Advanced characterization techniques for the knowledge-based design of hard coatings

(1)

Advancedcharacterizationtechniquesforthe knowledgeͲbaseddesignofhardcoatings

Dipl.ͲIng.MichaelTkadletz

beingathesisinpartialfulfilmentoftherequirementsforthedegreeofa DoctorofMontanisticSciences(Dr.mont.)

attheMontanuniversitätLeoben

Leoben,July2015

(2)

I

Financial support by the Austrian Federal Government (in particular from BundesministeriumfürVerkehr,InnovationundTechnologieandBundesministeriumfür Wissenschaft, Forschung und Wirtschaft) represented by Österreichische Forschungsförderungsgesellschaft mbH and the Styrian and the Tyrolean Provincial Government, represented by Steirische Wirtschaftsförderungsgesellschaft mbH and StandortagenturTirol,withintheframeworkoftheCOMETFundingProgrammeisgratefully acknowledged.

Affidavit

Ideclareinlieuofoath,thatIwrotethisthesisandperformedtheassociatedresearch myself,usingonlyliteraturecitedinthisvolume.

Leoben,July2015

(3)

II

Acknowledgments

MysincerestgratitudeisduetoProf.Dr.ChristianMittererforthepossibilitytocompose thisthesisattheChairofFunctionalMaterialsandMaterialSystems.Iamgratefulforhis supervision,advice,patienceandalltheconfidenceheplacesinme.Christian,youare leadinganoutstandinggroupandIamproudtobeapartofit!

FurtherIwouldliketothankao.Prof.Dr.JozefKeckesforhissupport,supervisionand especiallyforhispatience.Jozef,thankyouforsharingyourknowledgewithmeandforall thefunwehad!

Iamgratefultoao.Prof.Dr.ReinholdEbnerandtoMag.AlexandraPurkarthofer,managing directorsoftheMaterialsCenterLeoben(MCL),forgivingmetheopportunitytocarryout mythesiswithinanMCLproject.IalsowouldliketothanktheMCLstaffandcoͲworkersfor theirsupport.Especially,IwouldliketothankBernhardSartoryforallthesupportand fruitfuldiscussions,hisneverendingmotivationandtheoutstandingworkhedoesonthe REMͲFIB.Bernhard,wereallydidgreatstufftoghether,thankyou!

Also,IwouldliketoexpressmygratitudetoMScMariannePenoyandDIClaudeMichotte from CERATIZIT Luxembourg and to Dr. Christoph Czettl and DI Markus Pohler from CERATIZITAustriafortheirsupportandthefruitfuldiscussions.

Iamalsoverygratefultothewhole“ThinFilmGroup”andmystudentcoͲworkersforthe greatworkingatmosphereandsolidarity.SpecialthanksgotoDr.OliverJantschner,sharing anofficewithyouwasgreatfun!

EspeciallyIwouldliketothankmyformerofficemateandsupervisorDr.NinaSchalk.Nina, therearenowordstothankyouforallyoursupport,patienceandfriendship!

Iwouldliketothankmyparentsandmybrotherfortheirloveandsupport.Withoutyouthat wouldneverhavebeenpossible.

Lastbutnotleast,IwouldliketoexpressmygratefulnesstoDaniela,forbeingbymyside, forherpatienceandforbringingsomuchloveandhappinessintomylife.Iloveyou!

(4)

III

1. Introduction

Hardandwearresistantcoatingsarewidelyusedinthefieldofcuttingapplications,fordies andmoldsaswellasforotherapplicationsundersevereoperatingconditions[1–3].The highhardnessofthecoatingsresultsinasignificantlyimprovedresistanceofthetools againstabrasivewearandtheirhighchemicalstabilityisanimportantfactorminimizing chemicaldegradation due tofor example oxidation[4–7]. The demand to constantly improve theproductionthroughputbyincreasingcuttingspeedand feedratefurther increasesthechemical,thermalandmechanicalloadsonthetoolsandconsequentlyalsoon the coatings. This promotes further optimization of thecoatings, where typically the chemicalcompositionandmicrostructureareadjustedorpostͲdepositiontreatments,like thermal annealing or mechanical blasting, are applied [8–12]. Thus, microstructural gradientsandconsequently,gradientsofresidualstress,hardnessandotherproperties whichareeitherdirectlyorindirectlyaffectedbythemicrostructurecanbetailored[13–15].

Additionally,thehightemperaturesduringservicecanresultinchangesoftheseproperties [16,17]. Thus, the interest in crossͲsectional and high temperature characterization techniqueshasbeengraduallyincreasingintherecentyears.Theaimofthisthesisisthe evaluation and further enhancement of characterization techniques as well as the establishmentofnoveltechniquesforadvancedcharacterizationofhardcoatings.

Withinthefirstpartofthethesis,extensivescanningandtransmissionelectronmicroscopy investigationswere performed ontheweartracksformed onarcevaporatedTiAlTaN coatingsbyballondisktestingatroomtemperature(RT)andat700°C.Specialattention waslaidtogrowthdefectssuchasdroplets,wherefocusedionbeam(FIB)techniqueswere usedtogetadeeperinsightintothecoatingmaterial.AcombinationofFIBcutandslice techniques,energydispersiveXͲrayspectroscopy(EDX)andgrayscaleimagecorrelationwas appliedtoobtaintomographicalimagesofselecteddroplets,providingaccesstothe3 dimensionaldistributionoftheirchemicalcomposition.

Recently,Keckeset al. [18]introducedanapproach,referred toassynchrotronXͲray nanodiffraction,whichenablesthedeterminationofdepthdependentresidualstrainasa functionofthecoatingthicknesswitharesolutionof100nmorevenless.Sincethis approachislimitedtonanocrystallinesamples,duetothesmallsizeoftheXͲraybeamwhich isusedtoprobethesample(beamdiameter<100nm),itwasmodifiedwithinthesecond partofthethesistoimprovethediffractionstatisticsforcoarsegrainedmaterialstoa satisfyinglevel.AnXͲraybeamwithdimensionsof100nminheightand10μmwidthwas implementedatthebeamlineID13attheEuropeanSynchrotronRadiationfacility(ESRF).

(6)

MichaelTkadletz Introduction

2

Measurementsonexemplarychemicalvapordeposited(CVD)DͲAl₂O3coatings,withagrain sizeinthemicrometerrange,havebeenperformedandthelineͲfocusedXͲraybeamhas beenevaluatedforthedeterminationofdifferentstressgradients,introducedbyblasting treatmentswithdifferentblastingmedia[19].

Besides techniques forthedeterminationof chemical, microstructural and mechanical properties,atoolͲboxofmethodsforthedeterminationofthethermoͲphysicalpropertiesof hardcoatingswasestablishedwithinthisthesis.ThethermoͲphysicalpropertiesofhard coatingshaveamajorinfluenceonthelateralanddepthdependenttemperaturefields duringservice,whichconsequentlyresultinstressgradients[20].Thus,specialattentionwas paidtotheinvestigationofthethermalconductivity,heatcapacityandthermalexpansionof wearresistantcoatings.ForTiN,AlNandTiAlNhardcoatingsgrownbyphysicalvapor deposition(PVD),biaxialstresstemperaturemeasurements(BSTM)andhightemperatureXͲ raydiffraction(XRD)weresuccessfullyappliedtorevealthecoefficientofthermalexpansion (CTE)ofthesamples.Bothtechniqueswerecomparedtoeachotherandtheresultsare discussed.Differentialscanningcalorimetry(DSC)wasusedtoinvestigatethespecificheat capacityofthesamplesandacomparisontovaluesprovidedbytheNationalInstituteof StandardsandTechnology(NIST)isprovided.Measurementsofthethermalconductivityof TiAlNcoatingswithdifferentmicrostructures,usingtimeͲdomainthermoreflectance(TDTR), wereperformedtoinvestigatetheeffectofmicrostructuraldesignonpreservationofalow thermal conductivity which is stable also afterannealing at 1200 °C for 120 min. A comprehensivereview,representingthelastpartofthethesis,summarizestherecent advancesinthecharacterizationofmicrostructure,chemicalcompositionandmechanical, tribological and thermoͲphysical properties of hard coatings, taking into account the methodsdevelopedand/oradaptedwithinthiswork.

(7)

MichaelTkadletz Coatingtechnology

3

2. Coatingtechnology

2.1. Coatingdeposition

Typicalcommerciallyappliedhardcoatingsarebasedonnitridesoroxidesofoneormore (transition)metalslikeAl,TiorCr,sometimesdopedwithsmallamountsofB,CorSi.

