CHEMICALVAPORDEPOSITIONOFYTTRIA-STABILIZEDZIRCONIAASA THERMALBARRIERCOATINGFORGASTURBINEENGINES By VENUGOPALVARANASI ADISSERTATIONPRESENTEDTOTHEGRADUATESCHOOLOFTHE UNIVERSITYOFFLORIDAINPARTIALFULFILLMENTOFTHE REQUIREMENTSFORTHEDEGREEOFDOCTOROFPHILOSOPHY UNIVERSITYOFFLORIDA 2004 ACKNOWLEDMENTS First,IwouldliketothankDr.TimAnderson,whoismychiefadvisorandthe AssociateDeanforResearchfortheCollegeofEngineeringattheUniversityofFlorida. Iappreciatehispatienceandhisknowledgeinthefieldofthermodynamics.Iappreciate allhiseffortstohelpmeunderstandhowimportantthermodynamicsistothefieldof chemicalengineering. Iwouldalsoliketothankmycommittee,whohavebeenveryinstrumentalinmy developmentasaprofessional.Dr.Crisalleshowedmetheessenceofteamworkandhard workwhenIparticipatedintheteachingassistantprogramattheUniversityofFlorida. OthercharacteristicsthatDr.Crisalleshowedmewereorganization,focus,and thoroughness. Mostimportantly,Dr.Crisallewillbealifelongfriend. Dr.Weaverhas givenmetremendousinsightintosurfacechemistry. Ialsoconsiderhimagreatfriend andhopethatheandIcancollaborateinthefuture. SpecialthanksaregiventoShirleyKelly,Peggy-JoThran,DebbieSandoval, JaniceHarris,andNancyKrell.Thesewonderfulladiesaremembersofthesupportstaff intheChemicalEngineeringDepartmentattheUniversityofFlorida. Iwouldliketo thankDr.GerhardFuchsforhisinsightintothesubjectmatterofthisthesis. Finally,I wouldliketothankDr.TheodoreBesmannfortheopportunitytoworkatOakRidge NationalLaboratoryandforhisguidanceinachievingthegoalsofthiswork. 11 1 TABLEOFCONTENTS Page ACKNOWLEDMENTS ii ABSTRACT vi CHAPTER INTRODUCTION 1 1 2 BACKGROUND 4 2.1 Introduction 4 2.2 ThermalBarrierCoating(TBC)System 5 2.3 EvaluationoftheTBCSystem 6 2.4 Zirconia-BasedCeramics 7 2.5 CurrentIssueswithTBCs 10 2.5.1 ThermalBarrierCoatingFabrication 10 2.5.2 ThermalBarrierCoatingLifetime 1 2.6 ChemicalVaporDeposition(CVD):AViableAlternative 13 2.7 ChemicalVaporDepositionofYttria-StabilizedZirconia(YSZ) 14 2.7.1 ChlorideCVDofYSZ 15 2.7.2 ParamtersInfluencingChlorideCVDGrowthRates 17 2.7.3 Metal-OrganicCVD(MOCVD)ofYSZ 24 2.7.4 ParamtersInfluencingMOCVDGrowthRates 29 3 EQUILIBRIUMMODELING 47 34..11 ThermochemiealProperties 47 3.1.1 PureSubstances 47 3.1.2 PhaseDiagramfortheZr-0System 48 3.1.3 PhaseDiagramtheY0i,5-Zr02System 51 3.2 ChemicalVaporDepositionEquilibriumCalculation 54 3.2.1 TheZr-Cl-O-C-H-InertSystem 54 3.2.2 TheZr-Y-Cl-O-C-H-InertSystem 57 3.2.3 TheZr-Y-O-C-H-InertSystem 60 3.3 Discussion 65 4 EXPERIMENTALAPPROACH 110 Materials 110 4.1.1 Reagents 110 4.1.2 SubstrateandSusceptor Ill iii 81 4.2 Methods Ill 4.2.1 PressureandTemperatureControl 11 4.2.2 PrecursorPreparation 112 4.2.3PrecursorDelivery 113 4.3 ReactionChamber 11 4.3.1EffectofGeometry 120 4.3.2EffectofGasFlowandSubstrate/ChamberHeating 121 4.3.3EffectofChamberPressure 123 4.4WasteandEffluentHandling 124 4.5Post-Processing 124 4.6CoatingAnalysis 125 4.6.1CoatingCrystalStructureAnalysis 125 4.6.2CoatingMorphologicalAnalysis 129 4.6.3CoatingCompositionalAnalysis 130 5 RESULTSANDDISCUSSIONOFCHLORIDECVD 142 5.1 Results 142 5.2 Discussion 146 6 RESULTSANDDISCUSSIONOFMOCVDUSINGBETA-DIKETONATES 155 6.1 Results 155 6.2 Discussion 159 7 RESULTSANDDISCUSSIONOFMOCVDUSINGN-BUTOXIDES 189 7.1 Results 189 7.2 Discussion 