerallycompareadualreferencedevicewithadualpolymerdevice.ThescanratedependencecyclicvoltammogramsofaPBEDOT-Hx2-Pyr-PyrdualECdeviceisshowninFigure7.38.InFigures7.37and7.38canbeseenthatasthescanrateincreasesthecurrent,ipincreasestoo.InFigure7.39,thecyclicvoltammogramofthedualreferencedevice,forthe25mV/sscan

PAGE 200

Whenthepolymerlmsareintheirneutralstate(Vcell=0:0V)thedeviceisstronglyabsorbingattheboundarybetweenthevisibleandthenearinfraredregionwherethetransitionofPBEDOT-Hx2-Pyr-Pyris(Eg=1:4eV).Therefore,theabsorbanceofthedeviceinthisregionisveryhighcomparedtoallotherstates.Athigher

PAGE 201

Ingeneral,attheintermediatedopedstates,theabsorbanceoftheECcellliesbetweentheabsorbanceofthetwoextremastates,neutralanddoped.Althoughthisistruearoundtheregionwherethetransitionoccurs,thisisnolongerthecaseforhigherwavelengths.ThesechangescanbeseenclearerinFigure??wherefewerspectrahavebeenplottedthroughanarrowerspectralregion.Theelectrochromiccontrast%T

PAGE 202

Figure7.41depictstheabsorbanceofthedualreferencedeviceoverthesamespectralregionandthesameappliedvoltagerangeusedfortheactualdevice.Thesemeasurementswereperformedinordertobeabletodetermineifthedualreferencedeviceisresponsibleforanyoftheabsorbancechangesseeninthedualpolymerdevice.Forthespectralregionbetween350nm(28,571cm1)and2250nm(4444cm1)wheretheabsorbancechangesdiscussedaboveoccur,thechangeoftheabsorbanceofthedualreferencedeviceisnegligible.Therefore,theobservedchangesareduesolelytothedopinginducedstates,upontheapplicationofavoltage,ofthePBEDOT-Hx2-Pyr-Pyrlms.

PAGE 203

Figure7.42depictstheabsorbanceofthePBEDOT-Hx2-Pyr-PyrdualECdevice,wheretheabsorbanceofthereferencedevicehasbeensubtracted,atdierentvoltageswithintheappliedrange.Therefore,ideally,itdepictsthecombinedabsorbanceofthetwoPBEDOT-Hx2-Pyr-Pyrlmschangingfromaneutralstatetodopedstates. Figure7.43depictstheabsorbanceofthePBEDOT-Hx2-Pyr-PyrdualECde-vice,wheretheabsorbanceofthereferencedevicehasbeensubtracted,atvoltagesfrom0.0V(neutralstate)to+1:6V(intermediatedopedstate).Atthisvoltagerange,theabsorbanceofthetransitionisdecreasing,whereastheabsorbanceinthenearinfraredregionisincreasing.Figure7.44depictstheabsorbanceofthePBEDOT-Hx2-Pyr-PyrdualECdevice,wheretheabsorbanceofthereferencedevicehasbeensubtracted,atvoltagesfrom+1:6V(intermediatedopedstate)to+2:6V(fullydopedstate).Atthisvoltagerange,theabsorbanceofthetransitioncontinuoustodecrease,whereastheabsorbanceinthenearinfraredregionisalsodecreasing.

PAGE 204

Figure7.45depictsthetransmittanceoftheactualPBEDOT-Hx2-Pyr-PyrdualECdevice,atvoltagesfrom0.0V(neutralstate)to+2:6V(fullydopedstate).Theseresultsconrmthatafterapplyingavoltageof+2:6Vduringthepreviousmeasurements,thedeviceisstillworkinganditexhibitsanelectrochromiccontrastof65%at5000cm1.Therefore,thedecreaseofabsorbanceinthenearinfraredregionthatwasobservedisnotrelatedtoanydamageofthedevice.Strongabsorptionsinthemid-infraredregionfromtheelectrolytegelprohibitustodrawanyconclusion.Furtherinvestigationisneededinordertodeterminethesourceofthenearinfrareddecrease.7.6.3SwitchingTime

PAGE 206

Asquarewavewasappliedtothedevice,between0.0Vand+2.6V(in100secdoublepotentialsteps),whiletheabsorbanceatmaxwasmonitored.Forthisdevicethemaxwasfoundfromtheabsorbancespectrum,showninFigure7.40,tobeat880nm(11;365cm1).Thislongtermswitchingstabilityexperimentwasperformedbetweentheneutralanddopedstatesofthedevice. InFigure7.46canbenoticedthat,thedierenceintheabsorbanceofPBEDOT-Hx2-Pyr-PyrECdualdevicebetweentheextremastatesisdiminishingthroughoutthetimeoftheexperiment.Itshouldbekeptinmindthoughthat,thisisnotanoptimizeddeviceandthereforebetterperformancecouldbeachieved.7.7Conclusions

PAGE 207

Furthermore,twodierentcongurationswereusedandtransmittanceorab-sorbancemeasurementswereperformed.Therstconguration,ECdevicewithhole,employedPEDOTastheelectroactivelmwhichcanbeswitchedbetweenaneutralandap-dopedstate.Forthisreason,aholewaspunchedthroughthecounterelectrodetoallowustoprobethechangesofonlyonePEDOTlminthedevice.HavingthesecondPEDOTlmbeinthepathofthelightbeam,theelectrochromicchangesofonelmwouldcancelthechangesoftheotherandoverall,nochangewouldbeob-served.Thistypeofdevicesexhibitnicelytheelectrochromiceectoverthespectralrangefromvisibletomid-infrared,provingtheSWNTlmssuitablefortheuseinECdevices,andalsoshowedimprovedandpromisingresultssimplybycarefullyprepar-ingandhandlingofthedevice.DataanalysisandttingbasedontheDrude-Lorentzmodelwerealsopresentedandshowedthatwecanobservethedopinginducedstateswiththistransmissive/absorptiveECconguration,polaronicandbipolaronicstates,astheyhavebeendescribedintheliteratureandhavebeenobservedinreectivetypeofECdevicesliketheonespresentedinChapter5.Furthermore,indicativestabilityexperimentswereperformedandshowedpotential.Anoptimizationstudywillgreatlyimprovetheperformanceofthesedevices,sincetherearemanyparametersthatcanbeadjusted. ThiscongurationwasalsotestedbyusingthePBEDOT-Hx2-Pyr-Pyronbothelectrodesandthechangesintransmittanceupontheapplicationofdierentvoltageswererecordedforallthreestatesofthepolymerlm;oxidized,neutral,andreduced.Preliminaryresultsoftheredoxstabilitywereperformedandshowedpotential.Anop-timizationstudyisnecessaryandwillgreatlyimprovetheperformanceofthesedevices.

PAGE 208

Thesecondcongurationthatweused,dualECdevice,employedPBEDOT-Hx2-Pyr-Pyrastheelectroactivelm,whichisoneofafewpolymersthatcanbeswitchednotonlybetweenneutralandp-dopedstates,butalsobetweenneutralandn-dopedstates.Duetotheabilityofthispolymertoaccessp-andn-dopedstates,adualdevicecanbebuiltusingthesamepolymeronbothelectrodes.ThetwoPBEDOT-Hx2-Pyr-Pyrlmswereinitiallyintheirneutralstates,andupontheapplicationofvoltage,onelmbecameoxidizedwhiletheotheronebecamereduced.Theabsorbancespectrumofthisdualdevicewasrecordedoverthespectralrangefromvisibletonearinfraredatdierentappliedvoltages.Asthestateofthepolymerlmsinthedevicechangedfromneutraltodoped,wewereabletofollowthedepletionofthetransition,whichforthispolymeroccursattheborderofthevisibleandnearinfraredregion(Eg=1:4eV)andtheriseofthedopinginducedinterbandtransitionsfurtherintheinfraredregion.Moreover,indicativeswitchingstabilityexperimentswereperformedandshowedpotential.Anoptimizationstudyisexpectedtoimprovetheperformanceofthesedevices. Inconclusion,weconstructedatransmissive/absorptivedualECdevicethatex-hibitselectrochromicchangesfromthevisibletotheinfraredregion.Thisisthersttransmissive/absorptivedualECdevice,toourknowledge,thatexhibitselectrochromismintheinfraredregion.Thesubstitutionoftheelectrolytegelthatwasusedinthiswork,withadierentonethatwillnotexhibitsuchabigabsorbancebelow3600cm1,willprobablyextendtheworkingrangeofthesedevicesdeeperintheinfraredregion.Theperformanceofdevicesofthistypewillcertainlybenetfromoptimization.

PAGE 209

Inchapter5,anextensivestudyonreectiveelectrochromicdeviceswaspresentedindetail.Thisstudywasbasedonreectancemeasurements,thatledtoadeepunder-standingoftheopticalandelectrochemicalpropertiesofconjugatedpolymers,andtheconstructionofelectrochromicdeviceswithenhancedperformance.VariablereectiveelectrochromicdevicesbasedonPProDOT-Me2andPBEDOT-NMeCzwereinitiallyoptimizedtoexhibitcontrastratiosof60-70%inthemid-infraredregionwheretwostrongabsorptionpeaks,waterabsorption(O-Hstretchingmode),andC-Hstretchingmode,werehinderingtheperformanceofthedevices.Anoveltechniquewasdevelopedinordertobuildwater-freedevices,andanewsampleholderwasdesigned,thatallowscontroloftheelectrolytegelthicknessandthus,signicantlyreducestheC-Hsignature.However,theuseofthenewsampleholderhasamajordrawback;duetotheappliedpressure,theslitsontheworkingelectrodearesqueezedtogetherandthus,thediusionoftheelectrolytegelthroughtheslitsandtheiontransportisverydicult.Asacon-sequence,theswitchingtimeofthedeviceisveryslow.Morespecically,theswitchingtimeforanECdevicemountedinthenewsampleholderincreasedbyafactorof103,comparedtotheswitchingtimeofthesamedeviceundernopressure. InadditiontotheabsorptionbandsduetoO-HandC-HstretchingmodesthatwereinitiallyhinderingtheperformanceofthistypeofreectiveECdevicesinthemid-infraredregion,andtheslowswitchingtimesduetothenewsampleholderdesign,thereweremoreproblems,thataectedthewholespectrumrangeandneededtoberesolved.First,theslitsontheworkingelectrodearenotat,hencetheytendtoscatterandabsorbpartoftheincidentlight.Therefore,theintensityofthesignalthatreachesthedetectorislowduetothediusionofthereectedlight,andtheeectiveelectrochromiccontrast200

PAGE 210

Aneorttowardstheoptimizationofthistypeofdevicesledtothereplacementoftheslittedgold-coatedMylarbyahighlyporousmetallizedmembrane.Thismem-braneisapolycarbonatemembrane,purchasedfromOsmonicsInc.,with10mporesize,onwhichathinlmofgoldhasbeendepositedthroughametalvapordepositionprocess(MVD).Thegoldlayerneedstohavesucientthicknessinordertoyieldhighreectance,whileatthesametimemaintainingtheporousnatureoftheexiblemem-brane.Typically,agoldlayerof50nmmeetsadequatelytheserequirements[42].Thisreplacementfacilitatesthediusionoftheelectrolytegelbyshorteningthepath-lengththattheionshavetotravel,andthusthedopingprocessbecomesrapidandmorehomo-geneous.Therefore,aneventhinnerlayerofelectrolytegelcanbeused,whichresultsinafurtherreductionoftheC-Habsorptionband.Underthesecircumstances,betterswitchingtimesareachieved,oftheorderofsub-seconds,andhighreectancecontrastfortheentireregionfrom0.9mto5.0mareexpected[42,75]. Measurementsofthespecularreectanceinthefarinfraredregionshowedrela-tivelylowvaluesforgold,approximately40%asshonwinFigureA.1.Forthemid-infraredregionthespecularreectancewasevenlower,varyingbetweenapproximately10%and30%,FigureA.2.Measurementsofdiusereectancehaveshownhighvalues,approximately70%fornearinfrared(NIR)[42].Itisexpectedthatdiusereectancewillbehighformid-andfarinfraredregionstoo.Inadditiontospecularreectancemeasurementsofthehighlyporousmembranes,transmittancemeasurementswereper-

PAGE 211

FigureA.3showsthenewPProDOT-Me2ECcell,thestructureofwhichisthesameastheonedescribedinchapter5,atitstwoextremastates.TheswitchingtimeforthisECdeviceis0.2seconds.Lifetime,orlongtermstability,testsrevealedthatwhenthistypeofdevicesareassembledininert(Ar)environment,theycanbeswitchedforhundredsofthousandsofcycleswithlessthan10%lossofelectrochromiccontrast,%Rinthenearinfrared[38,42]. FigureA.4showsthespecularreectanceofaPProDOT-Me2ECcell,astheoneshowninFigureA.3forpartofthefarinfraredregion,from20mto200m.Sincethespecularreectanceofthesubstrate,goldonmembraneisverylow,about40%intheaboveregion,wedidnotexpectthereectanceofthecell,whichalsoincludesapolyethylenewindow,anelectrolytegellayer,andan150nmPProDOT-Me2lmonthegoldonmembranesubstrate,tobehigh.Theelectrochromiccontrastofthecellwas

PAGE 214

TableB.1:Drude-Lorentzttingparametersforp-dopedPEDOTlmsofthreedierentthicknessesonSWNTs/PET. 205

PAGE 215

TableB.2:Drude-LorentzttingparametersforthePETlm.

PAGE 216

TableB.3:Drude-Lorentzttingparametersfortheelectrolytegellayer. TableB.4:Drude-LorentzttingparametersforaPEDOTlminthreedierentstates;p-doped,slightlyp-doped,andalmostneutral. p-dopedslightlyp-dopedneutral!pj!jj!pj!jj!pj!jj

PAGE 217

FigureC.1showsthecyclicvoltammogramofPBEDOT-Hx2-Pyr-PyronITO/Glassinsolutionforanappliedvoltagerangebetween-2.2Vand+0.9V.Inthismeasurements,bothreductionstateshavebeenreachedandthepotentialforeachpeakisshownversusSCE.Thecolorsofthepolymerlminn-doped(rstreduction),neutral,andp-dopedstatesarealsoshown.PBEDOT-Hx2-Pyr-Pyrisoneofthefewpolymerscapableofformingap-dopedstateaswellasastablen-dopedstate.Then-dopedstateofelec-troactivepolymersisusuallynotstableduetothehighreactivityofoftheanionchargecarriertooxygenandwater. FigureC.2showstheabsorbanceofPBEDOT-Hx2-Pyr-PyronITO/Glassinso-lutionforanappliedvoltagerangebetween-1.09Vto+0.81VvsSCE.Inthisgure,208

PAGE 218

ThemeasurementspresentedinthisAppendixhavebeenperformedbymycollab-oratorintheChemistrydepartment,TimothySteckler.

PAGE 219

C.K.Chiang,C.R.Fincher,Y.W.Park,A.J.Heeger,E.J.LouisH.Shirakawa,S.C.Gau,andA.G.MacDiarmid.Phys.Rev.Lett.,39:1098,1977.[2] H.LethebyXXIX.J.Chem.Soc.,15:161,1862.[3] A.G.MacDiarmid,M.Akhtar,C.K.Chiang,M.J.Cohen,J.Kleppinger,A.J.Heeger,E.J.Louis,J.Milliken,M.J.Moran,D.L.Peebles,andH.Shirakawa.JournalofElectrochemistrySociety,124:C304,1977.[4] H.Shirakawa,E.J.Louis,A.G.MacDiarmid,C.K.Chiang,andA.J.Heeger.JournaloftheChemicalSociety-ChemicalCommunications,16:578,1977.[5] A.J.Heeger.Rev.Mod.Phys.,73:681,2001.[6] A.G.MacDiarmid.Rev.Mod.Phys.,73:701,2001.[7] H.Shirakawa.Rev.Mod.Phys.,73:713,2001.[8] W.R.Salaneck,R.H.Friend,andJ.L.Bredas.PhysicsReports,319:231,1999.[9] A.J.Heeger,S.Kivelson,J.R.Schrieer,andW.P.Su.Rev.Mod.Phys.,60:781,1988.[10] J.L.Bredas.HandbookofConductingPolymersII.Eds.,T.A.Skotheim,MarcelDekker,Inc.,NewYork,1986.[11] W.P.Su,J.R.Schrieer,andA.J.Heeger.Phys.Rev.B,22:2099,1980.[12] J.L.BredasandJ.B.Street.Acc.Chem.Res.,18:309,1985.[13] R.S.KohlmanandA.J.Epstein.PhysicalPropertiesofPolymersHandbookII.Eds.,J.E.Mark,AIPPress,Woodbury,NewYork,1996.[14] W.P.Su,J.R.Schrieer,andA.J.Heeger.Phys.Rev.Lett.,42:1698,1979.[15] Z.G.SoosandS.Ramasesha.Phys.Rev.B,29:5410,1984.[16] R.Kiebooms,A.Aleshin,K.Hutchison,F.Wudl,andA.J.Heeger.SyntheticMetals,101:436,1999.[17] P.Chandrasekhar.ConductingPolymers,FundamentalsandApplications:APracticalApproach.Ashwin-UshasCorp.,Inc,1999.[18] K.Sengupta,I.Zutic,H.J.Kwon,V.M.Yakovenko,andS.DasSarma.Phys.Rev.B,63:144531,2001.210

PAGE 220

K.Takimiya,M.Kodani,Y.Kataoka,Y.Aso,T.Otsubo,T.Kawamoto,andT.Mori.Chem.Mater.,15:3250,2003.[20] N.F.Mott.JournalofSolidStateChemistry,88:5,1990.[21] N.F.MottandE.A.Davis.PhilosophicalMagazine,17:1269,1968.[22] N.F.Mott.Rev.Mod.Phys.,40:677,1968.[23] R.Menon,C.O.Yoon,D.Moses,andA.J.Heeger.HandbookofConductingPolymers.2ndedition;Eds.,T.A.Skotheim,R.L.Elsenbaumer,J.R.Reynolds,MarcelDekker,Inc.,NewYork,1998.[24] C.R.Fincher,M.Ozaki,M.Tanaka,D.L.Peedles,andA.G.MacDiarmidL.Lauchlan,A.J.Heeger.Phys.Rev.B.,20:1589,1979.[25] J.Hwang,D.B.Tanner,I.Schwendeman,andJ.R.Reynolds.Phys.Rev.B.,67:115205,2003.[26] P.J.Nigrey,A.G.MacDiarmid,andA.J.Heeger.Chem.Commun.,96:594,1976.[27] J.H.Burroughes,D.D.C.Bradley,A.R.Brown,R.N.Marks,K.Mackay,R.H.Friend,P.L.Burns,andA.B.Holmes.Nature,347:539,1990.[28] I.D.Parker.J.Appl.Phys.,75:1656,1994.[29] J.R.Platt.J.Chem.Phys.,34:862,1961.[30] S.K.Deb.Appl.Optics.,Suppl.,3:192,1969.[31] G.V.Granqvist.Phys.ThinFilms,17:301,1993.[32] R.J.Mortimer.ElectrochimicaActa,44:2971,1999.[33] D.R.RosseinskyandR.J.Mortimer.Adv.Mater.,13:783,2001.[34] R.J.Mortimer.Phys.Rev.Lett.,26:147,1997.[35] C.J.Schoot,J.J.Ponjee,H.T.vanDam,R.A.vanDoorn,andP.J.Bolwijn.Appl.Phys.Lett.,64:23,1973.[36] P.R.SomaniandS.Radhakrishnan.MaterialsChemistryandPhysics,77:117,2003.[37] H.W.Heuer,R.Wehrmann,andS.Kirchmeyer.AdvancedFunctionalMaterial,12:89,2002.[38] A.A.Argun,P-H.Aubert,B.C.Thompson,I.Schwendeman,C.L.Gaupp,J.Huang,N.J.Pinto,D.B.Tanner,A.G.MacDiarmid,andJ.R.Reynolds.Chem.Mater.,16:4401,2004.[39] R.B.Bennett,W.E.Kokonasky,M.J.Hannan,andL.G.Boxall.Uspatent5,446,577.1995.