CommonrepresentativesareTiN,TiAlNordifferentpolymorphsofAl2O3.Thesynthesisof coatingsutilizingvapordepositionmethodsrequiresthethreefollowingbasicsteps:(i) generationofthevaporphase,(ii)transportofthevaporfromsourcetosubstrateand(iii) depositiononthesubstratebycondensation,nucleationandcrystalgrowth.Techniques whicharetypicallyusedforthedepositionofhardandwearresistantcoatingsareeither physicalvapordeposition(PVD)orchemicalvapordeposition(CVD)[21–23].

2.1.1 Physicalvapordeposition

InPVD,solidmaterials,socalledtargets,areusedwhichcaneitherconsistofmetals,alloys orcompoundslikenitridesoroxides.Inthecaseofcompoundtargets,thetargetsmoreor lessexhibitthefinalchemicalcompositionofthecoating.AnotherpossibilityisreactivePVD, wheremetallictargetsareusedtodepositcompounds,whereasothercomponentssuchas NorOareprovidedasreactivegas[23].Further,forPVDdifferentprocessescanbe distinguishedbythewayhowthetargetisevaporated.Twofrequentlyappliedtechniques aresputterdepositionandcathodicarcevaporation(CAE)[22].

AschematicofatypicalsputterprocessfordepositionofmetaloxidesisshowninFig.1[24].

Duringdeposition,aworkinggas(typicallyAr)isprovidedinthechamberandbyapplyinga voltage between thetargets(cathode)andthe chamber (anode),aplasmais ignited.

Subsequently,thepositivelychargedAr⁺ionsareattractedtothenegativelychargedtarget andhititssurface.DuetothemomentumtransferfromtheAr⁺ionstothetargetatoms, collisioncascadesareinitiated.Thosecascadesrecoilwithinthetarget,andiftheenergyof thetargetatomsbecomeshigherthantheirbindingenergy,theyareejected.Typical ionizationratesforsputteringarebelow1%,thus,theejectedparticlesaremainlyneutrals.

Afterejection,theatomsmovethroughthechamberonaballisticpath.Iftheatomshitan obstacle,whichcanbeeitherthechamberwallorthesubstrate,theycondenseanda coatingisformed.IthastobenotedthatalthoughFig.1illustratesformationofthe compoundinthegasphaseonly,compoundformationmightoccuralsoatthetargetaswell asatthesubstratesurface[21,22,25–27].

(8)

4

Fig.1:Illustrationofatypicalsetupforsputterdepositionofmetaloxidecoatings[24].

Theefficiencyofthedepositionprocessandthequalityofthedepositedcoatingscanbe furtherenhanced,ifunbalancedmagnetronsputteringisapplied.There,anincreaseofthe plasmaionizationrateandtheionfluxonthegrowingfilmcanbeachievedbyaspecial arrangement of a magnetic system behind the target. In principle, a magnetron configurationtrapselectronswithinthemagneticfieldinthetargetnearregion,yieldinga higheriondensityandconsequently,increasedefficiencyofthesputterprocessanda characteristicerosiontrackonthetarget.Anunbalancedmagnetronconfigurationallows themagneticfieldtoreachthenegativelybiasedsubstrateasthemagneticpolesareof differentstrength.Thus,acertainamountofelectronsreachesthesubstratenearregion, againpromotingionizationofatomsinthatregion.Theincreasedamountofionsinthe substratenearregionresultsinanamoreenhancedionbombardmentofthegrowing coating,whichimprovestheadherence,increasesthedensityandresidualstrainofthe coating[25,27,28].

Incontrasttosputtering,duringCAEtheevaporationofthetargetmaterialisachievedbyan electricarcdischargemovingacrossthetargetsurfacewhichisshownschematicallyin Fig.2a.Duetothehighcurrentdensity(10⁶to10¹²A/m²)ofthesmallcathodespot(10^Ͳ8to 10^Ͳ4m),i.e.thespotwherethearcmeetsthetarget,hightemperaturesariseandthetarget materialisevaporatedexplosively.CAEprocessesareusuallyperformedunderinertgasor, similartoreactivesputtering,inareactivegasatmosphere.Incontrasttosputterdeposition, veryhighionizationratesofmorethan90%canbereachedduringCAE.Thisoffersthe possibilitytoadjustthekineticenergyandpathoftheparticlesimpingingonthesubstrate

(9)

5

byapplyingabiasvoltage.DuringCAE,theevaporatedspeciesalsoshowhigherionization states,seeFig.2awheredifferentlychargedTiionsareshown[26,29–32].

OnemajordrawbackofCAEisthehighnumberofmoltenmacroscopicparticleswhichare ejectedfromthetarget,andsubsequentlyformdefects,socalleddroplets,withinthe depositedcoatings.Dropletsarecriticalastheyincreasethesurfaceroughness,provide diffusionpaths,fosteroxidationanddeterioratethemechanicalintegrityofthecoatings [31–35].Atypicaldropletsurroundedbycavities(indicatedbythewhitearrow)arisingfrom shadowingeffectscanbefoundinthesecondaryelectron(SE)scanningelectronmicroscopy (SEM)micrographinFig.2b.ThecrossͲsectionalSEͲSEMmicrographshowninFig.2cwas preparedbyFIBmilling.Itrevealsaloosebondingofthemetallicdroplettothesubstrate, whichwasovergrownbytheTiAlTaNcoatingduetotheongoingdepositionprocess.The surrounding cavities can also be identified in the crossͲsectional micrograph and are indicatedbywhitearrows.

Fig.2:(a)SchematicofaCAEdepositionprocess[36].SecondaryelectronSEMmicrographs ofadropletcaptured(b)onthesurfaceand(c)withinaFIBcrossͲsection[ownwork].

DuringsputteringaswellasCAE,theapplicationofabiasvoltage,i.e.anegativevoltagein therangeofͲ50toͲ200Vappliedbetweenthesubstrateandchamberisverycommon.The biasvoltagepromotestheionizedparticlefluximpingingonthegrowingcoating,which resultsinsignificantlyimprovedcoatingpropertiessuchashardnessanddensity.Further, thebiasvoltageisacommontooltocontrolandadjusttheresidualstresswithinPVD coatings[8,23,37,38].

PVDisalineͲofͲsightprocess;thus,oneinherentdrawbackisthestraightforwarddeposition whichcausesshadowingeffectsandprohibitsdepositionwithincavitiessuchasholesor behindedges.AmajoradvantageofPVDingeneralisthewidetemperaturerangewhichcan beappliedfordeposition(fromroomtemperatureupto700°C)andconsequentlyallows depositionofcoatingsalsoontemperaturesensitivesubstrateslikee.g.certainsteelsor evenplastics.Further,thewiderangeofaccessibletargetmaterialsandreactivegases

(10)

6

providesahighflexibilityandavastamountofpossiblecoatingmaterials.Theopportunity to adjust the depositionparameterstoagreatextent allowstailoring ofthecoating propertiestoahighdegree.AnothermajoradvantageofPVDisthepossibilitytodeposit coatingsundernonͲequilibriumconditions,whichfacilitatesthesynthesisofmetastable coatingslikeTiAlN[39,40].

2.1.2 Chemicalvapordeposition

InCVD,coatingdepositionisachievedbytheuseofprecursorswhichcanbesolid,liquidor gaseous.SolidprecursorssuchasAl,ZrorHfaretypicallychlorinatedwithinthedeposition plantinordertoevaporateandtransporttherespectivespeciesintothereactor.There,the volatile precursors typically react by thermal activation to solid products, which subsequentlyformthecoating.Theactivationofthechemicalreactioncanbefurther assistedbyplasma,laserormicrowavesinordertodecreasereactiontemperaturesor increasedepositionrates.Thecoatingsdepositedwithinthisworkweresynthesizedusinga socalledhotͲwallCVDreactor,wheretherecipientissurroundedbyaheatingsystem (furnace)inordertoachievethenecessarydepositiontemperatures,whichareintherange of700to1100°Cforhardcoatings.AnillustrationofanindustrialscalehotͲwallCVDreactor isshowninFig.3,includingtypicalexamplesofsolid,liquidandgaseousprecursorsaswell asthevacuumsystemandagasscrubberafterthereactor[22,27,41–44].

Fig.3:BasiccomponentsofanindustrialscalehotͲwallCVDreactor[42].

Duringdeposition,fivebasicstepsclosetothesubstratecanbedistinguished:(i)Diffusionof thefilmformingspeciesthroughaboundarylayerand(ii)adsorptiononthesubstrate surface.Thisstepisfollowedbysurfacediffusion,(iii)thefinalfilmformingreactionand(iv)

(11)

7

consequentialdesorptionofbyͲproducts.Inafinalstep(v)theremainingbyͲproductsare exhaustedfromtherecipient.