193 8 SUMMARY,CONCLUSIONS,ANDFUTUREWORK 218 8.1 SummaryandConclusions 218 8.2 RecommendationforFuturework 222 8.2.1 ImprovementsinCoatingDeposition 222 8.2.2 ThermalCycleTestingofCoatings 223 8.2.3DeterminationofaGrowthMechanism 224 APPENDIX A PDFCARDSFORUSEINX-RAYDIFFRACTIONANALYSIS 232 B CALIBRATIONFORTEMPERATUREMEASUREMENTS 239 IV REFERENCES 243 BIOGRAPHICALSKETCH 255 V AbstractofDissertationPresentedtotheGraduateSchooloftheUniversityofFlorida inPartialFulfillmentoftheRequirementsfortheDegreeofDoctorofPhilosophy CHEMICALVAPORDEPOSITIONOFYTTRIA-STABILIZEDZIRCONIAASA THERMALBARRIERCOATINGFORGASTURBINEENGINES By VennVaranasi August2004 Chair;TimAnderson MajorDepartment;ChemicalEngineering Thegasturbineengineusesanyttria-stabilizedzirconia(YSZ)coatingtoprovide thermalinsulationforitsturbineblades.ThisYSZcoatingmustbetetragonalincrystal sturcutre,columnarinmicrostructure,andbe100-250pmthicktoprovideforadequate protectionfortheturbinebladesinthesevereengineenvironment. Currently,YSZ coatingsarefabricatedbyelectron-beamphysicalvapordeposition(EB-PVD),butthis fabricationmethodiscostintensive.Chemicalvapordeposition(CVD)isamore commerciallyviableprocessingmethodandapossiblealternativetoEB-PVD.The depositionoftetragonalYSZfromgaseousmetalandoxidationsourceswerestudied. Achemicalequilibriumanalysismodeledthefeasibilityofdepositingtetragonal YSZforbothchlorideCVD(Zr-Y-C-O-Cl-H-Inertsystem)andmetal-organicCVD (MOCVD)(Zr-Y-C-O-Hsystem).PurethermochemicalpropertiesandtheassessedYSZ phasediagramwereusedinthisanalysis. Usingthemolarinputofmetals{{ny+«zr)and {ny/{ny+nzr)=0.08))asbases,equilibriumcalculationsshowedthattetragonalYSZ VI formationwasfeasible. TetragonalYSZformationwasfeasiblewithhighoxygen content{hq/{ny+nzr)>8)andhightemperature(T>100°C)inthecaseofchloride CVD(Zr-Y-C-O-Cl-H-Inert). TetragonalYSZformationwasfeasiblewithhighoxygen content(no/{ny+nzr)>5)andhightemperature(T>950°C)inthecaseofMOCVD (Zr-Y-C-O-H). AlthoughsolidcarbonformationdidnotappearinchlorideCVD, additionaloxygen{no/(ny+nzr)>32)andlowhydrogencontentrelativetocarbon{uh/ nc<2)wererequiredtoavoidsolidcarbonformationinMOCVD. Coatingsweredepositedusingasetofbaseconditionsderivedfromthechemical equilibriumanalysis. InchlorideCVD,YCI3wasnotincludedbecauseofitslowvapor pressure,thus,ZrCUwasoxidizedwiththeH2-CO2gasmixture. MonoclinicZr02 coatingsweredepositedatthethermochemicallyoptimizedconditions{hq/{ny+nz/)> 8,T>1004°C)withapproximately5.5pmh growthrate. Inmetal-organicCVD(MOCVD),liquidprecursorsolutionsofY-andZr-(3- diketonateandY-andZr-n-butoxideprecursorswereusedasthemetalsourcesandO2 gaswasusedastheoxidationsource. UsingtheY-andZr-P-diketonateliquidprecursor solution,tetragonalYSZwasdepositedwithalayeredmicrostructureapparentanda maximumgrowthrateofapproximately14pmh’’(ctivationenergy(Ea)of50.9±4.3kJ mol'*). Thegrowthrate(approximately43pmh withEa=53.8±7.9kJmol"')was improvedusingY-andZr-n-butoxideliquidprecursorsolutions,andthemicrostructure wascolumnar. Yet,two-phasedepositionofmonoclinicZr02andtetragonalYSZ occurred. Resultsofelectron-probemicro-analysisshowedthattheny/(ny+nzrratio waslessthan45%oftheny/(ny+nzrratiointheliquidprecursorsolution. vii CHAPTER 1 INTRODUCTION Demandforlowercostpowergenerationhasspurredresearcheffortstoward