PAGE 221

P.Chandrasekhar.Uspatent5,995,273.1999.[41] I.Schwendeman,J.Hwang,D.M.Welsh,D.B.Tanner,andJ.R.Reynolds.Adv.Mater.,13:634,2001.[42] P.H.Aubert,A.A.Argun,A.Cirpan,D.B.Tanner,andJ.R.Reynolds.Chem.Mater.,16:2386,2004.[43] A.A.Argun,M.Berard,P.H.Aubert,andJ.R.Reynolds.Adv.Mater.,17:422,2005.[44] P.M.S.Monk,R.J.Mortimer,andD.R.Rosseinsky.Electrochromism:Funda-mentalsandApplications.VCH,Weinheim,Germany,1995.[45] J.D.Jackson.ClassicalElectrodynamics.3rdedition,JohnWiley&Sons,Inc,NewYork,NY,1998.[46] D.J.Griths.IntroductiontoElectrodynamics.2ndedition,PartII,PrenticeHall,EnglewoodClis,NJ,1997.[47] M.Fox.OpticalPropertiesofSolids.OxfordUniversityPress,Inc,NewYork,NY,2001.[48] F.Wooten.OpticalPropertiesofSolids.AcademicPress,NewYork,NY,1972.[49] O.S.Heavens.OpticalPropertiesofThinFilms.DoverPublications,Inc,NewYork,NY,1991.[50] A.Vas{cek.Opticsofthinlms.North-HollandPublishingCompany,Amsterdam,Holland,1960.[51] AdvisorEditorZ.Knittl.Opticsofthinlms.AnOpticalMultilayerTheory.JohnWiley&Sons,Inc,NewYork,NY,1976.[52] A.W.Crook.J.O.S.A.,38:954,1948.[53] N.W.AshcroftandN.D.Mermin.SolidStatePhysics.ThomsonLearning,Inc,1976.[54] VACVacuum/Atmospheres.InertAtmosphere&VacuumDepositionEquipment.Manual,NewYork,NY.[55] A.J.BardandL.R.Faulkner.ElectrochemicalMethodsFundamentalsandAp-plications.3rdedition,Wiley,NewYork,2001.[56] C.A.Thomas.Donor-Acceptormethodsforbandgapreductioninconjugatedpolymers:theroleofelectronrichdonorheterocycles.PhDdissertation,UniversityofFlorida,Gainesville,Florida,2002.[57] K.DoblhoferandK.Rajeshwar.HandbookofConductingPolymers.3rdedition;Eds.,T.A.Skotheim,R.L.Elsenbaumer,J.R.Reynolds,MarcelDekker,NewYork,1998.

PAGE 222

K.D.MollerandW.G.Rothschild.FarInfraredSpectroscopy.Wiley,NewYork,1971.[59] R.J.Bell.IntroductoryFourierTransformSpectroscopy.AcademicPress,NewYork,1972.[60] P.L.Richards.SpectroscopicTechniquesfortheFarInfrared,SubmillimeterandMillimeterWaves.Ed.,D.H.Martin,Amsterdam,Holland,1967.[61] W.HerresandJ.Gronholz.UnderstandingFT-IRDataProcessing;Part2:De-tailsofthespectrumcalculation.BruckerAnalytischeMesstechnikGmbH,Karl-sruhe21,WestGermany,1985.[62] W.HerresandJ.Gronholz.UnderstandingFT-IRDataProcessing;Part1:DataAquisitionandFourierTransformation.BruckerAnalytischeMesstechnikGmbH,Karlsruhe21,WestGermany,1985.[63] E.L.DereniakandG.D.Boreman.InfraredDetectorsandSystems.Wiley,NewYork,1996.[64] E.Hecht.Optics.3rdedition,Addison-WesleyLongman,Inc,1998.[65] M.Abramowitz.Microscope:BasicsandBeyond.Vol.1,FellowNewYorkMicro-scopicalSociety,forOlympusAmericaInc.[66] M.W.DavidsonandM.Abramowitz.OpticalMicroscopy.Website:,http://micro.magnet.fsu.edu/primer/opticalmicroscopy.html,August2005.[67] CarlZeissMicroscopeSystems.MPM800MicroscopePhotometer:OperationInstructions.CarlZeiss,Inc.,Germany,1993.[68] CarlZeissMicroscopeSystems.Axioskop:Routinemicroscopefortransmittedandincidentlightuorescence,OperationInstructions.CarlZeiss,Inc.,MicroscopeDivision,Germany,1993.[69] K.D.Cummings.OpticalPropertiesofaRandomSmall-ParticleComposite.PhDDissertation,UniversityofFlorida,Gainesville,Florida,1982.[70] R.D.Rauh.ElectrochimicaActa,44:3165,1999.[71] P.TopartandP.Hourquebie.ThinSolidFilms,352:243,1999.[72] A.Bessiere,L.Beluze,M.Morcrette,V.Lucas,B.Vianna,andJ.C.Badot.Chem.Mater.,15:2577,2003.[73] H.Pages,P.Topart,andD.Lemordant.ElectrochimicaActa,46:2137,2001.[74] Y.H.Chang,K.Lee,R.Kiebooms,A.Aleshin,andA.Heeger.SyntheticMetals,105:203,1999.[75] P.Chandrasekhar,B.J.Zay,G.C.Birur,S.Rawal,E.A.Pierson,L.Kauder,andT.Swanson.AdvancedFunctionalMaterials,12:95,2002.

PAGE 223

J.S.HaleandJ.A.Woolam.ThinSolidFilms,339:174,1999.[77] J.Hwang.Electrochemicalspectroscopyofconjugatedpolymers.PhDdissertation,UniversityofFlorida,Gainesville,Florida,2001.[78] I.Schwendeman.Opticalandtransportpropertiesofconjugatedpolymersandtheirapplicationstoelectrochromicdevices.PhDdissertation,UniversityofFlorida,Gainesville,Florida,2002.[79] K.NakanishiandP.H.Solomon.InfraredAbsorptionSpectroscopy.2ndedition,Holden-Day,SanFrancisco,1977.[80] C.O.BanwellandE.McNash.FundamentalsofMolecularSpectroscopy.4thedition,McGraw-Hill,London,1994.[81] M.Cornick,M.Nikolou,D.B.Tanner,I.Schwendeman,andJ.R.Reynolds.InfraredspectralenhancementofanECbywaterremoval.REUproject,Univer-sityofFlorida,http://www.phys.u.edu/REU/2002/reports/cornick.pdf,August2005.[82] S.Iijima.Nature,354:56,1991.[83] R.H.Baughman,A.A.Zakhidov,andW.A.deHeer.Science,292:787,2002.[84] M.S.DresselhausandP.C.Eklund.AdvancesinPhysics,49:705,2000.[85] X.Blase,LorinX.,Benedict,EricL.Shirley,andStevenG.Louie.Phys.Rev.Lett.,72:1878,1994.[86] R.Saito,M.Fujita,G.Dresselhaus,andM.S.Dresselhaus.Appl.Phys.Lett.,60:18,1992.[87] M.Ouyang,J.L.Huang,C.L.Cheung,andC.M.Lieber.Science,292:702,2001.[88] J.W.G.Wildoer,L.C.Venema,A.G.Rinzler,R.E.Smalley,andC.Dekker.Nature,391:59,1998.[89] T.W.Odom,J.L.Huang,P.Kim,andC.M.Lieber.Nature,391:62,1998.[90] Z.Wu,Z.Chen,X.Du,J.M.Logan,J.Sippel,M.Nikolou,K.Kamaras,J.R.Reynolds,D.B.Tanner,A.F.Hebard,andA.G.Rinzler.Science,305:1273,2004.[91] T.Guo,P.Nikolaev,A.Thess,D.T.Colbert,andR.E.Smalley.Chem.Phys.Lett.,243:49,1995.[92] A.G.Rinzler,J.Liu,H.Dai,P.Nikolaev,C.B.Human,F.J.Rodriguez-Marcias,P.J.Boul,A.H.Lu,D.Heymann,D.T.Colbert,R.S.Lee,J.E.Fischer,A.M.Rao,P.C.Eklund,andR.E.Smalley.Appl.Phys.A,67:29,1998.[93] I.W.Chiang,B.E.Brinson,R.E.Smalley,J.L.Margrave,andR.H.Hauge.J.Phys.Chem.,105:1157,2001.

PAGE 224

Z.Chen.Electriceldinducedtransparencymodulationinsinglewallcarbonnan-otudeultra-thinlmsandamethodtoseparatemetallicandsemiconductingnano-tubes.PhDDissertation,UniversityofFlorida,Gainesville,Florida,2003.[95] M.E.Itkis,S.Niyogi,M.E.Meng,M.A.Hamon,H.Hu,andR.C.Haddon.NanoLetters,2:155,2002.[96] A.Ugawa,A.G.Rinzler,andD.B.Tanner.Phys.Rev.B,60:R11305,1999.[97] A.Ugawa,J.Hwang,H.H.Gommans,H.Tashiro,A.G.Rinzler,andD.B.Tanner.Curr.Appl.Phys.,1:45,2001.[98] S.M.Bachilo,M.S.Strano,C.Kittrell,R.H.Hauge,R.E.Smalley,andR.B.Weisman.Science,298:2361,2002.[99] P.G.Collins,K.Bradley,M.Ishigami,andA.Zett.Science,287:1801,2000.[100] X.Liu,T.Pichler,M.Knupfer,andJ.Fink.Phys.Rev.B,70:205405,2004.

PAGE 225

IwasborninAthens,Greece.Afternishinghighschool,IenrolledinthePhysicsDepartmentoftheUniversityofCrete,whereIobtainedaB.S.inphysics(2000),withaspecializationinmicroelectronics.ThethesisworkwascompletedworkingintheMicroelectronicsResearchGroup(MRG).In1999,undertheauspicesoftheErasmusstudentexchangeprogram,IstudiedforsixmonthsattheDepartmentofExperimen-talSolidStateIII,intheResearchInstituteforMaterials(RIM)oftheUniversityofNijmegeninHolland.Myinterestsatthattimewerefocusedinthegrowthandopticalcharacterizationofinorganicsemiconductors. In2000IjoinedthegraduateschoolofthePhysicsDepartmentoftheUniversityofFlorida.Afterayearofcorecourses,IjoinedProfessorTanner'sresearchgroup.Twojointprojectsimmediatelydrewmyattention:oneonelectrochemicallypreparedconjugatedpolymersincollaborationwithProfessorReynolds'groupfromtheChemistryDepartmentandProfessorRinzler'sresearchontransparent,singlewallednanotubelms.Bothprojectsappearedfascinatingfromtheverybeginning.Ihaveactivelypursuedresearchintheseareas,anendeavorthathasbeenveryrewardingsince.216


Permanent Link: http://ufdc.ufl.edu/UFE0012152/00001

Material Information

Title: In-situ Spectroscopic Studies of Single-Walled Carbon Nanotubes and Conjugated Polymers in Electrochromic Devices
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0012152:00001

Permanent Link: http://ufdc.ufl.edu/UFE0012152/00001

Material Information

Title: In-situ Spectroscopic Studies of Single-Walled Carbon Nanotubes and Conjugated Polymers in Electrochromic Devices
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0012152:00001


This item has the following downloads:


Full Text










IN-SITU SPECTROSCOPIC STUDIES OF SINGLE-WALLED CARBON
NANOTUBES AND CONJUGATED POLYMERS IN ELECTROCHROMIC
DEVICES












By
MARIA NIK(OLOU


DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA


2005















ACK(NOWLEDG1\ENTS

The past five years have been an intensely
journey. During that time, I have been very fortunate to have been surrounded by

exceptional people that greatly influenced my career and at the same time inspired me.

The first person to whom I would like to express my sincere gratitude is my research

advisor, Professor David B. Tanner, who welcomed me in his group, supported me and

encouraged me front the very beginning. The time I have spent working under his

supervision has proved to be an invaluable experience. Furthermore, I would like to

thank hint for his trust, and for allowing me to pursue a research path, which required

close collaboration with the C'h.~ ~!!I s-y Department.

In that different universe, a person that also had aml I i9 .r impact on my work, and

to whom I would like to express my sincere gratitude, is Professor John R. Reynolds,

who in reality has been a second advisor to me. For the last three years, he has welcomed

me in his group, and he has tirelessly supported me, guided me, being ahr-l- .- interested

in and appreciative of my work. I have to admit that I feel extremely privileged having

collaborated with these two excellent researchers and, above all, exceptional people.

I would also like to express my appreciation to many people: Professor Richard

Woodard, for his help organizing my transition from Greece and with my initial ad-

justnlent to a new environment; Professor K~atalin K~anarks, for her help with the ex-

perintents, data analysis, guidance, input, and useful discussions involving the whole

project described in ('! .pter 6; also, Professor Andrew G. Rinzler for providing high

quality samples, and for .ll.--le .v patiently answering all my questions concerning the

subject of nanotubes which is covered in ('! .pters 6 and 7. I am also indebted to my

supervisory coninittee, Professors Alan T. Dorsey, Arthur F. Heard, and Pierre Sikivie,

for their interest in serving on my coninittee.










Several coworkers have had an important role throughout the completion of the

work presented in this dissertation. Fr-on the C'h.~ Ins!-1ry Department I would like to

thank my collaborators: Dr Irina Schwendenian, for her help when I was just starting my

first project, and Dr Avni A. Argun for many useful and enjoi-,1.1-- discussions, as well

as for his help in the chemistry laboratory. I am also very grateful to Aubrey L. Dyer

for many useful and enjoi- .1.1.' discussions, her valuable help in the chemistry laboratory,

and her friendship. I would also like to thank Timothy Steckler for providing me with

the nionomer for the work that is outlined in ('! .pter 7, and Nisha Ananthakrishnan

for help with the AFM images. Finally, I would like to thank all remaining nienters

of Prof. Reynolds' group; their group meeting presentations and useful discussions were

an invaluable resource for me. I also wish to thank them for providing a pleasant, and

productive environment to work.

Front the Physics Department, I would like to thank my collaborators front Prof.

Rinzler's group: Dr Zhihong ('ll, is for providing the samples for the research outlined in

('!, Ilter 6 and for many useful discussions, Zhuangchun Wu for providing the samples

for the research outlined in ('! .pter 7, and also Jennifer Sippel Oakley for help with the

AFM images. Front Prof. K~anarks' group (Research Institute for Solid State Physics

and Optics, Hungarian Academy of Sciences, in Budapest, Hungary), I thank Ferenc

Borondics for collaboration on the research project outlined in ('! .pter 6. Last but not

least, I would like to thank the people in Prof. Tanner's group, my past and present

colleagues, the people I worked side by side with throughout my graduate years, for their

cooperation, useful discussions, and their friendship. I would also like to thank Matt

Cornick, an undergraduate student who joined our group through the REU program,

for his help with the experiments outlined in ('! .pter 5, and Nathan Heston for his help

with the last nmid-infrared experiment outlined in ('! .pter 7.

I would also like to acknowledge the help of the nienters of the machine shop,

especially Marc Link and John Van Leer, the people in the electronics shop, and the

cryogenics team.










Furthermore, I wish to thank all my friends in Gainesville for the fun times we had

together throughout these years. Finally, my deepest appreciation goes to my husband

and my family for their constant support and love.

This work is dedicated to the memory of Maria B. Panousi.

















TABLE OF CONTENTS


Page

ACKNOWLEDGMENTS . . ii

ABSTRACT . . viii

CHAPTER

1 INTRODUCTION . . 1

2 CONDUCTING, OR CONJITGATED, POLYMERS . . 5

2.1 Non-conjugated and Conjugated Polymers ... .. .. 5
2.2 Classification of Conjugated Polymers ... .. .. .. 9
2.2.1 Doping Mechanisms in DGSPs and NDGSPs .. .. .. .. .. 11
2.3 Theoretical Models ...... .. ... .. 17
2.4 Conductivity in Conjugated Polymers (CPs) ... .. .. 19
2.5 Aletal-Insulator (11-1) Transition ...... .... .. 21
2.6 Doping Induced Properties in Conjugated Polymers .. .. .. .. .. 22
2.7 Doping Methods for Conjugated Polymers ... .. .. 24
2.8 Fundamentals of Electrochronmisni ...... .... .. 28
2.9 Synthesis Methods of CPs . ...... ... .. :31
2.10 C'!I. .) ..terization Methods of Electrochrontic CPs .. .. .. 3:3
2.11 Electrochrontic Devices (ECDs) Based on CPs ... .. .. :36
2.12 General Applications of CPs . .... .. .. :38

:3 THIN FILM OPTICS . . 40

:3.1 Optical Processes and Optical Constants ... .. .. 40
:3.2 Interaction of Electromagnetic Waves with Matter .. .. .. .. 42
:3.3 Light Propagation Through a Planar Interface ... .. . .. 49
:3.4 Light Propagation Through a Single L ... ir Structure .. .. .. .. 5:3
:3.5 Light Propagation Through a Multi-L ... r- Structure .. .. .. .. .. 58
:3.5.1 Matrix Method ....... .... .. 61
:3.6 K~raniers-K~ronigf, or Dispersion, Relations .... .... .. 62
:3.7 Models for the Determination of Optical Constants .. .. .. 64
:3.7.1 Lorentz Model .. ... .. .. 64
:3.7.2 Drude Model ......... .. .. 68
:3.7.3 Drude-Lorentz Model . ..... .. 71
:3.7.4 Sunt Rules ........ ... .. 72











4 INSTRUMENTATION AND EXPERIMENTAL TECHNIQUES .

4.1 Dry Box.
4.2 Electrochemical Methods .. ........
4.2.1 Electrochemical Polymerization ....
4.2.2 Cyclic Voltammetry (CV)
4.3 Optical Methods
4.3.1 Spectroelectrochemistry
4.3.2 Interferometric or FTIR Spectrometer
4.3.3 Monochromatic Spectrometer .....


. 73


5 REFLECTIVE ELECTROCHROMIC DEVICES (ECDS) .

5.1 ECDs Fabrication ....... ..
5.2 In-situ Reflectance Measurements and All .k .
5.3 Results and Discussion .......
5.4 Enhancingf the Performance of ECDs .....
5.5 Conclusions .....