IncontrasttoPVD,oneofthemainadvantagesofCVDisthehighsocalledthrowingpower, describingthefeasibilityofcoatingdepositiononoutͲofͲsightareas,enablingdepositionalso onshadowedpartsofthesubstrate.ThelargereactorsusedinCVDallowtoprocesslarge batchsizes(e.g.15000halfinchcementedcarbideinserts),whichmakesdepositionof ratherthickcoatingsintherangeof10to15μmefficient[21,41,42].Further,thevariation ofprecursorsandreactionconditions,suchasforexamplethegaspressure,provideahigh flexibilityintermsofthestoichiometryofthedepositedcoatings.

2.2. Thinfilmgrowth

AsummaryofthebasiceffectsinvolvedincoatinggrowthisillustratedschematicallyinFig.4 [45].Aftercondensationofthefilmformingparticlesonthesubstrate,twobasicsurface interactionsmightoccur:(i)Surfacediffusionor(ii)reͲevaporation.Theparticleswhich diffusealongthesurfacemightbeadsorbedatspecificsitessuchassurfaceasperities,edges orimpurities.Inaddition,thereisalsothepossibilitythatseveralparticles,diffusingalong thesurface,formclusterswhichmighteitherdiminishagainorgrowfurther,iftheyreacha critical size which is then referred to as nucleation. A continuous coating grows by coalescence of multiple nuclei. With ongoing deposition, also interdiffusion between substrateandgrowingcoatingcantakeplace[21,23,25,45].

Fig.4:Interactionsofcondensedparticlesonasurfaceduringcoatingdeposition[45].

Afternucleation,dependingonthebindingenergiesoftheparticles,growthcanoccurin threedifferentmodeswhichareshowninFig.5.LayerͲbyͲlayer(FrankͲvanderMerwe) growthoccursinthecaseofstrongerbindingenergiesbetweenatomsofcoatingand substratematerial(Fig.5a),whileIsland(VolmerͲWeber)growthoccurswhenthebinding energybetweenthecoating particlesis stronger(Fig.5b). Layer plusisland(StranskiͲ

(12)

8

Krastanov)growthcanbeobserved,iffirstlayergrowthismorefavorablewhichchangesto islandgrowthafterdepositionofauniformlayer(Fig.5c)[25].

Fig.5:Typicalgrowthmodesofcoatings:(a)layerͲbyͲlayer(FrankͲvanderMerwe),(b)island (VolmerͲWeber)and(c)layerplusisland(StranskiͲKrastanov)growth[25].

During deposition, the growth kinetics which has a major influence on coating microstructureandthusproperties,canbestronglyinfluencedbythedepositionparameters such assubstrate temperature, working gas pressure andbiasvoltage.Movchanand Demchishin[46]werethefirsttointroduceastructurezonemodel(SZM)forthermally evaporated films which relates the homologous temperature (Ts/Tm, where Ts is the substratetemperatureandTmisthemeltingpointofthedepositedspecies)totheobserved microstructures.Basically,themodelconsistsofthreedifferentzones.Inthefirstzone,at temperaturesbelow0.3xTm,theadatommobilityisratherlowandthus,coatingsdeposited withinthiszoneexhibitahighroughnessandporositycausedbyshadowingeffects.Films depositedinthesecondzone,rangingfromtemperaturesfrom0.3to0.45xTm,showa smoothsurfaceandcolumnar,densestructure.Thethirdzone,abovetemperaturesof 0.45xTm, is defined by large equiaxed grains formed due to recrystallization and interdiffusion.

Later,Thornton[47]proposedanextendedmodel,includingthepartialpressureofa working gasinordertoexpandthe model tosputteredcoatings.Messieretal.[48]

introducedamodifiedmodelwhichincludedthebiasvoltageduring deposition.Both revisedmodelshaveincommonthattheyaddedafourth,socalledtransitionzonebetween zone1and2,inwhichthecoatingsexhibitadensebutfibrousstructure.In2010,Anders [49]revisedThornton´sSZM.BychangingthehomologoustemperatureTs/Tmintothe generalizedtemperatureT^*,therevisedSZMshowninFig.6aalsoconsidersthepotential

(13)

9

energyofparticlesarrivingatthecoatingsurface.Further,thepressureaxiswasreplacedby anormalizedenergy(E^*)axis,thus,consideringthekineticenergyofbombardingparticles.

Further,thepreviouslyunlabeledthirdaxisisusedtorepresentthecoatingthickness.With theseimprovements,theSZMdoesnotonlyrepresentthequalitativeinfluenceofbias voltageandpartialpressureonthecoatingmicrostructure,butalsoexpandsittotechniques exhibitingalargeionͲflux.Finally,Anders’SZMindicatesthetransitionfromtensileto compressivestressaswellasareasatlowandhighvaluesforE^*atsimultaneouslowT^*, whicharenotaccessible.

DuringCVDthemainfactorsinfluencingcoatinggrowtharethereactionkinetics,basically influenced by the deposition parameters. The deposition temperature, precursor gas throughput(i.e.gaspressure)andthefeedgascompositionaswellasthepresenceof additional dopants areknowntoinfluencethemicrostructure ofthegrowingcoating significantly.AschematicillustratingthreetypicalstructurezonesoccurringinatypicalCVD coatingisshowninFig.6b[44].There,zone1representsacoatingconsistingofcolumnar regularlyshapedgrainswithdomeͲlikecaps.Thiscanberelatedtohighsurfacediffusionand uninterruptedgraingrowthasaresultofhighdepositiontemperatures.Withinzone2,the grainsarestillcolumnarshaped,butinclinedtoacertaindegreeanddistinctfacetsonthe surfacecanbeobserved.Thischangeinshapefromzone1tozone2canbeattributedto lower deposition temperatures and thus, decreased diffusion, as well as increased supersaturationoftheprecursormixture.Zone3isdeterminedbythegrowthofrather small equiaxed grains causedby minimizeddiffusioneither at low temperatures, low pressuresorhighsupersaturations[44].

Fig.6:Structurezonemodelfor(a)PVDcoatingsaccordingtoAnders[49]and(b)CVD coatings[44].

(14)

10

2.3. PostͲdepositiontreatment

PostͲdepositiontreatmentsofhardcoatingsarewidelyemployedtofurtherimprovethe coating performance either by surface modification or by improving their mechanical properties.Frequently,mechanicaltreatmentssuchaspolishing,brushingorwetͲanddryͲ blastingareappliedinordertosmooththesurfaceandconsequently,reducefriction.

Moreover,afterblastingacertainamountofremainingblastingmaterial,socalledtransfer material,canbefoundonthecoatingsurfacewhichfurthercontributestothemodification ofthetribologicalbehavior[50,51].Besidesthesurfacemodification,blastingtreatments havebeenprovensuitabletotailortheresidualstressofhardcoatingswithinawiderange.

Depending onthe chosenblasting medium andblastingparameters,it is possible to significantlychangetheresidualstresswithintheuppermostregionofthesurfaceoreven downintothesubstrate[10–12,19,52–55].

EspeciallyinthecaseofCVDcoatings,whichtypicallyexhibittensileresidualstressintheasͲ depositedstate,blastinghasagreatpotentialtoimprovethemechanicalproperties[56].

Thus,wetͲanddryͲblastinghasrecentlygainedincreasinginterest,wherenotonlyachange inmagnitudebutalsoachangefromtensiletocompressivestresswasreportedrepeatedly [19,51,52,57,58].AnillustrationofawetͲblastingprocessusingtwodifferentblastingmedia isdepictedinFig.7a[12].AtypicalstressͲdepthgradientcomputedforblastingof42CrMo4 steelispresentedinFig.7b[59],exhibitingacompressivestressmaximumbelowthe surface,whichisgraduallydecreasingwithincreasingdepth,endingwithapronounced tensilepeakbeforereachingtheinitialstressstate.AccordingtoSchiffneretal.[59],the maximumintroducedresidualstressanditsdepthbelowthesurfaceaswellasthetotal affecteddepthcanbesignificantlyinfluencedbythesizeoftheblastingmediumandits velocity.Further,thedensityandthemechanicalpropertiesoftheblastingmedium,i.e.

hardnessandYoung´smodulus,playanimportantrole.Alsoblastingparameterslikeblasting pressure,workingdistance,workingangleandnozzlegeometryhavesignificantinfluenceon theeffectoftheblastingtreatment.

(15)

11

Fig.7:(a)SchematicofawetͲblastingprocessmodifiedafter[12].(b)ResultingstressͲdepth gradientafterblastingof42CrMo4steel(computedresult)[59].