lower-costmaterialsfabrication. Akeypowergeneratingdeviceknownasthegas turbine,hasgarneredmuchattentionbecauseofitsefficientpowergenerating capabilities. Yet,thegasturbineincurshighcapitalandmaintenancecostsbecauseofits complexmaterialsdesignandfunction. Thus,improvementsinmaterialsandfabrication willboostthegasturbine’soveralllifetimeandefficiencyandpotentiallyreducecapital andmaintenancecosts. Costefficiency,engineefficiencyandenginelifetimearepartiallylimitedby turbinebladeperformanceandcyclelife(Haynes1997,Miller1997). Theturbineblades (whicharethemovingpartoftheturbine)requireprotectionfromexposuretohot combustiblegasesassociatedwiththeengineenvironment. Athermalbarriercoatingis usedtoprotecttheseturbinebladesfromhightemperaturesandcontaminantsinthefuel andintakeair(Haynes1997,Marple2000). Thisstudyfocusesonthecost-efficientfabricationofthermalbarriercoatings. Theceramic,yttria-stabilizedzirconia(YSZ)isusedsinceithasadequatehigh- temperaturepropertiesnecessaryforgasturbineenvironments(Wahletal.2001a). Currently,YSZisfabricatedusingairplasmaspray(APS)orelectron-beamphysical vapordeposition(EB-PVD),however,thesetechniquesarecapital-andoperational-cost intensive(Wahletal.2001b). Aneconomicalalternative,chemicalvapordeposition 2 (CVD)showspromiseasapossiblereplacementforAPSorEB-PVD(Krumdiecketal. 2003,Wahletal.2001a,Wahletal.2001b,PulverandWahl1997). ThegoalofthisworkisthesuccessfuldepositionofYSZusingtheCVDmethod. Thisworkusesachemicalequilibriumanalysistoidentify,optimize,anddeterminethe effectofCVDconditionsongasandsolidphaseformation. TheCVDconditionsused wereatommoleratios(e.g.,theratioofthetotalnumberofhydrogenatomsinthesystem tothetotalnumberofcarbonatomsinthesystem,regardlessofwhichmoleculesthey residein),thesubstratetemperature,andthereactorpressure. TheseoptimizedCVD conditionswereusedasabasisforselectingexperimentalconditions. Further,the equilibriumanalysiswasusedtoexplainexperimentalresults. ExperimentswereattemptedinaCVDreactor. Avertical,cold-wall,stagnation flowreactorwasimplementedinthiswork. Amongthemyriadofpossibleprecursors, onlythreetypeswereused. ForchlorideCVD,HClreactedwithZrchipstoprovidea controllableZrvaporsource. ForMOCVD,Y-andZr-p-diketonatesandY-andZx-n- butoxideswereused. Sincetheexperimentsusedliquiddelivery,atetrahydrofuran (THF)solventwasusedtodissolvetheY-andZr-P-diketonatesandtolueneandn- butanolwereusedtodissolveY-andZr-n-butoxides,respectively. Toanalyzethecoatingdeposition,severalcharacterizationtoolswereused. For crystalstructureanalysis,x-raydiffraction(XRD)wasused. Forcoatingmorphological analysis,scanningelectronmicroscopy(SEM)wasused. Forcoatingcompositional analysis,electronprobemicro-analysis(EPMA)wasused. Thenextchapterfollowswithabackgroundintowhyyttria-stabilizedzirconiais usedasathermalbarriercoatingandwhyCVDcanbeusedtofabricatethismaterial. 3 Thefollowingchapterdelvesintothechemicalequilibriumanalysistooptimizetheatom ratiosandintensivevariables(temperatureandpressure). Then,thefollowingchapter describestheexperimentalapproachforCVDandmaterialscharacterization. The followingthreechapterspresentanddiscusstheresultsofexperimentalefforts. Thelast chapterthensummarizes,concludes,andrecommendspossibleavenuesforfuturework.