. . 104

..... 105
. . 109
. 110
.. .. 114
.. .. 120


6 FREE-STANDING SINGLE WALLED CARBON NANOTUBE FILMS 122

6.1 Carbon Nanotubes . .. .. ... .. 122
6.1.1 Electronic Structure of Carbon Nanotubes .. . .. 125
6.1.2 Density of States (DOS) of SWNTs .. . .. 127
6.1.3 Carbon Nanotube Synthesis and Purification .. .. .. .. .. 131
6.1.4 General Applications of CNs .... .... . 132
6.2 Free-Standing SWNT Films:
Transparency, and Optical Conductivity .... .. .. 133
6.2.1 Sample Preparation . ..... ... .. 133
6.2.2 In-situ Transmittance Measurements ... .. .. 136
6.2.3 Optical Constants, Data Analysis and Model Fit .. .. .. .. 146
6.2.4 Infrared Spectral Weights ..... .. .. 152
6.2.5 Conclusions .. ... ... .. .. 152


7 TRANSMISSIVE/ABSORPTIVE EC DEVICES . . 155


7.1 Sample Preparation


.


.... 155
.... 155
.... 157
.... 157
.... 159
.... 160
.... 160
.... 164
.... 166
.... 168
.... 168
**** 169


7.1.1 Purification of Laboratory C!. isis. II< and Materials ....
7.1.2 Electrode Preparation ......
7.1.3 Electrochemical Polymerization ......
7.1.4 Gel Electrolyte Preparation ......
7.2 PEDOT on SWNT Electrodes ........
7.2.1 Spectroelectrochemistry of PEDOT ......... .
7.2.2 Thickness Determination of PEDOT on SWNT Electrodes
7.2.3 Atomic Force Microscopy (AFM) Images ...... .
7.3 PBEDOT-HX2-Pyr-Pyr on SWNT Electrodes ......
7.3.1 Cyclic Voltammetry of PBEDOT-HX2-Pyr-Pyr .......
7.3.2 Spectroelectrochemistry of PBEDOT-HX2- FT-Py .E-****












7.4 Transmissive/Absorptive EC Device Construction ..... .
7.5 In-Situ Transmittance Measurements of EC Devices .......
7.5.1 Analysis of the Different L n. ris in the EC Device .....
7.5.2 Analysis of a PEDOT EC Device (with hole) ........
7.5.3 Results on PBEDOT-HX2-Pyr-Pyr EC Device (with hole) .
7.6 Towards the construction of a Dual Device ......... .
7.6.1 Cyclic Voltammetry of PBEDOT-HX2-Pyr-Pyr dual device
7.6.2 In-Situ Absorbance Measurements ......
7.6.3 Switchingf Time .......
7.7 Conclusions ......


.. .172
.... 174
.... 175
.... 176
.... 183
.... 189
.... 189
.... 191
.... 195
.... 197


APPENDIX


A NEXT GENERATION OF REFLECTIVE ECDS USING MICROPOROUS GOLD
ELECTRODES . . 200


B DRUDE-LORENTZ FITTING PARAMETERS . . 205


C PBEDOT-HX2-PYR-PYR . . 208

REFERENCES . . 210


BIOGRAPHICAL SKETCH . . 216















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

IN-SITU SPECTROSCOPIC STUDIES OF SINGLE-WALLED CARBON
NANOTITBES AND CONJUGATED POLYMERS IN ELECTROCHROMIC
DEVICES

By

Maria Nikolou

December 2005

C'I nI~ I s..i: David B. Tanner
Major Department: Physics

Our initial focus has been the optimization of the performance of reflective elec-

trochromic devices. In-situ reflectance measurements and extensive subsequent analy-

sis enabled us to construct devices with significantly enhanced performance. Variable

reflective electrochromic devices based on PProDOT-?10 and PBEDOT-N~leCz were

optimized to exhibit contrast ratios of 60 70 in the mid-infrared region. In the latter

region, two strong absorption peaks, O-H stretching mode and C-H stretching mode,

were hindering the performance of the devices. A novel technique was developed in

order to build water-free devices. In addition a new sample holder was designed that

permits control of the electrolyte gel thickness, and reduces the C-H signature.

Building upon the aforementioned techniques, we designed an infrared transmis-

sive/absorptive type of electrochromic cell. The construction of such a device had been

previously impossible due to the absence of a suitable, infrared transparent, conducting

material. An extensive study of single-walled carbon nanotubes as free-standing films

showed that these films have the highest transmittance among transparent conductors

in the 2 5 m spectral range, at very low sheet resistance. Transmittance measure-

ments on purified (p-doped) and vacuum annealed (de-doped) carbon nanotube films










were performed and analyzed front far infrared through ultraviolet, at temperatures he-

tween 50 K 300 K(. The result of this study led us to propose the construction of the

very first transmissive/absorptive EC device in the infrared region using SWNTs as the

conducting lI .0 c. The polymer that was used in the devices is PBEDOT-Pyr-Pyr which

can he both p- and n- doped.















CHAPTER 1
INTRODUCTION

The main focus of this work is the optical characterization of conjugated poly-

mers in different electrochromic device platforms, in an effort to optimize the perfor-

mance of these devices. Furthermore, an attempt to develop the first infrared trans-

missive/absorptive electrochromic device, led to extensive studies on single-wall carbon

nanotude (SWNT) films. This material proved to be the best candidate for the replace-

ment of the conventional indium-doped tin oxide (ITO), which served as the conducting

1 e. -r in electrochromic devices, for the infrared region. Preliminary results of the per-

formance of this type of devices are also presented in this work.

More specifically, the second chapter offers an introduction to the theoretical back-

ground that is necessary for the understanding of conjugated polymer systems. This

includes a presentation on the different types of conjugated polymers and their unique

characteristics, such as the doping mechanisms, and the induced structural and opti-

cal property changes. A brief description in the charge transport mechanisms, and the

metal-insulator transition occurring in conjugated polymers is also given in this chapter.

Different polymerization, doping, and characterization methods, and several theoretical

models that attempt to explain the behavior of these systems are introduced. Finally,

the phenomenon of electrochromism is defined, and general applications based on con-

jugated polymers are presented.

The third chapter is a review of the basic laws that govern the interaction of electro-

magnetic waves with matter. Optical processes and common techniques for extracting

the optical constants, which characterize each system, are introduced and explained.

Light propagation through different structures is described, and the formulas for the

calculation of the reflectance and transmittance through these structures, as well as the

related optical parameters are derived. The chapter concludes with the presentation of










the dispersion, or K~ramers-K~ronigf, relations and the models, Drude and Lorentz models,

that are used for the determination of the optical properties of the media.

The fourth chapter introduces the instrumentation, and experimental techniques

used throughout this work. A description of systems such as the dry box, or glove box,

and the different types of spectrometers, interferometric and monochromatic, is provided.

The concepts of Fourier spectroscopy are presented. Techniques for the production

and/or characterization of samples, such as electrochemical polymerization and cyclic

voltammetry are introduced.

In chapter five, the fabrication of reflective electrochromic devices is presented.

Details on the design, the .1--, ~!,11;... and the materials used are provided. In-situ

reflectance measurements, and analysis of the experimental results are discussed. The

problems that were encountered, and the no~i~ to overcome them in an effort to enhance

the performance of these devices are also presented. Finally, a design of a new type of

sample holder, which is part of the problem solution is described here.

C'!s Ilter six includes an introduction on the theoretical background that is neces-

sary for the understanding of carbon nanotube systems. This includes a presentation

on the different types of nanotubes and their unique characteristics, such as the de-

pendence of their electronic structure on the tube diameter and wrapping angle of a

graphene sheet, with no doping impurities present. Furthermore, the electronic density

of states (DOS) in carbon nanotubes is not continuous, as it is in graphite, but it di-

vides into a series of spikes caused by the quantum confinement of electrons in the radial

and circumferential directions. These spikes are known as van Hove singularities, and

they comprise a typical signature of 1D systems. Typical synthesis, purification tech-

niques and general applications of carbon nanotubes are mentioned, as well as a brief

description of the sample preparation. Transmittance spectra of as-prepared (purified),

p-doped, and vacuum annealed, de-doped, free-standing SWNT films of different thick-

nesses, are presented from the far infrared through the ultraviolet region, at different

temperatures. The aI~ i h--;- of the data includes the use of K~ramers-K~ronigf relations for










the determination of the optical properties of carbon nanotubes, optical conductivity

and infrared spectral weights, and the Drude-Lorentz model for the fittingf of the trans-

mittance data and the optical conductivity. The conclusions that were drawn from this

extensive study are presented at the end of the chapter.

In C'!. Ilter seven, transmissive/absorptive EC devices were built using two differ-

ent electroactive polymers and, for the first time to the best of our knowledge, SWNT

films were used as the conductive 1.s. ris in EC devices. Initially, we conducted a compar-

ison study between the performance of SWNT films, used as the conductive electrodes,

and the well studied and extensively used ITO. Cyclic voltammograms and absorbance

measurements in solution for both polymers used in the devices were performed, and

conclusion were drawn. Furthermore, two different configuration were used, and trans-

mittance or absorbance measurements, data analysis based on the Drude-Lorentz model,

as well as stability studies were performed. The first configuration, is an EC device with

hole on the counter electrode, emploi-. I PEDOT as the electroactive film which can be

switched between a neutral and a p-doped state. The second configuration, is a dual

EC device, emploi-- 4 PBEDOT-HX2-PyT-Pyr aS the electroactive film which is one of a

few polymers that can be switched not only between neutral and p-doped states, but

also between neutral and n-doped states. The constructed transmissive/absorptive dual

EC device that exhibits electrochromic changes from the visible to the infrared region.

This is the first transmissive/absorptive dual EC device exhibits electrochromism in the

infrared spectral region. Further optimization studies are needed for both types of de-

vices. The conclusions that were drawn from this extensive study are presented at the

end of the chapter.

In spectroscopic studies, different energy units are often used for different tech-

niques, different parts of the electromagnetic spectrum, and between different disciplines.

In general, the most commonly used unit for the infrared region of the spectrum is the

frequency or wavenumber, expressed in cm l, whereas for the visible region of the spec-

trum, is the wavelength expressed in nm. Throughout this work these units are used










interchangeably. Hopefully this fact does not cause any confusion to the reader, since

these units are simply related to each other: v(ent-l)= 104 p(m)= 107/A(nnt). The

relation between the frequency and energy units is provided in Table 1.1.


Table 1.1: Relation between energy units.



1 ent-l 0.124 nieV
1 nieV 8.0658 ent-]















CHAPTER 2
CONDUCTING, OR CONJUGATED, POLYMERS

The discovery of conducting polymers in 1977 by Alan J.Heeger, Alan G. Mac-

Diarmid, and Hideki Shirakawa [1] created a new research field that has been studied

intensively since. Although conducting polymers had been initially introduced in 1862

when poly(aniline), or PANI, was synthesized [2] it was the collaborating effort in 1977

of the 2000 Nobel laureates [3-7] which led to the discovery of poly(acetylene) (PAc),

that invigorate the interest of the research community. This work showed that organic

conjugated polymers have the ability to be doped over the full range from insulator to

metal, and offered the promise of a new type of polymers: materials which exhibit the

electrical and optical properties of metals, or semiconductors, while retaining the attrac-

tive mechanical properties and processing advantages of polymers. This new generation

of polymers created a new field of interdisciplinary research, on the boundary between

chemistry and physics, and new enormous potential applications.

2.1 Non-conjugated and Conjugated Polymers

Polymeric materials, such as plastics, date back to the 19t" century. These mate-

rials, also called non-conjugated or saturated polymers, were basically known for their

mechanical and chemical properties, and were considered to be electrically insulating

materials. This is due to their electronic configuration which is 1S2 2s2 2p2, and thus,

the carbon (C) atoms have their four valence electrons tied up in covalent bonds. These

polymers have an sp3 hybridization, which means that each carbon has four bonds with

equivalent energy, four a-bonds. The a a* gap, or energy band gap, E,, in condensed

matter terminology, is large, i.e., 8-10 eV, and therefore non-conjugated polymers are

electrically insulating and transparent in the visible region.










Conducting, or conjugated, polymers are the most recent generation of polymers,

and their electronic configuration is fundamentally different. In conjugated polymers,

carbon atoms have the following hybridization: sp2pz, which leads to three equivalent

o--honds and one xr-hond from the remaining pz atomic orbital. The xr-honding, which

occurs when two pz orbitals of successive carbon atoms along the backbone overlap, re-

sults in electron delocalization, and hence in charge mobility along the polymer chain [8].

This delocalization is responsible for the unique properties of conjugated polymers. The

r-~r* gap, or energy gap, E,, is relatively small, i.e., 1-4eV, and this is the reason

for the semiconducting behavior of conjugating polymers. Therefore, the polymer chain

symmetry pbovi- an important role in determining the electronic structure, and conse-

quently careful design of the monomer unit along with doping can result in systems with

metallic properties. The bond formation of conjugated polymers is shown schematically

in Figure 2.1.


p orbital p orbital





Cpi bond





sigma
bond

Figure 2.1: Formation of o--hond by strong overlapping of two sp2- OrbitalS and formation
of xr-honds by weak overlapping of the pz-orbitals of two successive carbon atoms.


Each carbon atom along the backbone of conjugated polymers has one unpaired xr -

electron, the orbital of which overlaps strongly with the orbitals of the nearest unpaired

xr electrons and weakly with orbitals of unpaired xr electrons in different polymer chains.

The first one is known as intrachain interaction and the second as interchain interactions.

These bonds, strong inside the polymer chain and weak, van der Waals type, between










different chains, make the system essentially quasi-one dimensional in the sense that

the charge carriers move free only along the backbone of the polymer chain [9]. Based

on this intrinsic low dimensional geometry of polymers, such as poly(acetylene), early

theoretical studies treated them as one-dimensional metals, and expected them to have

equal C-C bond lengths along the chain Figure 2.2.










crystal lattice









k,=Tc/2at x/a
electron density of states
dispersion relation
without distortion


Figure 2.2: The polymer chain treated as one-dimensional metal, with equal C-C bond
lengths along the polymer chain.


However, according to Peierl (1955) the ground state of such a one-dimensional

metal is unstable with respect to a structural distortion, which results in the creation of

alternating double and single bonds, as shown in Figure 2.3. As a result, a spontaneous

symmetry breaking occurs, and an energy gap opens at the Fermi level rendering the

material semiconductor [9]. These symmetry hII, I1:;14 or Peierl's distortion, doubles the

unit cell, as shown in Figure 2.4, and concentrates the electronss between alternating

pairs of carbon atoms. This is consistent with conjugated polymers having alternating

single and double bonds, or longer and shorter bonds (1.446 A~ and 1.346 A~ respectively















































kp-R/2a 1
states
dupe~rsion relation
with a per~iodlic distortion


Figure 2.4: Periodic distortions or defects for systems with half filled band. E, is the
band gap cause by the distortion; kF is the Fermi wavevector; a is the size of the unit
cell before the distortion.

The lowering of the symmetry lowers the energy of the occupied states and sta-

bilizes the distortion. Thus, the xr band is split into two subbands, a fully occupied

xr band (also called the HOMO: Highest Occupied Molecular Orbital, valence band, or


88 88
2a



yc


[10]) along their backbones. Figure 2.5 illustrates the bond alteration for a number of

conjugated polymers.


metallic state
(unstable)

dimerization
S(Peierls transition)
semiconducting state
(~n-~n* band gap)



Figure 2.3: Peierls transition or dimerization, trans-poly(.ll 1vlilene) is shown as an ex-
ample.


**
rstal lattice


E


electron density of


X/a










and therefore higher in energy than the aromatic phase or benzonoid (or benzoid) type

of structure (A). Other polymers included in this category are cis-PAc, PTh, PPP, as

shown in Figure 2.5.

2.2.1 Doping Mechanisms in DGSPs and NDGSPs

Initially, theoretical models from condensed matter physics that apply to conven-

tional semiconductors were adopted by workers, in an effort to understand the doping

mechanisms in conjugated polymers. However, it was soon realized through experi-

mental results that the doping mechanisms in these organic materials is fundamentally

different. Although conjugated polymers develop xr and xr* orbital bands that can be

thought of as the valence and conduction bands of a conventional semiconductor, the

doping process is quite different. The introduction of charges in these polymer systems

leads to a structural distortion of the lattice in the vicinity of the charge, which lowers

the energy of the system and stabilizes the charge.

More specifically, upon doping of conjugated polymers, different types of excita-

tions occur depending on which of the categories, DGSPs or NDGSPs, the polymer

belongs. For degenerate ground state polymers, e.g., trans-poly n .- tvilene, the intro-

duction of electrons, and/or holes, to the polymer chain, creates a domain wall that

separates regions of different bonding structure, i.e., phase A and B Figure 2.8. These

excitations were initially called im!-11 but later, in view of the fact that the domain

wall is a nonlinear shape preserving excitation which propagates freely along the poly-

mer chain, they were called -. .!ltons" [9]. Solitons are topological excitations, and since

they convert phase A structure to phase B structure, and vice versa, in a perfect infinite

chain they can only be created or destroyed in pairs.

While the soliton has an obvious effect on the lattice distortion pattern of the

polymer chain, it also has a remarkable effect on the electronic structure. The localized

distortion gives rise to a single localized electronic state in the middle of the energy

band gap region. This mid-gap state is a solution to the Schroidinger equation in the

presence of a structural distortion and therefore, can accommodate 0, 1, or 2 electrons.










At higher doping levels, the formation of a charged hipolaron, or radical dication

or dianion, is the stable excitation for NDGSPs. A hipolaron is a bound state of two

charged solitons of the same charge, or two polarons of the same charge whose neutral

solitons annihilate each other. Thus, hipolarons possess charge but carry no spin. The

creation of a hipolaron generates two syninetrical nmid-gap energy states, as in polarons

but located closer to the center of the gap. When these nxid-gap states are empty

the hipolaron is positive (p-type doping), whereas when they are fully occupied the

hipolaron is negatively charged (n-type doping). These excitations are delocalized over

several monomer units, generally six to eight, and thus can propagate along the polymer

chain.

As the doping level is further increasing, the individual hipolaron levels described

above, shown in Figure 2.13, coalesce into hands. These hipolaron hands arise front the

depletion of electronic states front the valence and conduction hand edges, which results

in a concomitant increase of the xr xr* gap. At these high doping levels, hipolarons can

give rise to high conductivity upon the application of an electric field. ESR experiments

have proven that in highly conductive conjugated polymers (CPs) the charge carriers

are spinless, and therefore hipolarons, since no signal can he detected [12].

1\ost cases considered in literature refer to positively charged polarons and hipo-

larons, i.e., p-type doping or oxidation of the conjugated polymers, because most CPs

cannot form stable negatively polarons or hipolarons.