StressͲdepthgradientssimilartoFig.7bcanalsobeexpectedforhardcoatings.Reportson wetͲanddryͲblastedhardcoatingsconfirmtheinfluenceoftheblastingparametersand blastingmedia.Inliterature,thedeterminationoftheintroducedresidualstressisoften basedondepthaveragingtechniques,thus,onlyprovidingmeanvaluesoftheintroduced residualstress.There,theresultingstressvaluesmightappearsimilar,althoughtheactual gradients might be verydifferent. Thus,the determination ofthe actual stressͲdepth gradientsiscrucialtoevaluatetheimpactoftheblastingtreatmentandtorelateittothe cutting performance of a coated tool. With ongoing development of depth sensitive techniquessuchasangledispersiveorenergydispersiveXRDmeasurements,investigations ofstressͲdepthgradientscausedbyblastingtreatmentsbecameavailablebutremainstilla challengingtask[52].Recently,Keckesetal.[18]introducedatechniqueapplyingcrossͲ sectionalXRDusingasynchrotronXͲraybeamwithadiameterbelow100nmtolamellaeof thesamplesintransmissiongeometry–atechniquewhichisreferredtoassynchrotronXͲ raynanodiffraction.AschematicofthistechniqueisshowninFig.8a[18].Thisapproach allowsthedirectdeterminationofthelatticestrainand,consequently,stressasafunctionof thecoatingthicknesswitharesolutionbelow100nmbyrecordingandevaluatingDebyeͲ Scherrer rings for each scanning step. A critical comparison of a nanodiffraction measurementandadepthsensitivelabbasedapproachcombinedwithenergydispersive synchrotroninvestigationsperformedonadryͲblastedsampleisshowninFig.8b[57].The comparisonrevealsexcellentagreementwithHertziantheoryandcomputedresultsinterms oftheshapeofthestressͲdepthgradient.Alsothestressmagnitudeobtainedbyboth measurementsisingoodagreement.However,incontrasttoXͲraynanodiffraction,thelab

(16)

12

basedapproachdoesnotallowtoresolvethedistinctstressreliefnearthesurface,whichis acommoneffectafterblasting.Thus,thedescribedXͲraynanodiffractiontechniqueoffers the possibility to investigate the effect of blasting treatments on hard coatings with outstandingresolutionandprecisionrathereasily.

Fig.8:(a)SchematicofthesynchrotronXͲraynanodiffractionsetup[18].(b)Comparisonof laboratoryandsynchrotronmeasurementsofablastedTiNsample[57].

Besidesthealreadymentionedmechanicalposttreatments,alsothermalposttreatment strategiesmightbeappliedtocertainhardcoatingsinordertoincreasetheirhardnessby agehardeningortoreducethecoefficientoffriction.Fig.9a[60]showstheevolutionofthe hardnessofaTiAlNcoatingasafunctionoftheannealingtemperature,wherethepeak valuearound900°Cisattributedtoagehardeningandthedecreasinghardnessathigher annealingtemperaturestotheformationofwurtziticAlNdomains.Incomparison,theTiN referenceshowsadecreasinghardnessstartingalreadyatannealingtemperaturesof400°C duetorecoveryofstressesandrecrystallization[60,61].Frictionreductioncanbeeither achievedbyoutdiffusionofelementsforminge.g.rutile[51],whichactsaslubricant,orby depositionofsolidlubricantssuchascarbonislandsonthecoatingsurface[62].InFig.9b [62],aSEMmicrographofacarbondepositonaTiAlTaNhardcoatingisshownwhichwas formedduringannealinginmethaneat900°Cfor1h.Duetothiscarbondeposit,the coefficientoffrictionofthecoatingscouldbedecreasedfrom0.6to0.2[62].

(17)

13

Fig.9:(a)EvolutionofhardnessofTiNandTiAlNwithincreasingannealingtemperature[60].

(b)SEMmicrographshowinganinhomogeneouscarbondepositontopofaTiAlTaNcoating, responsibleforlowfrictionbehavior[62].

(18)

MichaelTkadletz Hardcoatingmaterials

14

3. Hardcoatingmaterials

3.1. TiAlNCoatings

Incuttingapplications,TiAlNbasedcoatingsarefrequentlyappliedduetotheiroutstanding mechanicalpropertiesatroomtemperatureaswellasatelevatedtemperatures.Sincethe depositionofTiAlNutilizingCVDisstillverychallenging,themajorityofavailableTiAlN coatedtoolsareproducedbyPVD.ThestablestructureofTiNisthefacecenteredcubic(fcc) rockͲsaltstructure,exhibitingareasonablehardnessandwearresistance.However,the oxidationresistanceofTiNisratherpoor,whichusuallylimitsthelifetimeandoperating temperatureduringcutting.AlNexhibitsawurtzitic(w)structurewithasignificantlylower hardness,thus,itisusuallynotusedforcuttingapplications[39,63,64].However,the additionofAltoTiNleadstotheformationofametastablefccͲTi1ͲxAlxNsolidsolutionby substitutionofTibyAlatomsinthefcclatticeuptoxvaluesofabout0.65to0.75,asshown inFig.10[65].Thereby,anincreasingAlcontentresultsinanincreasinghardnessand consequentlyenhancedwearresistance.Further,theoxidationresistanceissignificantly improvedduetotheformationofastableAl2O3scaleonthecoatingsurface[63,66,67].

Abovexvaluesof0.65to0.75,thewurtziticstructurebecomespredominant,whereAl atomsaresubstitutedbyTiatoms[33,68–70].Thechangefromfcctowurtziteathighx valuesresultsinapronounceddecreaseinhardnessandwearresistance[33,66,67,70].

Fig.10:EvolutionofcrystalstructureofTi1ͲxAlxNasafunctionoftheAlcontent[65].

Attemperaturesabove800°C,themetastablefccͲTi1ͲxAlxNisknowntoundergospinodal decompositionintofccͲTiNrichandfccͲAlNrichdomains.Thisresultsinasignificantincrease in hardness due to age hardening caused by coherency strains (see Fig. 9a). If the temperatureisfurtherincreased,themetastablefccͲAlNdomainstransformintostablewͲ AlN,resultingindeteriorationofhardnessandwearresistance[33,60,70].

Recently,alloyingoftransitionmetalsasathirdcomponenttotheTiAlNsystemhasbeen intensivelyinvestigated[71,72].PossibletransitionͲmetalͲnitridecandidatesforalloying,

(19)

15

includingsomeimportantpropertiesaswellastheirrelativelatticemismatchwithrespectto afccͲTiAlNreference,aredepictedinFig.11a[8].Thecorrespondingincreaseinhardness duetoalloyingaswellasthehardnessevolutionduetoannealingofselectedcoatingsis showninFig.11b[8].ForalloyingwithHforNbasignificantincreaseinhardnessatroom temperaturecanbeobserved,whiletheagehardeningeffectislesspronounced.Alloying withYslightlyincreasesthehardnessatroomtemperature,butwithincreasingannealing temperaturethehardnessdeterioratesdramatically.AlloyingwithTaresultsinthemost pronouncedincreaseinhardnessatroomtemperature,whichcanbe retainedduring annealingalthoughnodistinctagehardening canbeobserved.Further,asignificantly improvedtribologicalbehavioraswellasincreasedoxidationresistancecanberelatedto alloyingwithTa.DuetothesignificantpositiveeffectofTa,TiAlTaNcoatingsaretoday standardcoatingsforcuttingapplications.

Fig.11:(a)SelectionoftransitionmetalnitridessuitableascandidatesforalloyingtoTiAlN and(b)theireffectonthehardnessandagehardening[8].

Mayrhoferetal.[56,73]reportedontheinfluenceofmicrostructureandresidualstresson themechanicalproperties,inparticularthehardness,ofvarioussputteredhardcoatings,see Fig.12a[73].Theresidualstresscanbetailoredby,e.g.,varyingthebiasvoltageduring deposition.Therespectivemicrostructureformedatthesedifferentbiasvoltagesenablesto adjustthemechanicalpropertiesofthedepositedcoatingswithrespecttotheirthickness,as recently shown by Daniel et al. for sputtered CrN coatings [74]. Thereby, the XͲray nanodiffraction approach [18] described in the previous section and crossͲsectional nanoindentationmeasurements[74]canbehelpfultoolstoinvestigatethemicrostructural evolutionof hard coatings.Exemplary,strainplots derived fromXͲray nanodiffraction measurementsontwodifferentTiAlNcoatingscanbefoundinFigs.12bandc.Bothcoatings

(20)

16

were~6μmthick.Duringdeposition,thebiasvoltageofthecoatingshowninFig.12bwas changedfrom Ͳ50Vto Ͳ40Vafterhalfofthedepositiontime,resultinginasignificant reducedresidualcompressivestrain.ForthesampleshowninFig.12b,thebiasvoltagewas Ͳ40Vfortheinitialthirdofthedeposition,followedbyͲ50VforthesecondthirdandͲ40V forthelastthird.Againitcan beobservedthatincreasingthebias voltageleads to significantlyincreasedresidualcompressivestrain.Thequantitativelydeterminedstrain valuesforbothcoatingsarecomparableforͲ40VaswellasforͲ50Vbiasvoltage.Beneficial designsequences,whichstillneedtobedetermined,mightbeakeytofurtherenhance TiAlNbasedcoatings.