The concept of doping in CPs is unique for many reasons. First, the doping process

is reversible, and the de-doping process produces the neutral polymer with little or no

degradation. Second, it involves the formation of mobile excitations, such as solitons,

polarons, and/or hipolarons, that couple to lattice vibrations. Third, unlike doping in

conventional semiconductors, in CPs the dopant atoms do not substitute atoms within

the lattice but are positioned between the polymer chains, and donate or accept electrons

front the polymer backbone [13]. Finally, the doping level can he controllably adjusted,

and thus the conductivity can he tuned over several orders of magnitude in the same










material. Conductivities varying front the insulating or senticonducting regime, when

the CP is in the pristine form, to highly conducting regime, when the CP is fully doped,

can he easily obtained.

2.3 Theoretical Models

Conjugated polymers are complicated systems, mainly due to the many degrees of

freedom they possess, e.g., o- hands, vr hands, lattice vibrations, intrachain and interchain

interactions, and the nonlinear phenomena which are characteristic of one dimensional

systems. Furthermore, real life makes the situation even more difficult, since the polymer

chains have limited conjugation lengths, may contain defects or impurities, and cross-

linking. 1\oreover, different synthetic methods can produce quite different morphologies,

and hence different properties, for the same polymer. All these factors make a complete

theoretical description a real challenge.

The highly anisotropic character shown by these one dimensional materials causes

the electronic motion to be easy only along the backbone of the polymer chain. There-

fore, the properties of these materials are mainly governed by the following types of

interactions among the unpaired electrons that occupy the highest molecular orbitals in

the solid: (1) the overlap of the wave functions of these electrons between .Il11 Il-ent sites

in the polynteric chain; (2) the interactions of the electrons with their surroundings, in

particularly with the lattice vibrations (electron-phonon coupling); and (3) the Coulomb

interaction between the electrons (electron-electron coupling). Various theoretical niod-

els, based on the simple tight finding model, have been developed, taking into account

some of the above interactions in an effort to provide a more realistic description of these

system.

A very well known model, used for the theoretical treatment of conjugated polymer

systems, is the Su-Schrieffer-Heeger (SSH) model. This method was initially applied to

poly(acetylene), and yielded useful information for the interpretation of experimental

data. This SSH treatment of polynteric systems includes the electron-lattice vibration

interactions, onlits the electron-electron interactions, and models the CP chain, or lat-









tice, potential in the form of classical springs between neighboring chain sites. Therefore,

the SSH Hamiltonian is written as follows [9, 11, 14]:




HS'SH n+, C:ti I:L~ +1,s~LnS C r c,s, ) + I, u,)2 + g (2.1)


where it has been assumed that the xr electrons can be treated in the tight binding

approximation with a hopping integral t, 1,,, which can be expanded to first order

around the undimerized state (before the Peierl's distortion, all bonds are of equal

length) :



tn+l,n = to + a~(U,+1 U,), (2.2)

where to is the hopping integral for the undimerized state, and a~ is the electron-lattice

displacement (phonon) coupling constant. This linear approximation is valid, since the

bond-length c~hangles are small, of the order of 0.08 A. In equation (2.1), c~,, and c,,S

create and destroy xr electrons of spin +1/2 on the uth repeated unit. These creation

and annihilation operators satisfy the anticommutation relations of fermions. The second

term in this equation corresponds to the o- bonding energy, which has been expanded to

second order around the undimerized state. The first order term can be neglected due

to symmetry. Finally, the third term in equation (2.1) is the kinetic energy of nuclear

motion, where M is the total mass of the repeated unit. This term can also be written

as MGu, = p,, where p, and u, satisfy canonical commutation relations:





Another model used for the theoretical treatment of conjugated polymer systems,

is the Pariser-Parr-Pople (PPP) model. This model considers the electron-electron cou-

pling, or Coulomb interaction, while neglecting the electron-phonon coupling. The

PPP Hamiltonian is written as follows [15]:













HPP= t ~,i(ci c;,risc~,cs 1s)+Ui nCnz e,4i~ + ~ C (ni l 1)(q,-1) (2.4)
i,s i i,i/

where



ne, =~~l c c:,y c s, (2.5)

is the number operator. U, 1*i are a way to parameterize the electron-electron interac-

tion. The potential U is effective only when two electrons occupy the same site, whereas

1,4, is effective for electrons that occupy different sites.
In addition to the ones generally described above, there are other well known

models used for the theoretical treatment of conjugated polymer systems. These include

the Hubbard model, the extended Hiickel model, the Valence Effective Hamiltonian

(VEH) method, the Peierl-Froihlich method, and more.

2.4 Conductivity in Conjugated Polymers (CPs)

Since the discovery of conjugated polymers [1] the main goal has been the re-

placement of conventional metals and inorganic semiconductors with these low cost and

light weight materials. The un-doped, or pristine, materials are generally semicondu-

ctors with conductivities typically in the range 10-10-10-5 S/cm, but upon doping their

conductivity can increase by several orders of magnitude and render them metallic.

Poly(acetylene), which initially was the most studied CP, upon heavy doping with io-

dine can achieve very high electrical conductivity, close to the electrical conductivity of

copper (~-106 S/cm). However, its high chemical instability under ambient conditions

has precluded the use of P(Ac) from commercial applications, and has confined it to its

scientific aspects. Figure 2.14 shows the range of electrical conductivities of the most

common conjugated polymers as the doping level is varied.

Experimental studies have shown [5] that the electrical conductivity of CPs im-

proves as the degree of chain extension and chain alignment is increased. The studies










conduction mechanisms. Some of the most well known models are the 1\ott Variable

Range Hopping (VRH) model, Sheng model, and the K~ivelson model [17]. Each of these

models is valid under different conditions, such as different doping levels, temperature

ranges, etc.

Finally, given that conjugated polymers show tunable conductivities in their doped

states with values, in some cases, that can reach those of conventional metals, one cannot

help but wonder if there is a possibility that these systems will exhibit superconductiv-

ity. Although conjugated polymers share many features with organic materials, such as

the tetllsr .ss, tltetraselenafulvalence family, (TIlTSF)2X [18], and methylenedithiote-

traselenafulvalene family (\!l )T-TSF)X1.27--1.29 [19] where X is usually a halogen, which

exhibit superconductivity, this phenomenon has not been observed yet for doped conju-

gated polymeric systems. Workers in the field appear to be optimistic towards this idea

but there is still a lot of progress that needs to be done. Currently available materials

are barely metallic with electronic properties which are dominated by disorder, render-

ing the characteristic mean free paths in the region for disorder-induced localization.

Therefore, the first step in the direction of truly metallic conjugated polymers is the

improvement of the materials quality, which will eventually result in longer mean free

paths, ideally of the order of the monomeric unit.

2.5 Metal-Insulator (M-I) Transition

X-ray studies have shown that CP systems generally consist of ordered, highly

conductive, regions, surrounded by less ordered or amorphous material [13]. As the

degree of disorder increases in a metallic system, there is a point at which the mean free

path becomes equal to the interatomic spacing, and that results in the localization of

the charge carriers which renders the material in a non-conducting state. The criterion

for this metal-insulator transition, proposed by loffe and Regel in 1960, is defined as:


kFla1 1


(2.6)










where kF is the Fermi wavenumber, and I is the mean free path. The metallic regfime

exists for kF1l > .

Based on the above criterion, 1\ott stated that for a metal-insulator transition to

occur, the disorder should be sufficiently large that kFl > 1 [20-22]. In the limit where

kFl
thus, all states become localized. In this case, the system becomes an insulator and is

called a Fermi glass. Even though Fermi glasses have a continuous density of states and

no energy gap, they behave as insulators as a result of the spatially localized states at

the Fermi level.

In conjugated polymeric systems, this metal-insulator transition is very interest-

ing. This, is due to the ability to control the critical regime by varying the degree of

disorder of the system, or by applying external pressure and/or magnetic fields. The

critical behavior, close to the phase transition, has been observed by several workers

in a number of conjugated polymers, such as poly(acetylene), poly(pyrrole), poly(para-

phenylenevinylene), poly(aniline) [23], as well as in poly(3,4-ethylene dioxythiophene)

[16], in a relatively wide range of temperatures. Although metallic behavior has been

demonstrated for conjugated polymers, the truly metallic regime for which kFl > 1 has

not been achieved yet.

2.6 Doping Induced Properties in Conjugated Polymers

In a preceding section, it was described how the hand structure of DGSPs and

NDGSPs is modified upon doping. These structural changes can he observed through

experimental measurements, i.e., spectroscopic measurements or ESR experiments, per-

formed on conjugated polymers in different doping levels. For the case of DGSPs, there

are three important signatures of charged soliton formation. First, the generation of

localized structural distortions is associated with localized vibration modes (lattice vi-

brations or phonons). These characteristic vibration modes are known as infrared active

vibration, or IR AV, modes because they are active in the infrared region. These soli-

ton induced modes, which can he observed by spectroscopic measurements in the mid-










infrared frequency range, show intensities proportional to the doping level [24]. Second,

the electronic transitions associated with the generation of the localized mid-gap energy

state can also be observed by spectroscopic measurements, in the near infrared frequency

range. Finally, the charge storage in spinless solitons can he verified by electron-spin

resonance experiments.

For the case of NDGSPs, there are also significant signatures of charged polaron

and hipolaron formation. First, the generation of localized structural distortions which

is associated with phonon modes. These polaron and hipolaron induced IR AV modes are

active in the mid-infrared region, and can he observed by spectroscopic measurements,

as it will been seen later in this section. Second, the electronic transitions associated

with the generation of the two, symmetrically placed, mid-gap energy states can also

be observed by spectroscopic measurements in the near infrared frequency range. Fi-

nally, the charge storage initially in charged polarons with spin 1/2 and, as the doping

level increases, in charged, but spinless, hipolarons can he verified through electron-spin

resonance experiments. These experiments show a small signal at low doping levels

that grows as the doping increases and saturates at intermediate doping levels, consis-

tent with the formation of charged polarons with spin 1/2. At higher doping levels the

formation of hipolarons commence, and therefore the signal is gradually lost [12].

Figure 2.15 shows a schematic band diagram, and the absorption spectra of poly

(3,4-ethylene dioxythiophene), a NDGSP, in the neutral, slightly p-doped, and heavily

p-doped states. In the neutral state, only the x x*, or E,, transition is possible and

therefore, only one absorption hand appears in the spectrum. When the polymer is in

its slightly p-doped state, the x x* transition diminishes and, due to the formation of

positively charged polarons, two symmetric mid-gap states are generated. This results

in two additional absorption hands in the spectrum. Finally, upon heavy p-doping, the

created positively charged hipolarons, move the two symmetric mid-gap states closer

to the center of the hand gap and the HOMO and LUMO (the valence and conduction

hands) further apart. Therefore, the x xr* transition is total bleached and, based on the










of the doping method depends on the nature of the polymer, and the application is in-

tended for.


*'! CI. II... doping by charge transfer.

This method involves charge transfer redox chemistry, oxidation (p-type doping,

the system loses electrons) and reduction (n-type doping, the system receives elec-

trons), and produces conjugated polymers with high electrical conductivities. As

the chemical doping level increases, the electronic structure evolves to that of a

metal [6].

This method was the one used in the initial discovery of the ability to dope conju-

gated polymers by charge-transfer redox chemistry [1]. The oxidation, or p-doping,

of poly(aniline), PANI, was achieved by exposing the polymer to iodine vapors, and

the reduction, or n-doping, involved treatment with sodium naphthalenide. h-

nlical doping can also be achieved with protonation by acid-base chemistry. This

type of doping leads to an internal redox reaction, and for the case of PANI to the

conversion of enteraldine base, a semiconductor, to enteraldine salt, a metal [5].

The charge transfer redox chemistry, oxidation and reduction, is illustrated in the

followingf examples:

(CP), + -nxrl2 .r [(CP .JI), (2.7)


for oxidation, or p-type doping, and


(CP),z + [Na+(Naphthalide-)] [(Na ),(CP)'"],z + (-\ph II,~ .I ldeo) (2.8)



for reduction, or n-type doping.

Materials produced hv chemical doping have very high electrical conductivities,

and can he used in applications as transparent electrodes, antistatics, electrontag-

netic interference (EMI) shielding, and intrinsic conducting fibers.









*Electrochemical doping.

This method of doping is the one used throughout this work. Although chemical

doping is an efficient process that can yield fully doped and high quality materials,

it has a n, l ini- disadvantage: there is no way to control the d..ph.lr and obtain in-

termediate doping levels. Attempts to reach such levels resulted in inhomogeneous

doped films. Electrochemical doping came as a solution to this problem [26]. In this

case, an electrode supplies charge to the conjugated polymer in the redox process,

while ions from a supported electrolyte diffuse into (or out of) the polymer chain

for electronic charge compensation. The doping level is determined by the "cell

vol .g ,which is the potential difference between the conducting polymer on the

working electrode (WE) and the counter electrode (CE). Homogeneous doping can

be achieved at any intermediate level by simply adjusting the "cell vcoll I,. and

wait for the system to reach equilibrium, indicated by the current through the cell

approaching zero. In our experiments, we required that the current had dropped

to about 1 of its peak value to consider that the doping level had stabilized.

Electrochemical doping is illustrated in the following examples:


(CP),, T+[i*F/]slultion i [(CP)+"(BF,),]n, + Lielcctrode (2.9)



for oxidation, or p-type doping, and



(CP),, + Lielectr-o de [(Lil),(CP)-m.,, + [Li (B3FJ)]solution (2.10)


for reduction, or n-type doping.

Materials produced by electrochemical doping can be used in electrochemical bat-

teries for charge storage, light emitting electrochemical cells, and electrochromic

applications, e.g., -in! I.t wind~i- optical switches, and low energy di pE.~--s










In both methods described above, chemical and electrochemical doping, the in-

duced electrical structure is permanent until the system is purposely "un-doped", i.e.,

the charge is removed or chemically compensated.


Photodoping.

Using this method, the semiconducting polymer is locally oxidized (hole creation)

and reduced (electron creation) by photoabsorption, which leads to charge carrier

separation. The created electron-hole pairs are separated into "free" carriers:


(CP), + he [(CP)+ (CP)- ],z (2.11)



where x is the number of electron-hole pairs. This number depends upon the

competition of the pump rate with the recombination rate. The "photoconduc-

tivity" lasts only until the excitations are either trapped, or have d.l i\-- 1 back

to the ground state. Following the photoexcitation from the ground state to the

lowest excited state with the proper symmetry, the recombination or decay of

an electron-hole pair to the ground state can he either radiative (luminescence) or

non-radiative. Some conjugated polymers, e.g., PPV, and PPP, show luminescence

with high quantum efficiencies while others, e.g., PAc and Pth, do not [9].

Photodoping by photoexcitation produces high-performance optical materials, which

are suitable for photovoltaic devices, and also provides a route for materials with

tunable nonlinear optical (NLO) response for electro-optic and optical devices,

e.g., waveguides.

C'I! I ge injection at a metal-semiconducting (j S) polymer interface without counter

10118.

Electrons and holes can he injected into an empty xr* (HOMO or conduction hand),

and a filled xr (LUMO or valence hand) bands respectively from metallic contacts:


(2.12)


[(CP),z ,e-], [(CP) "],










and

([(CP),z +.e-], [(CP)-"],. (2.13)


This method is fundamentally different from chemical and electrochemical doping,

because there are no counterions introduced in the system, although the poly-

mer becomes oxidized or reduced. In the case of charge injection at a metal-

semiconductor interface, electrons reside in the xr* hand, and/or holes at the xr

hand only as long as biasing voltage is applied. Then the injected electrons and

holes recombine with the emission of radiation (electroluminescence).

This electron or hole injection at a metal-semiconducting polymer interface is

particular useful for applications such as organic field effect transistors (FET's),

and light emitting diodes (LED's) [27, 28].


2.8 Fundamentals of Electrochromism

The electrochromic effect is defined as the ability of some materials to reversibly

modify their electronic structure upon doping induced by the application of an external

voltage. This leads to a change in the optical absorption spectra, and therefore a change

in the color of the material. The term "le'lectrochromism" was -II_a---- -1. by J. R. Platt

in 1960 [29] in analogy to the effects of I1,, a un n 1.1~omism" which describes the change

in color upon the application of different temperatures, and "photochroun!!-in which

describes the change in color produced by light.

Electrochromic materials change color in a reversible way by an electrochemical

reaction, and they can he classified into three categories based on their electronically

accessible optical states, or simply stated, based on their capability to access different

colors. The first category includes materials that have one colored state, i.e., in this state

the material absorbs within the visible region, and one transparent or bleached state,

i.e., in this state the absorption is outside the visible region. This class of materials is

mostly used in transmissive/absorptive type of devices. The second category includes

materials that have two distinctive colored states, i.e., both states absorb within the










visible region, and the last category includes materials that are able to access multiple

states, or multiple colors. These latter materials are often called poly-electrochromic, or

it is said that they exhibit multi-color electrochromism. The last two classes of materials

are used in reflective type of devices.

Many different materials, inorganic and organic, exhibit electrochromism. Among

them are the inorganic transition metal oxide systems, and especially the high hand

gap semiconductor tungsten oxide, WO:3, which, has been the main focus of research

for the last three decades, [30-34]. Tungsten oxide, upon reduction, changes from being

transparent in the visible region to having a blue color:



WO:3 + rM+ + re- M,WO:3. (2.14)


The fabrication of thin films of, amorphous or p.II~i- I i-- .11;!. WO:3 requires sputtering

under high vacuum, which is a complicated and expensive process. This high manu-

facturing cost, in addition to other reasons such as the very long lifetime requirements,

as well as the insufficiently fast response times, proclaims that there is still work that

needs to be done.

Other inorganic materials that exhibit electrochromism are mixed oxides of vana-

dium (V), molybdenum (Mo), niohium (Nh), titanium (Ti), nickel (Ni), cobalt (Co), and

iridium (Ir), phthalocyanine metal complexes, and also transition metal ]!.:: .. i. l1..ilso~-

tallates, such as Prussian Blue (PB). Prussian Blue, an example of poly-electrochromic

material, has a blue color in its is II us .!" form, and upon reduction it becomes trans-

parent, the so-called Prussian White (PW) but also known as Everitt's salt:



[FelllFell(CN),]- +e-, [FellFell(CN),]-2, (2.15)


while partial oxidation of Prussian Blue results in Prussian Green (PG), which, as indi-

cated by the name, has a green color:











3[FelllFell(CN)6] i [FIIFeI CN)6iae 2 6C1)6]- + 2e-, (2.16)

and further oxidation yields Prussian Brown, which has a yellow-golden color:



[FelllFell(CN)6]- [F IFeI C1N6] + e-. (2.1'7)

Another family of materials that exhibits the electrochromic effect includes small

organic molecules, such as bipyridilium salts, otherwise known as viologens. The most

well known viologens are the 1, 1'-dimethyl-4, 4'-bipyridilium or otherwise called methyl

viologen (ilV) [32], and the 1,1l'-di-n-heptyl-4, 4'-bipyridilium, also known as heptyl

viologen [35, 36]. This type of materials upon reduction undergoes a change from a

transparent to a colored state, with the colored state depending on suitable choices of

nitrogen, or alkyl, or other group substitutions. Until recently, the only widespread elec-

trochromic commercial application is the automatic rear-view dimming mirror system by

Gentex, called Night Vision Safety (NVS), which utilizes solution-phase electrochromic

viologfens. Despite the success of this reflective device, the development of other elec-

trochromic systems, e.g., H rtwn- for buildings, have not shown the expected

breakthrough in the market yet.

Conjugated polymers are the third family of electrochromic materials, and the

one that has gained a lot of attention in the recent years. Although not as developed

as the materials discussed earlier in this section, their popularity is basically due to

the fact that conjugated polymers are easier to process than inorganic electrochromic

materials, and they offer the ill I i ~r advantage of color tunability. More specifically, their

color can be tailored through structural modification of the repeat unit, i.e., monomer

functionalization and copolymerization, or through the use of blends, laminates, and

composites. Hence, one can find conjugated polymers in any of the three categories

mentioned above for the use in transmissive/absorptive or reflective type of devices, and

in a wide selection of colors. Conjugated polymers also promise rapid response times,










high contrast ratios, and long lifetimes, which are very important characteristics for

the development of commercial electrochromic devices. Of all conjugated polymers the

derivatives of poly(pyrrole) (PPy), poly(aniline) (PANI), and poly(thiophene) (PTh)

are the most widely studied [32, 33, 36-38].

The fact that electrochromic materials have been intensively studied the last

decades, and lately have been emploi- II in numerous applications, forced the definition of

electrochromism to be modified in order to fit within the demands of the modern world.

Therefore, although previously electrochromism was the reversible and visible change of

the color associated with the reduction-oxidation, or doping/de-doping, process of an

electrochromic material, this definition has been extended to include a wider spectral

range modulation. This spectral range now covers ultraviolet (UV), visible (Vis), near

infrared (NIR), mid infrared (illRt), far infrared (FIR), and microwave (\!W) regions.

In the case of these regions, "(0 .1. corresponds to the response of the detectors at the

different wavelengths, and can be studied as the change in the transmittance, and/or

reflectance, induced by doping or de-doping of the electrochromic material.

2.9 Synthesis Methods of CPs

The synthesis of conjugated polymers by oxidation methods can be generally di-

vided into two main classes, chemical and electrochemical polymerization. In the che-

mical case, the basic mechanism can be simply described as follows. A radical ion is

generated from a monomer molecule with the use of a chemical oxidant, and then cou-

ples to another monomer molecule, which results on the generation of a dimer radical

ion. This coupling between the radical ion and the monomer molecule occurs due to

the abundance of monomer molecules in the bulk of the reaction environment. These

reactions propagate until the completion of the polymer chain.

In the electrochemical case, the polymerization is also initiated by the generation

of a radical ion, which occurs at an electrode surface by oxidation via an applied electric

potential. Due to the fact that in the electrode vicinity, where the reactions take place,

the concentration of radical ions is large, radical-radical coupling occurs and this, leads to










the formation of a dication. The dication upon proton elimination generates a neutral

dimer, which oxidizes to a radical cation dimer. These reactions progress until the

completion of the polymer chain.

Electrochemical polymerization is the method that has been used throughout this

work because it provides a quick and easy way for the deposition of CP films on vari-

ous substrates. This process employs a three-electrode configuration that consists of a

working electrode (WE), an auxiliary (counter) electrode (CE), and a reference electrode

(RE). The working electrode is the electrode where the polymerization of interest takes

place, and there is a number of different materials, solid or flexible substrates, that can

be used. The choices include indium-doped tin oxide (ITO) on glass or PET, PEDOT-

PSS on PET, and thin metal films on grids, when transmittance measurements need to

be performed, whereas, solid platinum (Pt) or gold (Au), and gold on Mylar, are used

when reflectance measurements need to be performed. The counter electrode provides

the required current to sustain the developing processes at the working electrode. In

order to ensure that the electrochemical reactions taking place on its surface are not

limiting, the area of the counter electrode needs to be larger, or at least similar, to the

area of the working electrode. A Pt Afla, which is a piece of Pt foil and Pt wire welded

together, is usually emploi- 4 as a counter electrode. Finally, the reference electrode

provides control of the applied electrical potential. There are several different types of

reference electrodes that are commercially available, such as saturated calomel, Hg2 2~

(SCE), Ag/AgCl saturated in K(C1, and Ag/Ag* electrodes. In addition to these stan-

dard references, quasi- or pseudo-references can be used, eg., silver wire. The latter

ones need to be calibrated every time before their use. For the polymerization, the WE,

CE, and RE electrodes are placed in a monomer solution, and they are connected to

a potentiostat/galvanostat. This three-electrode arrangement prevents large currents

from passing through the reference electrode, and change its potential.

The electrochemical deposition of the CP on a conductive surface of choice can

be achieved in several different owsi~. In the case where the applied electric potential










is held constant during the deposition, the method is called potentiostatic deposition

and it generally yields polymers with consistent morphology. If the current that passes

through the electrodes is constant, the method is called galvanostatic, which generally

yields polymers of poorer morphology when compared to the films produced by the

previous method. Both methods can use either the time or the charge as a parameter

for the termination of the deposition process. Control of charge is desirable when a film

of specific thickness is needed.

Cyclic voltaninetry (CV) is another method that can he used for CP deposition.

In this method, the potential is repeatedly cycled over a specified voltage range, while

the resulting current is measured. The obtained voltaninogrant is a display of current

density as a function of the applied voltage. The scan rate, expressed in niV/s, is a

dynamic parameter and can he changed in order for different types of reactions, fast or

slow, to be followed effectively. This method results in films of comparable quality to

the potentiostatic method.

2.10 Characterization Methods of Electrochromic CPs

1\any characterization methods have been developed in an effort to gain a deeper

understanding of the electrochrontic processes in CPs. A very important, and widely

used, characterization method is spectroelectrochentistry. This method provides a way

to probe the electronic structure of CPs through the optical changes occurring upon

doping. Hence, information about the energy hand gap, and the nxid-gap states created

upon oxidation or reduction, can he deduced by carefully studying the resulting spectra.

Spectroelectrochenlistry can he performed using the three electrode configuration,

described in the previous section, to initially characterize the conjugated polymer film

in monomer-free solution. In this method, different potential values are applied, and

after the system reaches an equilibrium, the absorption spectrum is measured. Figure

2.16 shows the spectra of a PEDOT film deposited on ITO/Glass substrate, at differ-

ent doping levels, from which the absorption of the nionomer-free solution, and of the

ITO/Glass substrate have been subtracted.










the hipolaron absorption is observed. These structural and optical changes are reversible

through de-doping (reduction) of the polymer system.

Spectroelectrochemistry measurements can also be performed in reflective or trans-

missive, electrochromic devices. For devices, the three electrode configuration is replaced

by a suitable holder. In order to apply a voltage across the devices, the counter and

the reference leads have to be connected to one another. Different potential values

are applied and, after the system reaches equilibrium, the absorption spectrum is mea-

sured. From the resulting spectra, useful information can he deduced concerning the

electrochromic mechanisms within the device.

Another important characterization method consists of a kinetics experiment,

which allows for measuring of the switching times between two extreme electrochromic

states, neutral and doped for the polymer films in solution or in devices. In this ex-

periment, a square-wave potential is applied at specified time intervals, while concur-

rently the absorhance at Ana,, is monitored, where Xna,,. is the wavelength of maximum

electrochromic contrast, and can he determined from the spectroelectrochemistry mea-

surements described above. Furthermore, the same experimental setup can he used to

perform a long-time redox switching stability, or lifetime, test. By applying the square-

wave potential over a long period of time, hundreds or even thousands of cveles, one

can study the degradation of the performance of the polymer film itself, or in the device

under investigation. The switching times and the lifetimes can vary significantly not

only between different polymers, but also for the same polymer, e.g., when different

dopants are used.

Other characterization methods include cyclic voltammetry; a technique that pro-

vides information about the potential values at which oxidation and reduction occur

for each polymer, in-situ conductivity measurements, and in-situ colorimetric analysis.

This latter method accurately defines the color, taken into account the sensitivity of the

human on; and the electrochromic contrast ratios in conjugated polymers. Although










these techniques provide useful information, they have not been used in the course of

this work.

2.11 Electrochromic Devices (ECDs) Based on CPs

Electrochromism is the ability of electrochromic materials to reversibly change

their transmittance, or reflectance, upon the application of an electric field. Therefore,

electrochromic devices which employ these type of materials, are used for the modulation

of light. An electrochromic device can he envisioned as an electrochemical cell, which

consists of two or more redox active materials separated by an electrolyte l~i-;r. Upon the

application of a small voltage, usually no more than a few volts, optical changes occur

due to electrochemical reactions that take place within the cell. The electrochromic

switching time of the device between two extrema states, e.g., colored and bleached, is

limited by the ion diffusion from one l~i-;r to another. Liquid electrolytes provide rapid

switching times due to high ion mobility, but cause faster degradation of the device

due to the risk of solvent evaporation or leakage. Gel and solid, electrolytes provide

slower switching times but they do not share the degradation problems with their liquid

counterparts, and therefore are the most commonly used in device applications.

There are two hasic categories of ECDs based on their function mode or, in other

words, the type of light modulation they perform. The first of these categories, is

the reflective ECDs in which the light incident to the device is controllably reflected.

The typical design of an ECD that operates in reflectance mode is an outward-facing

platform, and was originally developed by Bennett [39] and C'I 1...4 -lekhar [40]. More

specifically, the device consists of an outward-facing working electrode, usually gold

coated on a slitted Mylar or on a porous membrane, onto which an active electrochromic

CP is deposited. The porous membrane can he polycarbonate, polysulfome, polyester,

or nylon, and compared to the slitted Mylar it allows for faster and more uniform ion

diffusion [41, 42]. The counter electrode, usually gold on Mylar, does not contribute

directly to the modulation of light but the counter electrode EC polymer serves to

balance the charge. To assemble the device, the counter electrode is placed facing the










hack side of the working electrode, with a porous separator (polypropylene) soaked in

electrolyte in between them. A transparent window, placed on top of this configuration,

seals the device. This window should allow the proving of the optical properties of

the reflective ECD, and thus the choice of the material depends on which region of the

electromagnetic spectrum the device is needed to operate. Typical windows that are

used are puk.i~ I bylene (for almost over the entire spectrum from far infrared to visible),

zinc selenide (for the mid- and near infrared region), and glass (for the near infrared and

the visible regions).



transparent wi ndow
workig elctroe CPgold on slitted Mylar or
Porous membrane (WE)
counter electrode CP
porous separator soaked
in gel electrolyte
gold on Mylar (CE)
polyethylene substrate
(for support)


Figure 2.17: Schematic diagram of a reflective electrochromic device.


More details about the design of this type of devices are given in ChI Ilpter 5. An-

other type of fabrication of reflective ECDs includes patterning of the working electrode,

and use of more than one electrochromic CPs in the same device. This design provides

high resolution pixel devices in which each pixel can he addressed individually [38, 43].

The most common applications of reflective ECDs are dli ph although, there

is still a lot of work to be done for these devices to be widely commercial available.

Alany workers are putting a serious effort into making devices with faster switching

times, longer stability, and higher electrochromic contrast. Reflective ECDs based on

CPs offer many advantages compared to LCDs (Liquid Crystal Devices). They are easy

to process, have low production cost, low voltage requirements, offer a wide selection of

colors, and the quality of the image does not depend on the viewing angle. Furthermore,

reflective ECDs can he made flexible and, in any size and shape. Other applications










include mirror devices, optical shutters for thermal control in spacecraft, and battlefield

camouflagfe countermeasures against night vision sensors.

The second category of ECDs is the transmissive/absorptive device, in which the

incident light controllably passes through the device. The design of this type of ECDs

consists of two electrochromic CPs deposited on transparent electrodes which are facing

each other and are separated by an electrolyte 1.>. -r, as shown in Figure 2.18. The elec-

trochromic CPs emploi-. I1 in transmissive/absorptive ECDs, must he complimentary to

each other, a cathodically coloring polymer and an anodically coloring polymer, in order

for high contrast values to be achieved. A cathodically coloring polymer has a low band

gap, ideally around 1.8 2.2 eV, is colored in its neutral (undoped) state, and upon oxida-

tion it becomes transparent in the visible region. Whereas, an anodically coloring poly-

mer has a high hand gap, ideally higher than 3.0 eV, is transmissive in the visible region

when it is in its neutral state, and upon oxidation it absorbs the visible light. Therefore,

the ECD can he reversibly switched between a colored, absorptive, and a transmis-

sive, bleached, state. The conducting electrodes emploi-, .1 in these devices, depend on

the region of the electromagnetic spectrum the device is needed to operate. Typical

transparent electrodes for the visible region, consist of indium-doped tin oxide (ITO)

films deposited on glass, and ITO or poly(3,4-ethylene dioxythiophene)-poly(styrene

sulfomate) (PEDOT-PSS), deposited on poly(ethylene teraphthlalate) (PET) for flexible

ECDs. This type of devices can he used as -in! I.t wind ..\--- for cars, e.g., rear-view

mirrors, sunroofs, and for buildings, providing huge energy saving cost.

2.12 General Applications of CPs

Initially, this new class of polymers held the hope that will serve as replacements

of the existing heavy metals in weight sensitive applications, such as air and space

applications. However, due to instabilities of these systems under ambient conditions

alternative practical uses were exploited. The promise of of the so-called plI I-I !I" elec-

tronic devices attracted a lot of attention not only as basic research in academia but also












transparent wi ndow

work ng electrode C P conduct ng layer

counter electrode CP gel electrolyte

conducting layer
transparent wi ndow



Figure 2.18: Schematic diagram of a transmissive/absorptive electrochromic device.


from a commercial point of view. Although there are not many commercial applications

available yet, the future of plI 1- I w" electronics seems very promising.

In particular, conducting polymers exhibit interesting and important properties

not typically available in other materials. Their advantages over other materials, or-

ganic and inorganic, lay on their light weight, relatively inexpensive fabrication, easy

processibility and ability to be formed into odd shapes and sizes, low voltage require-

ments, compatibility with many organic liquid and solid electrolytes, and fast switching

color changes. In addition, a wide variety of colors is available with conjugated poly-

mers, as well as the ability of color tunability based for example on the doping level,

or the choice of dopant. These unique properties make possible a number of appli-

cations such as transparent electrodes, antistatics, electromagnetic interference (E1\l)

shielding, conducting fibers, electrochemical batteries, anti-corrosion coatings, sensors,

-! In~rt wind i--- optical switches, thermal control devices, low energy di-ptlli- light

emitting electrochemical cells, photovoltaic devices, organic field emission transistors,

and more [5, 17, 26, 27, 32, 3:3, 36, 38, 44].















CHAPTER 3
THIN FILM OPTICS

Optical experiments on thin films provide a valuable method of understanding the

properties of materials that are in a solid state form. The materials specific parameters,

directly available through these experiments are the frequency-dependent reflectance

R, or transmittance I, spectrum over a region of interest. This region does not refer

only to the relatively narrow visible region but extends and covers the whole range of

the electromagnetic spectrum from far infrared to ultra-violet. Fr-om the experimental

measurements, R and/or T, and calculations based on theoretical models, the dielectric

function can be deduced which is the property most directly related to the electronic

structure of the material under consideration. This way, the optical phenomena can be

quantified, and a number of optical constants can be derived describing the response of

the medium to light.

3.1 Optical Processes and Optical Constants

The basic optical phenomena that can be observed as light travels through matter

are reflection from a surface between two different media, and propagation within a

medium in the solid state form. The amount of light transmitted through a film relatively

to the amount of incident light depends on the way light propagates within each medium.

For example, if the frequency of the light wave is resonant with any of the transition

frequencies of the atoms in the medium, part of the light will be absorbed and thus, as

the light travels through the medium its intensity will be decreased. The intensity of

the transmitted light within the medium, is described by Beer's law:



I(z) = I(zo)e-az, (3.1)










where a~ is called the absorption coefficient and is a strong function of frequency, and zo

is the interface plane.

Another phenomenon that causes attenuation of the light intensity within a medium

is scattering. Scattering is responsible for re-directing the light to all possible directions.

Consequently, it diminishes the amount of light that proceeds in the forward direction

and therefore, the amount of light which transmits through the material. Thus, scat-

teringf has an effect similar to absorption, and the resulted attenuation of light can be

expressed in an analogous form as the equation above:



I(z) = I(zo)e-N"", (3.2)

where NV is the number of scattering centers per unit volume, and o- is the scattering

cross-section of the scattering center. A number of different factors can be responsible

for the scattering process, such as inhomogeneities, impurities, defects, etc.

Furthermore, the light waves travel within a medium with a smaller velocity com-

pared to the velocity of light in free space, and with a velocity that differs for different

materials. This phenomenon, called refraction, is described by the refractive index n.

The refractive index is, in general, a function of frequency and this effect is known as

dispersion. The refractive index n is defined as the ratio of the velocity of light in free

space c ( where c= 2.998x10sm s-l), to the velocity of light in a medium v:



a = c (3.3)


The velocity change causes the light rays to bent with respect to the normal on the

interface as the light goes through two different materials, and it is described by Snell's

law [45].

The two quantities that describe the propagation of the light wave within a

medium, the absorption coefficient a~ and the refraction index n, can be combined into

a single parameter called the complex refractive index:












ft = n + is, (3.4)


where the real part of the complex refractive index is the refractive index n, as defined

in equation (3.3), and the imaginary part n, is related to the absorption coefficient, as

it will be shown later. The imaginary part a is called the extinction coefficient.

In the above description of the phenomena that occur when light travels through

a medium, it has been assumed that the result of this interaction is independent of

the intensity of the light beam. Therefore the analysis is based on linear optics where

properties such as the refractive index n and the absorption coefficient a~ are taken

to be independent of the optical power of the source. This is indeed the case when

conventional light sources are used, and our research throughout this dissertation falls

into this domain. However, now that high power lasers are available, there are a number

of other phenomena that can occur as a light beam of high intensity propagates through a

medium. These are described by nonlinear optics. Nonlinear effects cause the properties

of the material under investigation to depend on the intensity of the light source. An

example of a nonlinear effect is the doubling of the frequency, where the frequency of

part of the beam doubles as it interacts with a medium.

3.2 Interaction of Electromagnetic Waves with Matter

The response of a material to an incident light beam can be described by the

followingf macroscopic vectors: the electric field E, the polarization P, the electric dis-

placement D, the magnetic field H, the magnetic flux density B, and the magnetization

M.

The application of an external electric field to an isotropic and homogeneous

medium tends to align the microscopic dipole moments within the atoms along the

external field direction. This produces a net dipole moment within the medium, and

therefore a polarization. More specifically, the polarization P is defined as the net di-










pole moment per unit volume, it is parallel to the applied electric field, and it can he

expressed as follows ]


(:3.5)


where X, is the electric susceptibility of the medium.

The electric displacement D of the medium is defined as


(:3.6)


(:3.7)


D = E + 47rP,


and hv combining the last two equations we can write



D = 2E


where



& ei + if2 = 1 71 e r


*38)


is the complex dielectric function, which is a very important parameter for the under-

standing of the interaction of light with matter.