Fig.12:(a)Hardnessofdifferentsputteredhardcoatingsasafunctionofgrainsizeand residual stress[73].(b) Variationofbiasvoltageduring depositionofsputtered TiAlN resultinginapronouncedstrainvariation[ownwork].

(21)

17

3.2. Al2O3Coatings

Al2O3 coatings arewidelyemployed forcutting toolsdue totheir chemicalinertness, corrosionresistanceandhighhardness,alsoatelevatedtemperatures.Sincethedeposition ofthesecoatingsisstillverychallengingusingPVD,commerciallyavailableAl2O3hard coatingsoncementedcarbideinsertsaretypicallydepositedusingCVD[75–77].Theyusually exhibitamultilayeredarchitectureconsistingofafewhundrednmthickTiNadhesionlayer, followedbyaTiCNbaselayerof6to10μmthickness.AfterdepositionoftheTiCNlayer, usuallyabonding/nucleationlayerofafew100nmisdeposited,whichisfollowedbythe actualAl2O3layerof6to10μmthickness[19,50,51,78].

Al2O3existsinseveralpolymorphs(D,N,J,T,G,FandK),with ɲͲAl2O3beingthestable phase. Using CVD, usually either NͲor DͲAl2O3 are deposited due to their favorable mechanicalpropertiesandwearresistance[79].IntheearlyyearsofCVD,themetastableNͲ modification,whichexhibitsaprimitiveorthorhombicstructure,waspredominant,because thedepositionofthestableDͲAl₂O3polymorphwasnotpossible[75,80–82].Atelevated temperatures,themetastableNͲAl2O3transformsintothestableDͲAl2O3whichresultsina volumeshrinkageofapproximately7%.Consequently,socalledsecondarycracknetworks areformedwhichdeterioratecoatingcohesionandthusperformance[80,83,84].Further, severalauthorshaveshownthatnotonlytemperaturebutalsootherparameterslike precursorgasmixture,depositiontime,dopingagentsandmechanicalactivationstrongly influencetheNĺDtransformation[77,85].Thus,asignificantadvancementofCVDAl2O3

wastheintroductionof directly nucleated DͲAl₂O3byutilizationofanucleationlayer betweenTiCNandAl2O3[80,83,86,87].

TheDͲAl₂O3polymorphhasarhombohedrallycenteredhexagonallatticebelongingtothe trigonal system, which is often described as a quasiͲhexagonal close packed oxygen superlatticewithAlatomsoccupyingtwothirdsoftheoctahedralinterstices[80,83].Asa result of this structure, many properties of DͲAl2O3 are clearly anisotropic, e.g., its mechanicalproperties.ThedirectiondependentYoung´smodulusofDͲAl2O3,perpendicular tothecorrespondingsurface,calculatedfromthesinglecrystalelasticconstantsisshownin Fig.13a[88,89].The hardnessandYoung´smodulusofsingle crystallineCVDDͲAl2O3 coatings, grown epitaxially on single crystalline sapphire substrates, determined by nanoindentation measurements, are shown in Fig. 13b. Although nanoindentation measurementsdonotprovidedataforoneexclusivedirection,asthestraininducedbythe indenterismultiaxial,asignificantinfluenceofthecrystalorientationonthehardnessand Young´s modulus can be observed. Similar to Fig. 13a, the coating with the (0001)

(22)

18

orientation(cͲaxisperpendiculartocoatingsurface)showsthehighesthardnessandYoung´s modulus, followed by the (112ำ0) oriented coating (cͲaxis parallel to coating surface), whereasthe(11ำ02)orientedcoatingrevealedthelowestvalues,whichismostlikelycaused bypronouncedtwinningalongthepyramidalplane[88].

Fig.13:(a)DirectiondependentYoung´smodulusofDͲAl2O3,perpendiculartothedrawn surface,determinedfromsinglecrystalelasticconstants.(b)HardnessandYoung´smodulus ofepitaxiallygrownCVDDͲAl2O3samples[ownwork].

Thepronouncedanisotropyofthemechanicalpropertiesisnotonlyevidentinthecaseof epitaxiallygrownsinglecrystallinecoatings.Ruppietal.recentlyreportedontheanisotropy ofpolycrystallinetexturedCVDDͲAl2O3coatings[9,82,83].Theyreportedonthesuperior mechanicalpropertiesof(0001)orientedcoatingsinnanoindentationexperimentsandhow thisknowledgecanbeusedtoenhancethecoatingsperformanceduringcutting[9].Thus, thecontrolofcrystalorientationandtextureisanimportantfactortoextendthelifetimeof DͲAl2O3hardcoatingsoncuttingtools.Inapplication,coatingswitha(0001)preferred orientation have proven to performvery well, whichis in goodagreementwith the calculatedresultsandthemeasurementsonthesinglecrystallineaswellaspolycrystalline coatings,exhibitingthehighestvaluesofhardnessandelasticmodulusforthisorientation.

AnexampleofthesuperiorperformanceofatexturedCVDDͲAl₂O3hardcoatingisshownin Fig.14.TheelectronbackͲscatterdiffraction(EBSD)polefiguremeasurementsdepictedin Fig.14aindicateapronounced(0001)textureforcoatingDincontrasttocoatingEwhich exhibitsa(101ำ0)texture.TheresultsofturningtestsonsteelareshowninFig.14b, demonstratingthattheflankwearcanbereducedbyafactor2from0.3mmforcoatingEto 0.15mmforcoatingDjustbytherightchoiceofthecoatingtexture[9].

(23)

19

Fig.14:(a)EBSDpolefiguremeasurementsofa(0001)anda(101๹0)texturedCVDDͲAl2O3

coating. (b) Results of turning tests performed with those coatings, where coating D correspondsto(0001)andEto(101๹0)[9].

AnotherexampleoftheinfluenceofthetextureisshowninFig.15,whereXRDpolefigure measurementsofdifferentreflectionsareshown foraCVD DͲAl2O3coatingintheas depositedconditionandafterdryblasting.Whilenosignificantchangesduetoblastingcan beobservedforthe0001and112ำ0reflections,the11ำ02reflectionchangesconsiderably.As alreadydiscussed,thisorientationistheweakestoneofthosethreeanditstendencyto twinningmightcausethispronouncedeffectoftheblastingtreatmentontheXRDpole figure[88].Thisexampleclearlydemonstratesthatnotonlytheblastingparameterslike pressure,workingangleanddistanceorblastingmediumplayanimportantrole,butalso thecoatingtobeblastedanditsorientationhavemajorinfluenceontheresultofthe treatment.

Fig.15:XRDpolefiguremeasurementsonasdepositedanddryͲblastedCVDDͲAl2O3coatings showingadistinctinfluenceofcrystalorientationontheeffectofablastingtreatment[own work].

(24)

MichaelTkadletz Conclusion

20

4. Conclusions

Theaimofthepresentthesisistoimproveexistingtechniquesaswellastointroducenovel approachesforthecharacterizationofhardandwearresistantcoatings.Theobtainedskills providethebasisforaknowledgebasedadvancementoftheinvestigatedcoatings.The performed experiments were mainly balanced between two topics: (i) crossͲsectional characterizationmethodssuchasscanning(SEM)andtransmissionelectronmicroscopy (TEM)utilizingenergydispersiveXͲrayspectroscopyandelectronbackͲscatterdiffraction (EBSD) as well as synchrotron XͲray nanodiffraction in order to reveal the chemical compositionandmicrostructureofthecoatingsasafunctionoftheirthicknessand(ii)the determinationofthermoͲphysicalpropertiesbymeansofdifferentialscanningcalorimetry (DSC)measurementsonpowderedcoatingstorevealtheirheatcapacity,hightemperature XͲray diffraction (XRD) and biaxial stressͲtemperature measurements (BSTM) for the determinationofthecoefficientofthermalexpansionandtimeͲdomainthermoreflectance (TDTR)forthedeterminationofthethermalconductivity.

Withinafirstexampletooptimizetheavailablecharacterizationmethodstothetribological degradationof hard coatings, the influence ofdroplets on the wear behavior of arc evaporatedTiAlTaNcoatingsindryslidingcontactshasbeenilluminated.CrossͲsectional EDXmapping,extensiveSEMimagingandTEManalysisrevealedthatdropletssignificantly contributetocoatingdegradationatroomtemperatureastheydeterioratethemechanical integrityofthecoatingsandthus,aresourcesforcrackinitiation.Further,itwasfoundthat thematerialreleasedduetodisintegrationofthedropletsactsasabrasiveweardebris, consequentlyincreasingwear.Testingatelevatedtemperaturesindicatedastabilization effectduetosofteningofthedropletsandsimultaneousoxidationwithincavitiesandvoids, leadingtoaselfͲhealingbehavior.