The current density J, which comes as a response of the medium to the application

of E, is related to the electric field as follows


J = E


(:3.9)


where


& o-1 + io2.


(:3.10)


is the complex conductivity of the medium. In general, the current density is the result

of the contributions arising from the bound and free charges: J = Jbowed + free, where

1 CGS units are used throughout unless otherwise specified.


P = XE










Jbound represents the localized motion of the restricted charges, and Jfree represents

the charge carriers that are free to move within the medium. Combining the equations

above, we find a useful relationship between the dielectric constant and the conductivity:


4~r
2 = 1 + i-. (3.11)


The last equation can also be expressed as follows



"' =T (3.12)

and



4xr

Another useful expression that relates the complex dielectric function and the real

part of the conductivity is the following:


4xri
E = e (W) + -ar (Lo) (3.14)


where ex(w) is the real dielectric function, while al(w) is the real part of the complex

conductivity, also called optical conductivity. An interesting case is the zero frequency

limit of the above equation, which specifies the response of the medium to static fields.

In this limit, ez (w = 0) becomes the static dielectric constant, while er (w = 0) becomes

the dc electrical conductivity.

In an analogous way, as the application of an external electric field, the applica-

tion of an external magnetic field to an isotropic and homogeneous medium produces a

magnetization M to the medium, which is proportional to applied magnetic field H:



M = XEmH (3.15)


where Xm is the magnetic susceptibility of the medium.









The magnetic flux density B of the medium is defined as:


(3.16)


and by combining the last two equations we can write


(3.17)





(3.18)


where


is the complex magnetic permeability.

The interaction of a medium with an applied electromagnetic field, is described by

Maxwell's equation's:


V D = 4xrpy



Vx E =
c iit
4xr 1 8D
V x H = -Jf + ,


(3.19)

(3.20)

(3.21)

(3.22)


where pf is the free charge density, and Jf is the free current density.

In the absence of external charges, pf = 0, and currents, Jy

equations can be written as follows:


0, Maxwell's


(3.23)

(3.24)


V -D

V -B


B = H + 4xrM,


B = pH,





I- = 1 + 4~Xm












1 BB
V x E = (3.25)
c iBt
1 BD
Vx H =(3.26)
c dt


Eliminating D, and H from the above equations results in:




V -E = (3.27)

V -B = (3.28)
1 BB
Vx E =(3.29)
c dt
pe1 dE
Vx B = .(3.30)
c iit


Using the following vector identity:



V x (V x A) = V (V A) V2A (3.31)

we obtain the final result:


pe1 82E
V2E =(3.32)
c2 dt2

which has the form of a wave equation with velocity:


1 pe1 c
4 v =(3.33)
v12 C2

Comparing the last equation with equation (3.3), and taking into account the fact that

for optical frequencies the magnetic permeability can be set equal to one, p = 1, the

refractive index can be expressed as follows


n =


(3.34)










This equation allows us to relate the refractive index, a parameter associated with the

way an electromagnetic wave propagates within a medium, to the dielectric function,

which is a constant directly related to the electronic structure of the medium.

Based on the above analysis, the solutions to Maxwell's equations will have the

form of a plane wave of angular frequency w: 2



E(r, t) = Eoei(kr--wt) (.7




H(r, t) = Hoei(k-r-wt) (.8


where Eo, Ho are constant amplitudes, in general complex numbers, and k is the wave

vector, which in a non-absorbing medium is given by:






where A is the free space wavelength A = 2xrc/w.

In the most general case of an absorbing medium, k is a complex number and it

represents the energy dissipation as the wave propagates through the medium:



k = -n~ L (n + iin). (3.410)
c c:

Substituting equation (3.40) to equation (3.37), and taking E along the z direction, we

obtain



E(z, t) = Eoei(witz/c-wt) = Eoe-'Fws/c i(mnZ/c-Wt) (3.41)
SIt must be noted that the signs in the phase argument are chosen arbitrary, and do not change the
physics of the problem. Therefore, a solution of the form:

E(r, t) = Eoei("t-k'r) (3.35)
and

H(r, t) = Hoei(wt-k'r), (3.36)
could equally be used.










Thus, in the case of an absorbing medium, there is an extra term in the wave equation

which represents the exponential decay of the wave within the medium. If we take into

account the fact that the intensity of a light beam is proportional to the square of the

electric field I ac EE* [45-47], and we compare this with equation (3.1), it can be shown

that the extinction coefficient a is proportional to the absorption coefficient a~:


2com 4xx~
a =, (3.42)


where A is the free space wavelength. Combining equations (3.8), (3.34), and (3.42), we

can relate 2, it, a, and a to each other:








62 = 2ns (3.44)




n = x + e + )4(3.45)


= / ex + (- + e )1(346




This analysis proves that the refractive index and dielectric constant are not independent

parameters, and thus, if we know one of them we can calculate the other.

In the analysis above, only isotropic and homogeneous media were considered.

In the case of highly anisotropic media, the polarization and induced currents lie in a

different direction from that of the electric field of the electromagnetic wave. In this

situation, the dielectric function becomes a tensor quantity. The response of the media

is well characterized by this tensor, however if the direction of the electric field is not

along one of the principal directions of the dielectric tensor, the analysis can be rather

complicated.









3.3 Light Propagation Through a Planar Interface

In order to determine the reflected and transmitted light at a plane surface between

two semi-infinite media of different dielectric properties, we need to apply boundary

conditions to the solutions of Maxwell's equations. These boundary conditions require

the tangential components of both the electric and magnetic field to be continuous at

the interface that separates the different materials.

We consider a plane wave, incident on a surface at z = 0, as shown in Figure 3.1.

The amplitudes of the incident, reflected, and transmitted waves of the electric vectors

are the following 3




Eo~ =E cos cof: Eo si ,, ].)(.9

Eo ogi (3.50)

E = [Eo cos of,: + E~Tp Sin ,-,r :('- i )~J (3.51)

E = o gi( -- )(3.52)
lp lp sin ~l~]ei~w"-l2?nl" sin~p 2rn zcos~p )(.3
2xn' si 1~~ 2xns cos 1
El = E gje"(""-sx-[x ]) (3.54)


where k = 2xu/AX, and the subscripts p, a denote the transverse magnetic polarization,

also known as TM, and the transverse electric polarization, also known as TE, respec-

tively. The incident, reflected, and transmitted waves of the magnetic field vectors can

be calculated from:



( LD)H = -(k x E) (3.55)

SThe sign convention for Maxwell's solutions used in this section is of the following form:

E(r, t) = Eoei(wt-k-r) (3.47)


H(r, t) = Hoei("t-k'r)


(3.48)











E Op


E *S


E ,


nn
"1
E*._

E 1,s





Figure :3.1: Light heani propagates through an interface between two media with different
optical properties.

The total components of the electric and magnetic vectors at the surface, at x = 0, for

the incident, reflected, and transmitted waves are




Eo, (o,~ + Eo ) cos cpo(:.)

Eoy = Eo ,+ Eo (:3.57)

Ho., = no(-Eo + Eo ) cos cFo (:3.58)

Hon = no(Eo* Eo;) (:3.59)

El,> = E,+, cos cpl (:3.60)

El, = EL (:3. 61)

Hi,. = -n iEl cos cpl (:3.62)

HI, = n IE {, (:.6:3)


The boundary conditions require the tangential components of both the electric and

magnetic field to be continuous at the interface, at x = 0, thus:













Eox = El, (3.64)

Eow = El, (3.65)

How = His (3.66)

How = H1, (3.67)


Substituting the total components of the electric and magnetic vectors at the surface,

z = 0, for the incident, reflected, and transmitted waves into the latter equations, we

obtain the well known Fresnel coefficients for reflection and transmission:


E~p n0 COS 1p n1 COS 0~
r = (3.68)



Ef, 2no cos co
t, (3.69C)



Eas no cos co n1 cos cpl
rs = (3.70)
Eas, no cos co + nl cos cpl


E' 2no cos co
is (3.71)
Eas, no cos co + nl cos cpl

The reflectance R, defined as the ratio of the intensity of the reflected light to the

intensity of the incident light on the surface, and the transmittance T, defined as the

ratio of the intensity of the transmitted light to the intensity of the incident light on the

surface are given by:





R, = r R, =,+ (3.72)




















?,_0~ CO 17 81 COS
Ls =o s\ (3.76)








R, = ocsco-n o p (3.77)



n1 2no cos co
=, n ocscl+n o p (3.78)



n1 2no cos co
=, n ocsco+n o p (3.79)


Furthermore, in case of normal incidence, on an isotropic non-absorbing medium, the

above equations become



R, = R, = oo-nn (3.80)




I= 1 = .(3.81)
(no +n1)2'

The above equations can be used even when the angle of the incident light does

not meet the medium surface at exactly right angle. For example, comparing the above

equations for reflectance for normal incident and an arbitrary angle, we can calculate, 4

the magnitude of the error approximation. Using no = 1 and nl = 2 we find that for

4 Using Snell's law to calculate cpl: nl sin cpl = no sin cpo.










an angle co up to 5.50 the difference between the obtained reflectance values lies within

1 Therefore, the choice of the right equation can be made depending on the accuracy

that is needed to be used in a particular experiment.

In case of normal incidence, on an isotropic but absorbing medium, the refractive

index nl, in the above equations should be replaced by the complex refractive index

nil = nl+ist, in order for the absorption to be taken into account. Therefore, reflectance

and transmittance are given by:


(no n1)2 ~2
R, = R, = (3.82)
(no + ur)2 2 ~



7,=i (3.83)
(no + n )2 2 ~

For other than normal incidence in an absorbing medium the calculations for the

reflectance and transmittance become tedious, and approximations need to be used for

each case. More details on this subject can be found in the references [48-52].

Although in our experimental part we are considering the case of normal incidence,

we are not dealing with single surfaces but with multi-1 li-o 1. I structures, this fact makes

the problem a lot more complicated.

3.4 Light Propagation Through a Single Layer Structure

We consider a single homogeneous, isotropic, non-absorbing, 111-;-r of thickness d

intercalated within two semi-infinite, non-absorbing, 111-;- rs of different refractive indices.

As the incident light beam meets the interface, part of the beam is reflected off, and

the rest of the beam is transmitted through the interface. These phenomena occur

every time the light beam encounters an interface between two media of different optical

properties. Therefore, the reflectance of, and the transmittance through a single 111-- r

of thickness d are obtained by summing the multiple reflected and transmitted parts of

the beam from the two parallel interfaces that define the single 1.s. -r structure. There

are different methods that can he ulsed for the determination of the Fresnel coefficients






































n,


of reflection and transmission for a single 111-; r. One way to deal with this problem is to

sum the amplitudes of the successively reflected and transmitted beams in terms of the

Fresnel coefficients as were determined in the previous section. Although this method

works well for the case of a single 1... -r, it becomes very complicated for the case of

multiple 1 ... r~s, and thus another approach is going to be used for this problem, which

is to work with the vector sums of the waves.

We consider an incident light beam on a l ... r of refractive index nl between two

semi-infinite media of refractive indices no, and n2, aS Shown in Figure 3.2.


Figure 3.2: Reflection and transmission
index nl, where the incident beam meets


through a singfle-11s-;-r structure of refractive
the surface at almost right angle.


The electric and magnetic vectors of the incident, reflected, and transmitted waves at

the first interface, at z = 0, are the following:


Eo, = (Eo e-ikoz +- E e+ik"oz) COS 0

Eoy = Eo e-ikoz + Eoe eikoz


(384

(3.85)


Eo













Ho, = no(-Eo e-ikoz + Eo e+ikoz)nO COS 0p (3.86)

Ho, = no(Eo~ e-ikoz -n E~p+ikoz (.7

Hoy = Ro(Eo e-ikoz E~p +ikoz (.8

E1. = (E e-ikiz' + Ez efikiz) COS 1I (3.89)

E = Ez e-ikiz + Ez e+ikiz (3.90)

His = ut(-E +e-ikiz" + Ez e+ikiz") COS 1p (3.91)

Hly = ni1(El e-ikl-z Eie+ikl-z) (3.92)


and the electric and magnetic vectors of the incident, reflected, and transmitted waves

at the second interface, at z = d, are the following:



E 2 = E -i~kazCS 2 (3.93)


EGy = E2 -ikg z (3.94)

H,, = 2(-E2 -ikgz) COS 2p (3.95)

HLy 12E2 -ikgz.' (3.961)


The boundary conditions require the tangential components of both the electric and

magnetic field to be continuous at the interfaces, at z = 0 and z = d, thus we obtain




(Eo+ + E~p) COS 0p = (E,+ + E-,) cos cp (3.97)

(Eo Eo )no = (E;L E,)nl (3.98)

Eo~ + Eo = E,+ + E, (3.99)

(-Eo+ + Eol)no cos co = (-E,+ + E-,)nl cos cp (3.100)













(E -iki dr + Ee+ikidl ) COS 1p + E2 -ikg2dr O (3.101)
(E e-ikid] G- E[e ikidl),1 = E -ikg2d] (3.102)

Ez e-ikid] + eikidi = E -ikgdl (3.103)


(-Ez e-ikid] +- E eikid])n1 COS 1p = E2 -ikg2dl,2 COS 2. (3.104)


Using the Fresnel coefficients, as were defined in the previous section, the latter equations

can be written as follows



E+ = Ez + -E- (3.105)

E = ( 1 E 316



ri1
Ef -ikd = E+ -ikgd (3.107)
S121



E~-ikidl 2E2- -ikgd' (3.108)
1 a2




The suffixes p and a have been omitted in the above equations because it has been

shown that, the relations between the vectors polarized in the plane of incidence and

those polarized in the perpendicular plane are the same [49]. One thing that should

be kept in mind while using these equations is that the Fresnel coefficients themselves

depend on the type of polarization. Based on these relations, the amplitudes of the light

beam in each medium can be expressed in terms of the amplitude of the incident beam,

and hence the reflected and transmitted amplitudes can be determined by:


E- rl + r26-i2S1
R = (3.109)
Ef+ 1 + rlr26-i2s1











E2' 1 26-i
T =aln (3.110)
Eo* 1 +rlr26i21

where


2xr
61 nldl cos cpl (3.111)


is the phase change upon the light beam traversing the film once.

Thus, the reflectance and transmittance are given by:


r:: + 2rlr2 COS 261 + r,
71 = RR* = (3.112)
1 + 2rlr2 COS 261 + r~r,


n2 n2 t2 2
T = TT* = 12 31)
no no 1 +2rlr2 COS 261 +r

For the simple case of normal incidence, the Fresnel coefficients expressed in terms

of the refractive indices have the following compact form:


n0 n1
ri = (3.114)
no + n1


n1 n2
T2 = (3.115)
nit + 8


2no
1 = (3.116)
no + n1


2ni
t2 (3-117)
nl + n2

Substituting the above Fresnel coefficients in the expression for reflectance and transmit-

tance, we obtain the expressions for reflectance and transmittance at normal incidence

in terms of the refractive indices of the media. In case of absorbing media, the real










refractive indices should be replaced by the complex refractive indices in order for the

absorption to be taken into account.

3.5 Light Propagation Through a Multi-Layer Structure

Applying the method that was described for the single 1.v. -r structure to a double

1... -r structure, as the one shown in Figure 3.3 we obtain for the reflected and transmitted

amplitudes the following:


Eo- 1 2-i2S1 -3-i2(S1 +Sz) 7123-i2S1
R= = r a e rre
Eot 1 + rlr26-i2S1 + 1736-i2(S1+ 62)+rzr36-i262



T = "e
Eot 1 + r r26-i2s1 + 1736-i2(61+62)+rzr36-i262 )


where


2xr
6m =


nmdm cos m ,


(3.120)


is the phase change upon the light beam traversing the 1 ... rrm = 1, 2, or 3.

Thus, the reflectance and transmittance will be given by:



71 = RR*


(3.121)


and


T = 83TT*.
no


(3.122)


Another way to express the Fresnel coefficients rl, T2, and T3 is by using the

amplitude and the phase change of the light reflected at an interface. For example, the

Fresnel coefficient, known also as effective Fresnel coefficient, corresponding to the light

reflected at the back surface of the second film Figure 3.3 can be written in the form:


(3.118)




(3.119)

















111

d I E + EP II
En,





E ,+ E,-




E 3+ n3




Figure 3.3: Reflection and transmission through a double-l ... r structure of refractive
indices nl, and n2-




1 + r2736-i2S z(.13

The first film, with refractive index of nl and thickness dl, can then be regarded as lying

on top of the surface of a medium whose Fr-esnel coefficient is g_. ''-, and therefore we

can write


i ri + g_ '--i261
1efor (3.124)

If we now consider a system of m l ... rs, the effective Fresnel coefficient of the last one

can be expressed as:











-i2Sm
eme (3.125)
1 +rmrm+1e-

where Qm and Om are given by:


2 +1 + 2rmrm+l cos 25m
g,+ (3.126)
m, 1 T7~+ r + 2rmrm,+1 cos 25m

and



Om = rim (m (3.127)


where


rm+l sin 25m
tan rlm = (3.128)
rm+ m+l cos 25m


rmTm+ sin 25m
tan (m = (3.129)
1 + rmrm+l cos 25m

For the (m 1)th lIn-;-r the effective Fresnel coefficient can now be calculated from:



iOmm1 mlem
1 e m-1 -ie,-aS (3.130)

Thus, the amplitude and the phase of the reflected light by a multi-1 0 r structure can

be calculated from the Fresnel coefficients at each interface and the thicknesses of the

films, by simply repeating this process until all lIn-;-rs have been taken into account.

In the case of a multi-las-; r structure of absorbing media, the real refractive indices,

introduced through the Fresnel coefficients must be replaced by complex quantities in

order to take into account absorption. In this case, the problem becomes very compli-

cated. Several methods of approaching this problem have been developed, and there are

a few approximations that can be applied based on specific cases. For the analysis of

our experimental data, the matrix method has been used.










3.5.1 Matrix Method

For a multi-1.s. ri structure, the matrix method can be demonstrated as follows.

Starting with the boundary conditions, we can write





Ef_, (eiSm-1E1 + rmeiSm1Eg) /tm (3.131)
EL, = (rme-iSm-iE1+esmig/m (3.132)



where 6m = kmdm, and rm, im can be derive the same way as equations (3.114), and

(3.116). The above recurrence relations can be written in a matrix form:


m--l ,1 is l is l mm(3.133)

E- m -iam-1 e-iSa1 E-
m-1l m~ m

Fr a system with a Ins-,-,,rs we requir to know,, the relations between, E1, n E

which will allow us to obtain the transmission coefficient, and between the Eo- and E'0,

for the reflection coefficient. Based on equation (3.133), we can write



G ,- tt~.. t,+ l ~l I(3.134)

0~ I ..nl n+1

where E, = 0 since there is no negative-going wave in the (n + 1)th lIns-c, and



Cm= iSm- qiSm- (3.135)
Em -ibm- 6-ibm-


Therefore, E1 ,, and Eo- can be expressed in terms of E,+, and this is the way we

calculate the transmission and reflection coefficients. This method is described in more

detail in (I Ilpter 4 of reference [49].










3.6 Kramers-Kronig, or Dispersion, Relations

The K~ramers-K~ronig technique is based on the law of causality, which states that

the response of the medium can not precede the external cause, in this case the applied

electromagnetic field, and it is also hased in the application of complex analysis. Gen-

eral relations, known as K~ramers-K~ronigf relations can he derived between the real and

imaginary parts of a quantity. Thus, for the real and imaginary parts of the complex

refractive index, and the dielectric function it can he written











and











where P denotes the principal part of the integration. As can he seen from the latter

equations, the real and imaginary parts of a linear response function are not indepen-

dent with each other. Therefore, the K~ramers-K~ronig relations allows us, for example,

to calculate the refractive index from the absorption coefficient and vice versa. This

method provides us with a very important tool because we can perform one measure-

ment which will provide us, i.e., the frequency dependence of the optical absorption, and

then calculate the dispersion, without the need to perform a separate measurement.

It should be noted that, for a physical system the response function taken as an

example the dielectric function, should satisfy the following relation:











E(w) = *(-w) (3.140)

which for a dielectric system can be expressed as:



e (0)= er(-w)(3.141)



62 0> = -62~(--wo) (3.142)

and it requires the real part of the dielectric function er, to be an even function of the

frequency w, and the imaginary part of the dielectric function 62, to be an odd function

of the frequency w.

The most common route to determine the optical parameters of a system is to

perform reflectance measurements and make use of the K~ramers-K~ronigf relations. It

was shown in the previous section that the Fresnel reflective coefficient can be expressed

in terms of the amplitude and phase change as follows



R = p(w: '' ', (3.143)

where



p(Li) = ( (3.144)

and


Im [R]
tan 8 = (3.145)
Re[ R]

which for the simple case where the incident light travels in vacuum no = 1 before it

meets an absorbing medium nl = n + is, the phase change is related to n and a of the

medium as follows












tan 8 = .(3.146)
1 n2 K2

Equation (3.143) can be written as:



In R = In p(w) + i0(w) (3.147)


and since the reflectance should also obey the law of causality, we can use the K~ramers-

K~ronig relations to calculate the phase change dispersion:



OO OO
SIn R(w') In R(w) 1 w' + w dR("')

0 0
(3.148)

The determination of the phase change 0(Lc), enables us to obtain the refractive index

n and extinction coefficient a through equation (3.146).

A few drawbacks of the K~ramers-K~ronig technique are the facts that only a sin-

gle bounce is taken into account, and the requirement that the measurements should

be performed over the complete frequency spectral range 0 < w < 00, which is not

practically possible. In a real experiment, the measurements are performed over a wide

spectral range and then, extrapolations techniques are used for the frequencies beyond

the measured frequency interval. These extrapolation procedures should be performed

cautiously because they can result to false calculations of optical quantities, and there-

fore to false conclusions.

3.7 Models for the Determination of Optical Constants


3.7.1 Lorentz Model

The Lorentz model is emploi-x I for insulating solids, and it describes the motion

mechanism of the bound electrons upon the application of an external field, and other

excitations, such as lattice vibrations, also known as phonons. The bound electrons are










treated as if they were attached to the nucleus by springs and, upon the application of

an electric field, they are subject to harmonic motion:


d2r d
nz + r, na -eEi,,, (:3.149)
mdt2 Yl n~

where ni, e are the mass and the charge of the electron respectively, and r is the electron

displacement from the equilibrium position. The second term on the left hand side of the

above equation represents the damping term and provides for energy loss mechanisms,

which in the case of solids are various scattering processes. The third term on the left

hand side represents the harmonic restoring force (Hooke's law) with which the electron

is bound to the core. The term on the right hand side represents the driving force,

where E1,e is the local electric field acting on the electron. We assume the local electric

field varies in time as Ele(t) oc e-iet, and that the displacement r has the same time

dependence. Therefore the solution to the equation of motion is


el1
r (t)= 2 z,,(t).(:3.150)


The induced dipole moment is given by:


62
p~/ =2~, er =2E,..(:3.151)

Assuming that the displacement r is sufficiently small, the dipole moment can he written



j) = er = cKc(w)Ei,,, (:3.152)


where the frequency dependent quantity c&c is the atomic polarizability. Combining the

last two equations, we can write for the polarizability of one electron atom:


621










It is obvious from the above expression that, the polarizability is a complex quantity

due to the inclusion of the damping term, and thus it has a phase difference compared

to the local electric field at all frequencies.

In case there are NV atoms per unit volume, the macroscopic polarization is



P = NV < p >= NEc < El,< >= XE. (:3.154)


In order to relate the microscopic atomic polarizability to the macroscopic electric sus-

ceptibility, we should determine the relationship between the microscopic electric field

El,,,. and the macroscopic electric field E. In general, < El,,, >/ E because the local

electric field is the average over atomic sites and not over regions between sites. For sim-

plicity, in the case of hound electrons can he assumed that the two fields are equal since

the inclusion of the restoring force contains all the necessary features for the description

of the optical properties of the system. Therefore, it can he written



P = NEcE = XeE (:3.155)


from which X, can he deduced


NVe2
Xe 2(:3.156)

Since the dielectric function is related to the macroscopic electric susceptibility as ?

1 + 4rX~e, it can he expressed in terms of the damped harmonic oscillators as:


4xNC~2
2=1 +(:3.157)






2=1 +> (:3.158)


where eL, is the plasma frequency in Lorentz model, defined as:











4xrle2
02 = (3.159)


The real and imaginary parts of the dielectric function can also be written as:




4xrle2 2 _
et = r2 ~2 = 1 +o L (3.160)
m ("0 L02 2 y2Lo2

and


4xrle22
62 = 2ms (3.161)
m ("0 L02 2 + 2Lo2

In case there are more than one characteristic resonant frequencies due to the

oscillation of the bound electron within the atom and to the lattice vibrations, or there

are more than one electrons per atom, the dielectric function is classically expressed as:


4xe2
S= 1 + (3.162)


where



Ny = N. (3.163)


Nyj and Lcl are the density of bound electrons and their resonance frequency respectively.

Quantum mechanically, the dielectric function is expressed as:


4xre2 If
E = + -2 '(3.164)


The wj represents the transition frequency of an electron between two atomic states of

energy difference AE = kycl, and the parameter fj is called oscillator strength and is a

measure of the relative probability of a quantum mechanical transition. In case of free









atoms with Z number of electrons, fj should obey the following oscillator strength sum

rule:



Cs = Z (3.16i5)

In this analysis, it was assumed that the electrons are in vacuum, where em = 1.

em is the contribution from the high frequency absorption, beyond the measured range.

However, in solid state em / 1 and thus in this case, the first term in equation (3.162),

or (3.164), should be replaced by em.

3.7.2 Drude Model

The Drude model is applicable to free charge carriers, free electrons in metals, and

it is obtained from the same equation of motion used in Lorentz model by simply setting

the restoring force term equal to zero:



F = kr = 0 (3.166)


where k = mw, is the spring constant which is chosen so that wo coincides with the

natural frequency of an atom. Therefore, the equation of motion for the Drude model is





m* +rt m* =1 -~t eE j m* + -v = -eEl (3.167)


where m* is the effective electron mass, and the damping constant y has been replaced

by 1/7r. The relaxation time, -r, characterizes the energy loss due to scattering in a way

similar to y. More specifically, -r is associated with collisions between the free charge

carriers and impurities, lattice vibrational phonons, or other scattering centers in metals.

Assuming that the velocity varies in time as vioc oc e-ist, the solution to the equation

of motion has the form:











e-r 1
v = E. (:3.168)
nz* 1 iwr

In case of NV free electrons per unit volume, the current density can he expressed as

follows



J = Nei)=E (:3.169)
nz* 1 iwr

From this last equation and equation (:3.9), we obtain the AC conductivity:



7D (0)= = *1 i (:3.170)

where


NCe2-
ao = (:3.171)


is called the DC conductivity, and it is the zero frequency limit of the Drude conductivity.

The real and imaginary parts of the conductivity in the Drude model are



aDI (:3.172)
1 + W2,r2




1 + W2r2'

The dielectric function in Drude model is given by:






where wDp is the plasma frequency in Drude model, defined as:


4;rle2
m (:3.175)









and typically lies in the visible or ultraviolet spectral region. The real and imaginary

parts of the dielectric function can be obtained either from equations (3.12) and (3.13)

and equations (3.172) and (3.173), or from equations (3.160) and (3.161) by simply

setting Lo' = 0, Y = 1/7, and m = m*:


eD1 ,2 (3.176)
1 + 2,r2



ED2 = 3-177)

In the limit of low frequency, where



w < 7(3.178)


the dielectric function from equation (3.174), can be written as


4xco~ 1 4xcor i
Du)=1--1 -- Lo (3.179)


thus, the real and imaginary parts of the dielectric function, are



ED =1- 4607- (3.180)



40ra
ED2 (3-181)

For sufficiently low frequencies w < 7- -, we have |el| < |62|, and therefore, equations

(3.45) and (3.46) will give


n = it. (3.182)


Using equations (3.41), (3.42), (3.180), (3.181), and (3.182) an expression for the skin,

or penetration, depth can be written as follows











2 c
5 (3.183)


The penetration depth provides a measure of the decay of the electric field within the

medium, and as can be deduce from the above equation, the higher the de conductivity

of the medium the shorter the penetration depth of the AC field at a frequency w.

3.7.3 Drude-Lorentz Model

The Drude-Lorentz model is used to describe the optical properties of materials

under investigation. The Lorentz model is emploi-, a for the bound carrier interband

transitions or the lattice vibrations, whereas the Drude model is used for the free carrier

intraband transitions. Thus, the total dielectric function can be expressed as follows






where em is the contribution from the high frequency absorption, beyond the measured

range, in vacuum em = 1, ED is the Drude dielectric function given by equation (3.174),

and 2L is the Lorentz dielectric function given by equation (3.162). Therefore, the

dielectric function can be expressed as:






This expression, obtained by combining the Drude and Lorentz models is used to fit the

experimental reflectance data and provides an alternative to the K~ramers-K~ronig method

for extracting the optical properties. The advantage of the Drude-Lorentz technique over

the K~ramers-K~ronig technique is the fact that it does not require the data to be obtained

over a wide spectral range. The fittingf procedure can be performed over a finite spectral

range, and no extrapolation processes are needed.

Another 1! r ~ ~ advantage of the Drude-Lorentz technique over the K~ramers-K~ronigf

is the fact that, this technique can be emploi-, II for the analysis of thin films and multi-









1 li-o 1. I structures. A separate parameter file is used for each lI .> .r, and E(w) is obtained

in terms of co,, and a set of three parameters, Lcl, Lcpy, and yj for every absorption

band. Then, from the dielectric function, the refractive index can be calculated through

equation 3.34, and finally, from the refractive index and the thickness of each lI .v.r, the
reflectance and transmittance of the maiilr'1 li-. 1r I structure can be determined.

More detailed information about the Drude and Lorentz models can be found in

the references [48, 53].

3.7.4 Sum Rules

Sum rules provide mainly a useful guide for the interpretation of the experimental

results, as well as a way of checking the validity of the experimental data. A sum rule

has already been introduced in a previous section, equation (3.165), but when we deal

with solids there are a few sum rules that are more important. The following sum

rule, expressed by the real part of the conductivity, states that the absorption energy

at all frequencies is constant. For the quantum mechanic Lorentz model, the oscillator

strength sum rule, or f-sum rule can be expressed



Ui~ldu = ~f3 316

with Lc given by equation (3.159), while for the Drude model the f-sum rule can be

expressed






Sum rules can frequently be found to be expressed in terms of an effective number

of electrons per atom NVeff, contributing to the optical properties over a finite spectral

range, as shown for example in reference [48].















CHAPTER 4
INSTRUMENTATION AND EXPERIMENTAL TECHNIQUES

This chapter describes the experimental equipment and the techniques used to

handle the materials, to grow the films, to monitor their stability, and to perform mea-

surements of absorbance, reflectance, and/or transmittance at near normal incidence

over a wide frequency range, from 20 cm-l to 40,000 cm- or equivalently 2.5 meV to

5 eV. In order to cover this entire region, various types of light sources, detectors, grat-

ings, beam-splitters, filters, and different types of spectrometers must be used. Narrow

regions of the electromagnetic spectrum are measured separately, using the appropri-

ate optical components, and then merged together to form the spectrum of the entire

frequency range of interest.

The following different spectrometers are used to measure the optical properties

of our samples:

StellarNet Photo Diode Array spectrometer (6250 cm-l to 52,600 cm- or equiva-

lently 0.77 eV to 6.5 eV),

Varian Cary 500 spectrometer (3030 cm-l to 57,100 cm- or equivalently 0.37 eV

to 7.0 eV),

Zeiss 800 MPM microscope photometer (4000 cm-l to 40,000 cm- or equivalently

0.52 eV to 5.0 eV),

Modified Perkin-Elmer 16U monochromator (1000 cm-l to 40,000 cm- or equi-

valently 0. 12 meV to 5.0 eV), and

Brucker 113v fast scan Fourier Transform Interferometer, or FT-IR (20 cm-1 to

5500 cm- or equivalently 2.5 meV to 0.68 eV).

A more detailed description of these spectrometers will follow later in this chapter.










4.1 Dry Box

Initially, all the samples were handled in ambient conditions. However, in-situ

reflectance and transmittance measurements showed two strong absorption bands, one

was coming from water (O-H stretching mode at 3570 cm l, or equivalently 2.8 pm) and

the other one was coming from C-H stretching (at around 3000cm -, or equivalently

3.3 plm), that hindered the performance of the samples in the mid-infrared region. In

order to overcome this problem, all the materials were handled, and the samples were

prepared in dry conditions. For this reason a dry box (VAC HE-series Dri Lab) was used,

shown in Figure 4.1. This system provides a working area, hermetically sealed from the

ambient environment, consisting of an inert atmosphere nearly free of moisture, oxygen,

and sometimes nitrogen if desired. The principal gas, which is the one used in our labs

too, is argon (Ar). Hence, helium (He), nitrogen (N2), Or any COmbination of the above

gases can also be used. An antechamber mounted on the side of the dry box is used

for passing materials in and out without disturbing the inert atmosphere of the box.

Any time the antechamber is exposed to ambient atmosphere for the insertion of articles

in the dry box there is a specific procedure that should be followed. This procedure

includes three cycles of evacuation and filling of the antechamber with inert gas before

any of the articles can be safely brought inside the dry box, without contaminating the

inert atmosphere. Finally, two butyl rubber gloves, mounted in the full view window,

provide easy access to the working area.

A very important part of this system is the chemical purifier mounted on the back

side of the dry box, shown in Figure 4.2. The gas within the box circulates continuously

through the purifier. The purifier canister contains a moisture absorbent (molecular

sieves) and an oxygen reducing agent Q1, which is a material consisting of finely divided

copper on an Alumina Matrix (made by Dow C'I. in... I1 Co~lop owr:) [54]. As the gas

from the glove box passes through the purifier, the absorbent removes the water vapor,

and the oxygen reducing agent removes the oxygen before the gas enters the controlled

atmosphere of the dry box again. The oxygen reducing agent, copper, combines che-


































Figure 4.1: Schematic diagram of Dry or Glove Box.


mically with the oxygen forming cuprous or cupric oxide. When the agent is saturated

with oxygen, the purifier is regenerated with the use of a special regeneration/forming

gas that contains :3 to 10 per cent hydrogen. However, it should be noted that there are

certain chemicals, like sulfur and sulfur compounds, such as H2S, SO2, SO3, etc., that

will poison the reactant material and they should not he used in the dry box. If their

use is necessary, a suitable trap should be installed in advance in the circulating path

to prevent these chemicals from getting into the purifier.

Other parts that can he mounted on the dry box are an oxygen analyzer (model

AO-:316-C, VAC, mounted on the dry box in our laboratory in the Physics department)

to monitor the trace oxygen in the controlled atmosphere, and a pedatrol (model HE-

6:3-P, VAC, mounted on the dry box in our laboratory in the Chemistry department) to

provide both automatic and manual pressure control of the atmosphere inside the dry

box. Due to the absence of an oxygen analyzer mounted on the dry box in the C!. ~!!I s-1y

laboratory, other methods had to be emploi-. I1 for oxygen and moisture testing of the

inert atmosphere. Since it is recommended that the atmosphere inside the dry box














~FLOWMETER
SPURGE CIRCULATION
Jg INLET
R) BLOEr.
.. REGENERRTION
I GRS-6 TO 10%

: 1 75 LBS MOLECULARSIV

5 LBS Q1 m PURIFIER



;I 4 LBS-- MOEUA IV


VRCUUM RN
DRI-LRB
VENT
LEGEND
S HRND VALVE
B SOLENOID VRLVE



Figure 4.2: Schematic diagram of the chemical purifier and the gas flow through the
chemical circulation.


he continuously monitored, a bottle of diethylzinc and a bottle of horon trifluoride

diethyletherate are kept inside at all times. Diethylzine is used for oxygen detection

and horon trifluoride diethyletherate is used for water detection. To perform the test,

the bottles should be uncovered and in case smoke is observed, then the atmosphere is

contaminated and a regeneration of the purifier is needed; in case smoke is not observed,

then the levels of oxygen and/or water are within the acceptable limits. However, it

should be noted that this test only provides an indication of the condition of the dry

box atmosphere.

4.2 Electrochemical IVethods


For the study and characterization of chemical systems, electrochemical methods

have been used throughout this work. Their application requires an understanding of the










fundamental principles of electrode reactions, and the electrical properties of electrode-

solution interfaces. The electrochemical measurements provide information about the

chemical changes caused by the passage of an electric current, and the production of

electrical energy by chemical reactions [55]. In the electrochemical systems studied in

this work, the processes that affect the transport of charge across the interface he-

tween chemical phases, for example, between an electrode (electronic conductor), and

an electrolyte (ionic conductor), are of main concern. The charge transport through the

electrode occurs by the movement of electrons or holes, whereas in the electrolyte phase

it occurs by the movement of ions. There are several different types of materials that

can he used as electrodes including solid metals (e.g., Pt, Au), liquid metals (e.g., Hg),

and semiconductors (e.g., ITO, Si). For the electrolytes, the most frequently used are

liquid solutions containing ionic species such as H+, Na+, Cl-, Li+, C104- in either

aqueous or non-aqueous solvent. In order for the electrochemical experiment to take

place, the solvent/electrolyte system must have sufficiently low resistance.

An important parameter in the electrochemical experiments is the measurement

and control of the cell potential, which is the difference in electric potential between the

electrodes in an electrochemical cell.l The magnitude of the potential difference at an

interface determines the direction and the rate of charge transfer [55].