To expand synchrotron XͲray nanodiffraction from nanocrystalline materialsto coarse grainedsamples,theconventionalsynchrotronXͲraynanodiffractionsetupwasmodified fromapointfocusedtoalinefocusedXͲraybeam.Usingthisapproachitwaspossibleto investigatemicrostructureandresidualstressofdryͲblastedCVDDͲAl2O3coatings,witha grainsizeinthemicrometerrange,asafunctionofthecoatingthickness.Theresults illuminatetheinfluenceoftwodifferentblastingmedia,i.e.withedgedandglobularshape, ontheresultingstressgradientswithinthecoatings.

Acomprehensivesurveyofadvancedcharacterizationtechniquesforthedeterminationof microstructural,mechanicalandthermoͲphysicalproperties,asithasbeenadaptedand

(25)

MichaelTkadletz Conclusion

21

optimizedwithinthisworkforthecharacterizationofhardcoatings,ispresentedinthelast part of the thesis. Methods like atom probe tomography, EBSD, synchrotron XͲray nanodiffraction and high temperature nanoindentation provide previously unrevealed insightsintomicrostructureandmechanicalpropertiesofwearresistantcoatings.Further, highͲtemperatureXRD,BSTM,DSCandTDTRprovideaccesstothermoͲphysicalproperties likecoefficientofthermalexpansion,heatcapacityandthermalconductivity.Thenow availablespectrumofadvancedcharacterizationtechniquesprovidesakeytoolͲboxforthe furtheroptimizationofhardcoatings.

(26)

MichaelTkadletz References

22

5. References

[1] V.Derflinger,H.Brändle,H.Zimmermann,Surf.Coat.Technol.113(1999)286–292.

[2] W.Kalss,A.Reiter,V.Derflinger,C.Gey,J.L.Endrino,Int.J.Refract.Met.HardMater.

24(2006)399–404.

[3] F.Klocke,T.Krieg,CIRPAnn.ͲManuf.Technol.48(1999)515–525.

[4] S.G.Harris,E.D.Doyle,A.C.Vlasveld,P.J.Dolder,Surf.Coat.Technol.146Ͳ147(2001) 305–311.

[5] G.S.FoxͲRabinovich,G.C.Weatherly,A.I.Dodonov,A.I.Kovalev,L.S.Shuster,S.C.

Veldhuis,G.K.Dosbaeva,D.L.Wainstein,M.S.Migranov,Surf.Coat.Technol.177Ͳ178 (2004)800–811.

[6] A.Buranawong,N.WititͲAnun,S.Chaiyakun,A.Pokaipisit,P.Limsuwan,ThinSolid Films519(2011)4963–4968.

[7] A.Inspektor,P.A.Salvador,Surf.Coat.Technol.257(2014)138–153.

[8] P.H.Mayrhofer,R.Rachbauer,D.Holec,F.Rovere,J.M.Schneider,in:S.Hashmi(Ed.), ComprehensiveMaterialsProcessing,Elsevier,Oxford,2014.

[9] S.Ruppi,Surf.Coat.Technol.202(2008)4257–4269.

[10] K.ͲD.Bouzakis,F.Klocke,G.Skordaris,E.Bouzakis,S.Gerardis,G.Katirtzoglou,S.

Makrimallakis,Wear271(2011)783–791.

[11] K.ͲD.Bouzakis,S.Gerardis,G.Skordaris,G.Katirtzoglou,S.Makrimallakis,F.Klocke,E.

Bouzakis,Surf.Coat.Technol.204(2009)1081–1086.

[12] K.ͲD. Bouzakis, E. Bouzakis, G. Skordaris, S. Makrimallakis, A. Tsouknidas, G.

Katirtzoglou,S.Gerardis,Surf.Coat.Technol.205(2011)128–132.

[13] R.Daniel,J.Keckes,I.Matko,M.Burghammer,C.Mitterer,ActaMater.61(2013) 6255–6266.

[14] R.Daniel,D.Holec,M.Bartosik,J.Keckes,C.Mitterer,ActaMater.59(2011)6631–

6645.

[15] A.Zeilinger,R.Daniel,T.Schöberl,M.Stefenelli,B.Sartory,J.Keckes,C.Mitterer, ThinSolidFilms581(2015)75–79.

[16] M.RebelodeFigueiredo,M.D.Abad,A.J.Harris,C.Czettl,C.Mitterer,P.Hosemann, ThinSolidFilms578(2015)20–24.

(27)

23

[17] D.G.Cahill,S.ͲM.Lee,T.I.Selinder,J.Appl.Phys.(1998)5783Ͳ5786.

[18] J.Keckes,M.Bartosik,R.Daniel,C.Mitterer,G.Maier,W.Ecker,J.VilaͲComamala,C.

David,S.Schoeder,M.Burghammer,ScriptaMater.67(2012)748–751.

[19] M.Tkadletz,J.Keckes,N.Schalk,I.Krajinovic,M.Burghammer,C.Czettl,C.Mitterer, Surf.Coat.Technol.262(2015)134–140.

[20] F.Kone,C.Czarnota,B.Haddag,M.Nouari,Surf.Coat.Technol.205(2011)3559–

3566.

[21] R.F.Bunshah,DepositionTechnologiesforFilmsandCoatings,NoyesPublications, NewJersey,1982.

[22] F.ͲW.Bach,T.Duda,ModerneBeschichtungsverfahren,WileyͲVCH,Weinheim,2000.

[23] R.A.Haefer,OberflächenͲundDünnschichttechnologie,TeilI:Beschichtungenvon Oberflächen,Springer,BerlinHeidelberg,1987.

[24] Clear Metals Inc. Ͳ Technology: Metal Oxide Deposition, http://clearmetalsinc.com/here/wpͲcontent/uploads/SputterͲillo.png,accessedJune 2015.

[25] M.Ohring,TheMaterialsScienceofThinFilms–DepositionandStructure,Academic Press,SanDiego,2002.

[26] B. Rother, J. Vetter, Plasmabeschichtungsverfahren und Hartstoffschichten, DeutscherVerlagfürGrundstoffindustrie,Leipzig,1992.

[27] H. Frey, Vakuumbeschichtung 1, Plasmaphysik, Plasmadiagnostik, Analytik, VDI, Düsseldorf,1995.

[28] W.D.Westwood,SputterDeposition,AVS,NewYork,2003.

[29] R.L. Boxman,D.M. Sanders,P.J.Martin,Handbook ofVacuumArcScienceand Technology:FundamentalsandApplications,NoyesPublications,NewJersey,1995.

[30] W.Kern,J.L.Vossen,ThinFilmProcessesII,AcademicPress,SanDiego,2012.

[31] D.M.Sanders,A.Anders,Surf.Coat.Technol.133Ͳ134(2000)78–90.

[32] H.K.Pulker,Vak.Forsch.undPrax.11(1999)25–25.

[33] A.Hörling,L.Hultman,M.Odén,J.Sjölén,L.Karlsson,J.Vac.Sci.Technol.A20(2002) 1815.

[34] U.Helmersson,M.Lattemann,J.Bohlmark,A.P.Ehiasarian,J.T.Gudmundsson,Thin SolidFilms513(2006)1–24.

(28)

24

[35] R.L.Boxman,S.Goldsmith,Surf.Coat.Technol.52(1992)39–50.

[36] HCVAC Ͳ Dongguan Huicheng Vacuum Technology Co., Ltd.: Arc Evaporation, http://www.hcvacuum.com/node/66,accessedJune2015.

[37] D.M.Mattox, HandbookofPhysicalVaporDeposition(PVD)Processing,William Andrew,Oxford,2010.

[38] P.H.Mayrhofer,C.Mitterer,L.Hultman,H.Clemens,Prog.Mater.Sci.51(2006) 1032–1114.

[39] H.A.Jehn,B.Rother,Int.J.Refract.Met.HardMater.14(1996)87–95.

[40] J.Kohlscheen,H.R.Stock,P.Mayr,Surf.Coat.Technol.120Ͳ121(1999)740–745.

[41] N.Schalk,SynthesisandPostͲtreatmentofHardCoatingsforCementedCarbide CuttingTools,PhDThesis,MontanuniversitätLeoben,2013.

[42] C.Czettl,DesignofCVDCoatingsforCuttingTools,PhDThesis,Montanuniversität Leoben,2013.

[43] K.L.Choy,Prog.Mater.Sci.48(2003)57–170.