4.2.1 Electrochemical Polymerization

All the electrochemical experiments have been performed inside the dry box using

an EG&PAR model 273A potentiostat/galvanostat. For the film deposition, a three

electrode configuration has been used, shown in Figure (4.3), that consisted of a glass

container, a working electrode (WE), an auxiliary (counter) electrode (CE), and a refer-

ence electrode (R E). The working electrode is the electrode where the polymerization of

interest takes place. For this work, a number of different working electrodes have been

used, such as Gold/ \!ylar, Gold/K~apton "~, ITO/Glass, and SWNTs/PET. A Pt flag,2

L Electrochentical cell: two or liore electrodes separated by all electrolyte phase.
SPt flag: a piece of Pt foil, with a relatively large area collpared to the WE, and a Pt wire welded
together.










used as the counter electrode, provides the required current to sustain the developing

processes at the working electrode. This arrangement prevents large currents from pass-

ing through the reference electrode and change its potential. There are several different

types of reference electrodes [56] that are commercially available, such as saturated

calomel (Hg2 2~) eleCOtOde (SCE), Ag/AgCl saturated in K(Cl electrode, and Ag/Ag*

electrode. However, for use in the dry box the choices are limited. In our experiments,

a silver wire (Ag) has been used as a pseudo-reference electrode. Fr-equent calibrations

have been performed versus Fc/Fc+ because the potential of the pseudo-reference elec-

trode changes with time, and sample solution deposition. The electrochemical polymer-

ization of the films has been carried out in 0.2 M electrolyte solution containing 10 mM

monomer, that is added in the glass container until a big portion of all the electrodes

is bathed in the solution, as shown in Figure 4.3. Finally, the electrodes are connected

to the potentiostat by either alligator clips, or micro clips since the passing currents are

small.
CE WE RE
















Figure 4.3: Schematic diagram of three-electrode configuration cell.



4.2.2 Cyclic Voltammetry (CV)

Cyclic voltammetry (CV) is one of several methods [57] used to characterize con-

ducting polymers. It is a simple and valuable technique that provides both quantitative

and qualitative information about the system under study. In this method, the current










density at the working electrode is measured as a function of the applied potential. 1\ore

specifically, the potential is repeatedly cycled over a specified voltage range while the re-

sulting current is measured. The obtained voltammogram is a display of current density

as a function of the applied potential. The scan rate expressed in mV/s is a dynamic pa-

rameter and can he changed accordingly in order for different types of reactions, fast or

slow, to be followed effectively. This cyclic voltammogram reveals information regarding

the electrochemical potential at which the oxidation and reduction processes occur for

each polymer, the degree of reversibility of the electrode reaction, and the electrochemi-

cal stability of electroactive species through repeated cycles. Important parameters of a

polymer cyclic voltammogram are the scan rate dependence of the anodic and cathodic

peak currents, Ir>, and I,><, the anodic and cathodic peak potentials, E,, and E,><, and

the reversibility of the potential wave.

4.3 Optical Methods

For the experiments that are going to be presented in this work, reflectance and/or

transmittance (in or out of solution) measurements had to be taken over a wide frequency

range, from 20 cm- to 45,000 cm l. To perform these experiments, various types of light

sources, detectors, gratings, heam-splitters, filters, and even different types of spectrome-

ters had to be used. Several regions of the electromagnetic spectrum have been measured

separately, using the appropriate optical components, and then merged together to form

the spectrum of the entire frequency range of interest.

4.3.1 Spectroelectrochemistry

Spectroelectrochemistry pIIl i- a key role in proving the optical changes that occur

upon doping and de-doping of a conducting polymer. It is an important method that

can also provide information about the electronic structure of the material, such as the

hand gap and the intraband states created upon doping.

Spectroelectrochemical experiments were performed using a StellarNet Photo Diode

Array (PDA) UV-Vis-NIR spectrophotometer (made by St. 11 I~ J. 1 Inc.) located inside










the dry box, and/or a ITV/Vis-NIR Varian Cary 500 spectrophotonieter (Varian Optical

Spectroscopy Instruments) located on the bench top.

The StellarNet spectrophotonieter (EPP2000C ITV-Vis and EPP2000 NIR In-

GaAs) is a portable fiber optic instrument used for absorhance/transmittance nica-

surenients in the ITV-Vis-NIR range. This spectrophotonieter utilizes a special concave

hologfraphic grating for abberation correction in order to provide better imaging. With-

out mirrors, the grating maintains low stray light, and the instrument losses are mini-

nmized. In addition, the absence of moving parts makes the instrument more stable and

very reliable. This system is equipped with a linear Photo Diode Array detector and a

tungsten halogen (tungsten krypton bulb) source. The spectrometer covers 190 890 nm

for the ITV-Vis region, and 900 1600 nm for the NIR region.

The ITV/Vis -NIR Varian Cary 500 spectrophotonieter is a double beam niono-

chrontator that covers the wavelength range from 175 -:3300 nm. The instrument is

equipped with a high performance R928 photoniultiplier detector for the ITV-Vis region

and an electrothernially controlled lead sulfide (PhS) photocell for the NIR region. The

available light sources for this system are a tungsten halogen source with quartz window

for NIR and visible regions, and a deuterium are ITV source.

For typical thin film polymer samples deposited potentiostatically on a number of

different substrates, a three-electrode cell configuration has been used, as the one de-

scribed above, to allow potential application while monitoring the absorption/transmission

spectra. The samples are placed in nionomer free solution, and the electrodes are con-

nected to an EG&PAR model 27:3A potentiostat/galvanostat. For EC devices, only the

ITV-Vis- NIR Varian Cary 500 spectrophotonieter is used, in which the three-electrode

cell is replaced by a suitable holder available in the Cary 5 kit. In order to apply a volt-

age across the EC devices, the counter and the reference leads have to be connected to

one another. Finally, for both cases, EC polymer films on electrodes and in devices, the

experiment is performed by sequentially stepping the applied potential in either 0.1 V

or 0.2 V increments, starting front a potential at which the polymer is in its neutral










form and ending at a potential value for which the polymer is fully doped, while the
absorption/transmission spectr smntrd

The UV/Vis NIR Varian Cary 500 spectrophotometer provides one more method

of EC polymer on electrode or in device characterization in addition to the one outlined

above. Using the same experimental set up we can perform a kinetics experiment and

measure the speed with which the material is able to switch between two extrema states.

In this case, a single wavelength is used, the wavelength of maximum contrast which can

he determined from the experiment described above, and a square potential waveform is

applied at desired time intervals. The percentage transmittance (T' .) is monitored at

the wavelength of maximum contrast X,,a, as a function of time, while the EC polymer

on electrode (in solution), or in device, is repeatedly switched between the two extrema

states.

4.3.2 Interferometric or FTIR Spectrometer

An interferometer or Fourier Transform Spectrometer is a type of spectrometer

mostly used in the infrared region. The spectrometer relies on two 1! in r~ advantages

compared to the monochromatic spectrometer:


Jacquinot (or throughput, or i4tendue) advantage, which is the ability of interfe-

rometers to collect large amounts of energy at high resolution due to the use of

circular apertures, which have high throughput, and


Fellgett (or multiplex) advantage, which is the ability of interferometers to receive

information about an entire spectral range during one scan. In other words, all

frequencies in the region of interest are sampled simultaneously.


As a result of these advantages an interferometer has large resolving power, fast

sampling times, reduced stray light, and high signal-to-noise ratio (S/N) [58-60]. How-

ever, even though the Jacquinot (or throughput or i4tendue) advantage holds for low

and high frequencies, the Fellgett (or multiplex) advantage is lost for higher frequencies.

This is due to the fact that although for the infrared region the noise is usually detector









noise which is independent of the signal, for higher frequency regions the noise depends

on the source intensity (photon noise) [59].

More specifically, in the infrared region the signal-to-noise ratio for an interferom-

eter is proportional to the square root of the total time T, required for a scan of a broad

band Av:



(S/NV)I oc T1/2, 41

but in the visible region, the signal-to-noise ratio is also proportional to the square root

of the source intensity I:



(S/NV)I oc [(T/M)I(aY)1/2, (4.2)

where M~ is the number of spectral elements of width by in a broad band Av = v92 vl-

Comparing the above results with the signal-to-noise ratio for a monochromator, which

in the infrared region is



(S/NV)M Oc (T/M~)1/2, 43

and in the visible region is



(S/NV)M Oc [(T/M)I(aY )1/2, 44

we conclude that although the interferometer has an advantage over the monochromator

at low frequency regions due to the Fellgett advantage, equations (4.1) and (4.3), at

higher frequencies this advantage is lost and the two spectrometers provide the same

signal-to-noise ratio, equations (4.2) and (4.4) [58, 59].

There are several advantages and limitations for both types of spectrometers. The

Fellgett (or multiplex) advantage makes the interferometer an excellent choice for the

low frequency regions but the loss of it at higher frequencies in combination with the










availability of stronger sources and more sensitive detectors in that region, make the

monochromator a more suitable choice for high frequency regions.

Fourier Transform Infrared (FTIR) Spectroscopy

The principle of interferometry is based on the idea of the Michelson interferometer

which is depicted in Figure 4.4. There exist several variations of the interferometer:

Prism, Grating, Fabry-Perot, Lamellar and others, but here only the simplest version is

going to be described. As shown in Figure 4.4 the Michelson interferometer consists of

two plane mirrors, one stationary (MZ/l) and one movable (M22), and a beam splitter.

The light beam coming from the source is divided by the beam splitter into two ideally

equal parts. One part is reflected by the beam splitter to the stationary mirror and has

a fixed path length, and the other part is transmitted through the beam splitter to the

movable mirror and has a path length that can be varied by translating the movable

mirror. The two beam parts are reflected from the mirrors and they are recombined

again at the beam splitter. Due to the path length difference or optical retardation of

the two beams, which is twice the displacement of the movable mirror from its balanced

position (ZPD) a phase difference equal to 2xub6 is introduced, where v is the wave

number and it is defined as v = 1/X. The partial beams are spatially coherent and will

interfere when they recombine.

In more detail, assume that the source is emitting monochromatic plane waves of

the form:



Es(z) = So(z)ei et- .vz (4.5)

In a Michelson interferometer, after the amplitude splitting takes place, there are two

beams that have traveled different distances zz and z2, before they recombined at the

beam splitter. As was mentioned above, each beam has undergone one reflection and

one transmission at the beam splitter.










.M/L1
(fixed mirror)


Beam splitter


.M2
(movable mirror)


Figure 4.4: Schematic diagram of the Michelson Interferometer.


If rbs IS the reflection coefficient and tbs the transmission coefficient of the beam

splitter, then the resultant field of the recombined beams, usingf the superposition law,

will be


En(zi, z2 V)= bs bsO 0 i(wt-2xurrzl) + i(wt-2xurzz)]dv


(4.6)


and the intensity irradiancee or flux density) In(zi, z2, v) at the detector will be


In(zi, z2, V)= R 1, 2~ V)R 1, 2~, V) = 2So Ebtbbs 2[ + COS(2xuT(zi -- z2)1


(4.7)


where zz z2 = 6 is the optical path difference between the two beams. Hence, equation

(4.7) can be written











In(6, v) = 2So Ebsbs bs 2 1 COS(2iTub)]. (4.8)

In the case of a polychromatic source emitting a continuous spectrum from v = 0 to

v = 00, the total intensity at the detector will be




In(O) = In(i, v)d, 2|rbs bs2 jtt @di+ t9C) COS(2xub0)dii (4.9)


where the first term on the right side of the above equation is constant and it represents

the total intensity emitted by the source. At zero path difference 6 = 0 the intensity at
the detector is


In(0) =,,, 4|b b () (4.10)
0 \/V

Thus, equation (4.9) can be written


[I() n() =2rb s2 V)CS2u~y (4.11)



in which the quantity [Ia(6) In(0)] is known as interferogram. The Fourier cosine

transform of the interferogfram provides the actual spectrum:


1 1
BR(b, V) oc SJ2(y) =TIbsl [I i(5) -r I(0)] cos(2iTub)db. (4. 12)


However, in practice it is not possible to measure an interferogfram over an infinite

path difference. The finite optical path displacement results in the introduction of

numerous peaks into the transformed spectra, in addition to the main, which is one

centered approximately at v = 0. These peaks need to be corrected because they cause

a "l .1 ., of spectral intensity, the intensity is not localized to the main peak at v = 0

anymore but it is distributed to all peaks. These peaks are called side lobes, or "feet",










and the corrective procedure used for modifying the basic Fourier transform integral is

called apodization. For the correction of this error, the interferogram is multiplied by a

function, the apodization function, which vanishes outside the range of the acquisition

data. Their product is then Fourier transformed, in order to give the actual spectrum.

There are several apodization functions that can he used, the simpler ones are the

rectangular, or "b<.::- I1 shown in Figure 4.5:


1, if
Gl(x) = O I> (4.13)



and the triangular, also shown in Figure 4.5:




1 -, if IX L



where L is the maximum path difference. Other apodization functions that can he used

are the trapezoidal which is shown in Figure 4.5, the Happ -Genzel, the Blackman -

Harris, and the Norton -Beer (week, medium, or strong). Further details on the effects

and the choices of other apodization functions used, can he found in the literature [58,

59, 61]. It must he noted though, that the convolution of the apodization function with

the interferogram will result in reduction of the resolution sR, where sR ~ 1/L.

Another assumption that was made in this section, which is not ak- .--s true in

a real experiment is that the interferogram is symmetric with respect to the zero path

difference (ZPD) 6 = xl x2 = 0. Usually, the sampling of the interferogram starts at

6 = 0, and the interferogram function has a maximum value at this point; but if this

is not the case, then all sampling points are displaced by the same small amount. This

results in an .l-i-inin.! It 10 interferogfram with respect to the 6 = 0 point, and it introduces

a phase error.





























a) Boxcar














b) Trangular














c) Trapezoidal


Figure 4.5: Several apodization functions (right), and their Fourier transforms (left)
plotted together.










Furthermore, in the calculation of the spectrum, equation (4.12), a symmetrical

interferogfram was assumed. However, in the case of an .mi-inin- f~ lical interferogfram,

a complete cosine and sine Fourier transformation is needed for the calculation of the

spectrum. The application of the cosine Fourier transformation alone will result in a

distorted spectrum, and the creation of spurious lines. Unfortunately using the complete

Fourier transformation, twice as many points need to be collected in order to achieve

the same spectral resolution which translates into longer times for data acquisition.

Furthermore, the use of a phase correction mode is necessary in order for the phase

error to be corrected, and for the scanning time to be kept short.

Finally, the data acquisition requires the use of a digital computer for the calcu-

lation of the spectrum from the interferogfram function. In order for this calculation to

take place, the recorded data have to be digfitized into a number of discrete values. For

this reason, the interferogfram is sampled at steps of path difference Av. The discrete

nature of the real experiment can be handled mathematically by using Dirac delta func-

tions. The interested reader can find more details about this process in the literature,

references [59, 62]. The spectrum that is obtained from the sampled interferogram is

periodic, it repeats itself at multiples of Av. If the repeated spectra overlap then an

error effect, called aliasingf or "foldingt is introduced. For this error to be avoided, the

maximum frequency of the true spectrum must obey the following condition:



vmax (4. 15)


Other errors can occur from electronic filtering, misalignment, optical effects caused

by various parts of the instrument optics, such as the use of non-ideal mirrors (ideal:

101' reflective), and non-ideal beam splitters (ideal: non-absorbing, 50' transmissive

and 50' reflective), and not accurate adjustment of the movable mirror. These errors

introduce a phase factor that usually can be taken care of by the use of a phase corre-

ction mode. The most commonly used correction mode, which is also the one used in

our equipment, is the Mertz but there are several more that can be used.









Brucker 113v Interferometer

The reflectance, and transmittance measurements in the far infrared (FIR), and

mid-infrared (jllRt) regions were obtained with a Brucker 113v fast-scan Fourier trans-

form, or FT-IR, interferometer. The covered frequency range is from 20 cm-l to 5500 cm-]

The spectrometer, a schematic diagram of which is shown in Figure 4.6, consists of four

main chambers: source, interferometer, sample, and detector chambers. During the mea-

surements the system is evacuated to avoid the appearance of absorption bands from

water (H20), and carbon dioxide (CO2) in the spectra. The sample region is divided

into two identical chambers, one for reflectance and the other for transmittance mea-

surements with the use of special designed optical stages. The reflectance stage is shown

in Figure 4.7. For far infrared measurements (20 cm-l 700 cm- ), a mercury (Hg) arc

lamp is used as a source, and a liquid helium (He) cooled 4.2 K( silicon (Si) bolometer as

a detector, shown in Figure 4.8. In the mid-infrared region (400 cm-l 5500 cm- ), a glo-

bar (Silicon Carbide SiC) lamp is used as a source, and a room temperature pyroelectric

deuterated triglycerine sulfate (DTGS) as a detector.

The principle of the FT-IR Brucker interferometer is based on the Michelson inter-

ferometer described above. The light beam generated by the source, as shown in Figure

4.6 (a), goes through an automated circular aperture 3 to the beam splitter, mounted on

an automatic changer 4.6 (d), where is divided into two parts. Each part of the beam is

reflected by fixed mirrors, and after imaged onto the faces of the movable double-sided

mirror 4.6 (e), the two beams recombine at the beam splitter. Part of the recombined

beam is reflected by a series of mirrors, focused in the sample chamber 4.6 (i) or 4.6 (j),

and sent to the detector through another set of mirrors. The rest of the recombined

beam is reflected back to the light source. The recorded interferogram has a maximum

amplitude when the two arms of the interferometer have a zero path difference (ZPD).

When the double-sided mirror moves due to a scanning mechanism at a constant speed v,

a path difference of 6 = 4v/t, where the time t is measured from the ZPD, is introduced.
SThe aperture diameter can vary from 1.25 mm to 10 mm.





























I Source chamber
a Near-,mid- or far-IR sources
b Automated aperture


III Sample chamber
i Transmittance focus
3' Reflectance focus

IV Detector chamber
k Near-, mid- or far-IR
detectors


II Interferometer chamber
c Optical filter
d Automatic beam splitter changer
e Two-side movable mirror
f Control interferometer
g Reference laser
h Remote control alignment
mirror


Figure 4.6: Schematic diagram of Brucker (IBM) 113v FT-IR spectrometer.


Figure 4.7: Schematic diagram of the reflectance stage.

















CASE

-- NITROGEN
CAN


RADIATION
SHIELD
HELIUM CAN

12.50 inCOLD WORK
SURFACE




WINDOW
/r"HOLDER
T p OPTICAL
1.75 in .in102i AXIS


7.0 in62i





Figure 4.8: Schematic diagram of the liquid helium cooled detector.


In this system the mirror is air-bearing and is moving continuously. The calibration of

the position of the double-sided mirror is monitored by a He-Ne laser Figure 4.6 (g) and

a white-light reference interferometer.

The detected signal is amplified by a preamplifier, and sent to a 16-bit analog-to-

digital converter. The data are collected by a computer system, where the interferogram

is Fourier transformed to produce the single beam spectrum. The reflectance spectrum

is then calculated by taking the ratio of the single beam spectrum of the sample to the

single beam spectrum of the background reference, usually an aluminum (Al) mirror.

Additional calculations need to be done in this case, in order for the Al mirror/reference

reflectance to be taken into account.4 FOT CaSe Of tranSmittance measurements, the

4 The aluminum (Al) mirror is used as a reference for reflectance measurements but since it does
not reflect 100 % throughout the whole spectral range that is being used for, the spectrum needs to be
corrected. This is called "Al mirror correction".