[44] H.O.Pierson,HandbookofChemicalVaporDeposition(CVD),WilliamAndrew,New York,1999.

[45] H.Lüth,SolidSurfaces,InterfacesandThinFilms,Springer,Berlin,Heidelberg,2010.

[46] B.A.Movchan,A.VDemchishin,Phys.Met.Met.28(1969)653–660.

[47] J.A.Thornton,J.Vac.Sci.Technol.11(1974)666–670.

[48] R.Messier,J.Vac.Sci.Technol.A2(1984)500–503.

[49] A.Anders,ThinSolidFilms518(2010)4087–4090.

[50] A.Riedl,N.Schalk,C.Czettl,B.Sartory,C.Mitterer,Wear289(2012)9–16.

[51] N.Schalk,C.Mitterer,C.Czettl,B.Sartory,M.Penoy,C.Michotte,Tribol.Lett.52 (2013)147–154.

[52] M.Klaus,C.Genzel,H.Holzschuh,ThinSolidFilms517(2008)1172–1176.

[53] K.ͲD.Bouzakis,G.Skordaris,I.Mirisidis,N.Michailidis,G.Mesomeris,E.Pavlidou,G.

Erkens,Surf.Coat.Technol.200(2005)1879–1884.

[54] K.ͲD.Bouzakis,G.Skordaris,F.Klocke,E.Bouzakis,Surf.Coat.Technol.203(2009) 2946–2953.

(29)

25

[55] C.Barbatti,J.Garcia,R.Pitonak,H.Pinto,A.Kostka,A.DiPrinzio,M.H.Staia,A.R.

Pyzalla,Surf.Coat.Technol.203(2009)3708–3717.

[56] P.H.Mayrhofer,C.Mitterer,L.Hultman,H.Clemens,Prog.Mater.Sci.51(2006) 1032–1114.

[57] M.Stefenelli,J.Todt,A.Riedl,W.Ecker,T.Müller,R.Daniel,M.Burghammer,J.

Keckes,J.Appl.Cryst.46(2013)1378–1385.

[58] M.Bartosik,R.Pitonak,J.Keckes,Adv.Eng.Mater.13(2011)705–711.

[59] K.Schiffner,C.Drostegen.Helling,Comput.Struct.72(1999)329–340.

[60] P.H.Mayrhofer,A.Hörling,L.Karlsson,J.Sjölén,T.Larsson,C.Mitterer,L.Hultman, Appl.Phys.Lett.83(2003)2049–2051.

[61] P.H.Mayrhofer,F.Kunc,J.Musil,C.Mitterer,ThinSolidFilms415(2002)151–159.

[62] N.Schalk,C.Mitterer,I.LetofskyͲPapst,C.Czettl,B.Sartory,M.Penoy,C.Michotte, Tribol.Int.67(2013)54–60.

[63] S.PalDey,S.C.Deevi,Mater.Sci.Eng.A361(2003)1–8.

[64] M.Durandurdu,J.Phys.Chem.Solids69(2008)2894–2897.

[65] M.Kawate,A.Kimura,T.Suzuki,J.Vac.Sci.Technol.A20(2002)569–571.

[66] I.W.Park,D.S.Kang,J.J.Moore,S.C.Kwon,J.J.Rha,K.H.Kim,Surf.Coat.Technol.201 (2007)5223–5227.

[67] E.Spain,J.C.AvelarͲBatista,M.Letch,J.Housden,B.Lerga,Surf.Coat.Technol.200 (2005)1507–1513.

[68] W.D.Cho,H.Y.Sohn,M.Minerals,M.S.Meeting,ValueͲAdditionMetallurgy,TMS, SanAntonio,1998.

[69] A.Kimura,H.Hasegawa,K.Yamada,T.Suzuki,Surf.Coat.Technol.120Ͳ121(1999) 438–441.

[70] A.Hörling,L.Hultman,M.Odén,J.Sjölén,L.Karlsson,Surf.Coat.Technol.191(2005) 384–392.

[71] M.Pfeiler,C.Scheu,H.Hutter,J.Schnöller,C.Michotte,C.Mitterer,M.Kathrein,J.

Vac.Sci.Technol.A27(2009)554–560.

[72] M.Pfeiler,G.A.Fontalvo,J.Wagner,K.Kutschej,M.Penoy,C.Michotte,C.Mitterer, M.Kathrein,Tribol.Lett.30(2008)91–97.

(30)

26

[73] P.H.Mayrhofer,C.Mitterer,J.Musil,Surf.Coat.Technol.174Ͳ175(2003)725–731.

[74] R.Daniel,A.Zeilinger,T.Schöberl,B.Sartory,C.Mitterer,J.Keckes,J.Appl.Phys.117 (2015)235301.

[75] S.Ruppi,J.Phys.IV11(2001)847–859.

[76] A.Osada,E.Nakamura,H.Homma,T.Hayahi,T.Oshika,Int.J.Refract.Met.Hard Mater.24(2006)387–391.

[77] D.Hochauer,C.Mitterer,M.Penoy,C.Michotte,H.P.Martinz,M.Kathrein,Surf.

Coat.Technol.204(2010)3713–3722.

[78] D.Hochauer,C.Mitterer,M.Penoy,C.Michotte,H.P.Martinz,M.Kathrein,Surf.

Coat.Technol.203(2008)350–356.

[79] I.Levin,D.Brandon,J.Am.Ceram.Soc.81(1998)1995–2012.

[80] S.Vuorinen,L.Karlsson,ThinSolidFilms214(1992)132–143.

[81] S.Ruppi,A.Larsson,ThinSolidFilms388(2001)50–61.

[82] S.Ruppi,A.Larsson,A.Flink,ThinSolidFilms516(2008)5959–5966.

[83] S.Ruppi,Int.J.Refract.Met.HardMater.23(2005)306–316.

[84] J.Skogsmo,M.Halvarsson,S.Vuorinen,Surf.Coat.Technol.54Ͳ55(1992)186–192.

[85] M.Kathrein,W.Schintlmeister,W.Wallgram,U.Schleinkofer,Surf.Coat.Technol.

163Ͳ164(2003)181–188.

[86] A.A.Layyous,D.M.Freinkel,R.Israel,Surf.Coat.Technol.56(1992)89–95.

[87] M.Halvarsson,H.Nordén,S.Vuorinen,Surf.Coat.Technol.61(1993)177–181.

[88] V.Pishchik,L.A.Lytvynov,E.R.Dobrovinskaya,Sapphire:Material,Manufacturing, Applications,Springer,NewYork,2009.

[89] J.F. Nye, Physical Properties of Crystals: Their Representation by Tensors and Matrices,ClarendonPress,Oxford,1985.

(31)

MichaelTkadletz Publications

27

6. Publications

6.1. Listofincludedpublications

I. TheeffectofdropletsinarcevaporatedTiAlTaNhardcoatingsonthewear behavior

M.Tkadletz,C.Mitterer,B.Sartory,I.LetofskyͲPapst,C.Michotte SurfaceandCoatingsTechnology257(2014)95Ͳ101.

II. ResidualstressgradientsinɲͲAl2O3hardcoatingsdeterminedby pencilͲbeamXͲraynanodiffraction:Theinfluenceofblastingmedia MichaelTkadletz,JozefKeckes,NinaSchalk,IvanKrajinovic,Manfred Burghammer,ChristophCzettl,ChristianMitterer

SurfaceandCoatingsTechnology262(2015)134Ͳ140.

III. Advancedcharacterizationmethodsforwearresistanthardcoatings:Areview onrecentprogress

MichaelTkadletz,NinaSchalk,RostislavDaniel,JozefKeckes,ChristophCzettl, ChristianMitterer

InvitedreviewsubmittedtoSurfaceandCoatingsTechnology.

6.2. Publicationsrelatedtothiswork

IV. Restrictionsofstressmeasurementsusingthecurvaturemethodbythermally inducedplasticdeformationofsiliconsubstrates

ChristianSaringer,MichaelTkadletz,ChristianMitterer SurfaceandCoatingsTechnology274(2015)68Ͳ75.

(32)

28

6.3. Mycontributiontotheincludedpublications PublicationI

Forthefirstpublication,thecoatingsampleswereprovidedbytheprojectpartner Ceratizit Austria GmbH. I performed the ball on disk tests, nanoindentation measurementsandXͲraydiffractionmeasurements.Further,Ievaluatedthescanning electronmicroscopyinvestigations,whichwereperformedwiththehelpofBernhard Sartory.Theinterpretationofthetransmissionelectronmicroscopyinvestigations,which wereperformedwithIlseLetofskyͲPapst,wasalsodonebymyself.Theinterpretationof theobtainedresultsaswellasthedevelopmentofthepublicationconceptandwritingof thismanuscriptwasdonebymyown.

PublicationII

Alsoforthispublication,thecoatingsampleswereprovidedbyCeratizitAustriaGmbH.I performedtheXͲraydiffractionmeasurementsaswellasthesynchrotronmeasurements attheID13beamlinelocatedattheESRFinGrenoble,France.Theevaluationofthe synchrotrondatawassupportedbyProf.JozefKeckes.SEMinvestigationsaswellasthe EBSD measurements were performed together with Bernhard Sartory. The FEM simulations were provided by IvanKrajinovic. The development ofthemanuscript concept,planningandwritingwasdonebymyself.

PublicationIII

Withinthispublication,acomprehensiveoverviewoncharacterizationtechniquesfor hardcoatingsisprovided.TheconceptionandplanningwasdonetogetherwithProf.

ChristianMittererandDr.NinaSchalk.Iperformedtheextensiveliteratureresearchand collectionofresultsofthedifferenttechniques.Partsoftheprevioustwopublications areincludedwithinthispublicationrepresentingthesuccessofthedevelopmentand improvementofcharacterizationtechniqueswithinthisthesis.Further,Idepositedthe PVDcoatingsforthechapterdiscussingthethermoͲphysicalpropertiesmyself.TheHTͲ XRDmeasurementswereperformedbyDr.ManfredWießner(MaterialsCenterLeoben) andIevaluatedthedataandcalculatedthecoefficientsofthermalexpansion.The annealingtreatmentsforthethermalconductivityserieswereperformedbymyselfand the timeͲdomain thermoreflectancemeasurements were performed byDr. Markus Winkler (Fraunhofer Institute for Physical Measurement Techniques IPM, Freiburg, Germany).ForthemeasurementsofthespecificheatcapacitybyDSC,Ipreparedthe powdersafterdepositionandcarriedoutandevaluatedtheexperiments.Finally,Ialso wrotethemanuscript.

(33)

29

Conception and

planning

Experiments Analysis and interpretation

Manuscript preparation

PublicationI 100 80 100 100

PublicationII 100 100 90 100

PublicationIII 100 30 100 90

Supervisionnotincluded.

(34)

MichaelTkadletz PublicationI

30

PublicationI

TheeffectofdropletsinarcevaporatedTiAlTaNhard coatingsonthewearbehavior

M.Tkadletz,C.Mitterer,B.Sartory,I.LetofskyͲPapst,C.Czettl,C.Michotte

SurfaceandCoatingsTechnology257(2014)95Ͳ101

(35)

31

TheeffectofdropletsinarcevaporatedTiAlTaN hardcoatingsonthewearbehavior

M.Tkadletz¹,C.Mitterer²,B.Sartory¹,I.LetofskyͲPapst³,C.Czettl⁴,C.Michotte⁵

1MaterialsCenterLeobenForschungGmbH,Roseggerstraße12,AͲ8700Leoben,Austria

2DepartmentofPhysicalMetallurgyandMaterialsTesting,Montanuniversität,FranzͲJosefͲ Straße18,AͲ8700Leoben,Austria

3InstituteforElectronMicroscopyandNanoanalysis,GrazUniversityofTechnology, Steyrergasse17,AͲ8010Graz,Austria

4CERATIZITAustriaGmbH,MetallwerkͲPlanseeͲStraße71,AͲ6600Reutte,Austria

5CERATIZITLuxembourgSàrl,RoutedeHolzem,B.P.51,LͲ8201Mamer,Luxembourg

Abstract

Hardcoatingsdepositedbycathodicarcevaporationareoftencharacterizedbydroplets, deteriorating their surface roughness and oxidation resistance. Within this work, the responseofthesedropletstotribologicalloadingandtheircontributiontothetribological systemwasinvestigated.BallͲonͲdisktestsagainstAl2O3counterpartsweredoneonTiAlTaN coatedcementedcarbidedisksatroomtemperatureand700°C.SurfacesaswellascrossͲ sectionsthroughtheweartrackswereinvestigatedbyscanningandtransmissionelectron microscopy.Whiledropletswerefoundtocontributetocoatingdegradationbyproviding nucleationsitesforshearcracksandbythereleaseofabrasivefragmentsintothesliding contact,oxidationatelevatedtemperatureleadstoamoreefficientembeddingintothe surrounding coating matrix. This suggests an oxidation induced selfͲhealing process, contributingtotheenhancedwearresistanceofTiAlNͲbasedhardcoatingsatelevated temperatures.

Keywords:

Hardcoatings,TiAlTaN,arcevaporation,droplets,macroparticles,tribologicalproperties

(36)

32

1.Introduction

Thewearbehaviorofhardcoatingsplaysanimportantroleforcoatedcuttingtoolsaswear directlyaffectsthetoollifetime.Forcuttinginserts,thewearresistanceisoftenevaluatedby millingorturningtests,whichprovideinformationabouttoollifetimeandperformanceof differentcoatings.Thesefieldtestsincludealargenumberofvariableslikecuttingspeed, cuttingforceandcuttingtemperature,whichareoftennotknowntoasufficientextent.

Hence, itisdifficulttorelateacertainobservedcuttingorwearbehaviorto specific propertiesofparticularcoatingsystems[1Ͳ5].Anotherpossibilityistoperformtests,which arereducedtoasmallnumberofcontrolledvariables,likeballͲonͲdisktests[6].Thosegive detailedreproducibleinformationaboutthetribologicalpropertiesofdifferentcoating systems. Selected variations of their properties, for example due to changes in the depositionprocess[7,8],canbecomparedtoeachother.WearͲresistanthardcoatingsare often deposited by cathodic arc evaporation, with its inherent benefits of increased deposition and high ionization rates compared to sputtering. However, cathodic arc evaporatedcoatingsareoftencharacterizedbydroplets,whichnegativelyaffectthesurface roughness, the corrosion and oxidation stability (as they act as diffusion paths), and subsequentlythetribologicalproperties[9Ͳ11].Uptonow,acomprehensiveunderstanding oftheireffectonthewearbehaviorofthecoating–bothincuttingaswellasintribological tests–isessentiallymissing.

PreviousreportshaveshownthattheadditionofsmallamountsofTatotheTiAlNsystemdo notchangethecrystalstructureofthecoatingsignificantlybutitsmechanicalproperties,the decomposition temperature and oxidation resistance are significantly increased [12].

Moderncuttingoperationsareknowntoresultinveryhightemperaturesonthetools cuttingedge[13]. Hence, withinthisworkballͲonͲdisktestswereperformedatroom temperature(RT)and700°Concementedcarbidediskscoatedbycathodicarcevaporation with a stateͲofͲtheͲart commercial TiAlTaN coatingtogainknowledge aboutthehigh temperaturebehavioroftheusedcoatingsystem.ThesurfaceaswellascrossͲsectionsof theweartrackspreparedbyfocusedionbeam(FIB)techniqueswereinvestigatedusing scanningandtransmissionelectronmicroscopy(SEM,TEM).Specialemphasiswaslaidon the role of droplets in the performed ballͲonͲdisk tests, where possible mechanisms triggeringcoatingdegradationhavebeendetermined.

(37)

33

2.ExperimentalDetails

Thecoatinginvestigatedwithinthisworkwasdepositedonpolishedcementedcarbidedisks (30u4mm)consistingof77wt%tungstencarbide,12wt%mixedcarbidesand11wt%

cobalt.AnindustrialͲscalecathodicarcevaporationsystemoftypeOerlikonBalzersRCS usingadepositionprocessaccordingtoPfeileretal.[14]wasused.Thecoatingconsistsofa a0.4PmthickadhesionlayerofTiNandaa2.8PmthickTiAlTaNtoplayerwithacomposition ofTi0.38Al0.58Ta0.04N[14].

BallͲonͲdisktestswereperformedatRTand700°CwithaCSMhightemperaturetribometer applyingaloadof5Nforaslidingdistanceof300m,aweartrackradiusof7mmandAl2O3 balls(Ø6mm)ascounterparts.Thetestswereperformedatambientatmospherewitha relativehumidityof25%.WeartrackprofilesweredeterminedusingaNanofocusPsurf whiteͲlightconfocalinterferometer.XͲraydiffraction(XRD)measurementswereperformed utilizingaBrukerAdvanceD8diffractometeringrazingincidencemodewithaninclining angleof2°from20to85°withastepsizeof0.02°andameasuringtimeof1.2susing copperKDradiation.Fordeterminationofthecoatinghardness,aUMISnanoͲindenterfrom FischerͲCripps Laboratories was used. Due to the rather high roughness of the arc evaporatedcoatings,asmallareaofthesurfacewascarefullypolishedandplateautests applyingaforcedecreasingfrom25to5mNinstepsof1mNleadingtoanindentation depthof75to200nmwereperformed.Toachievereasonablestatistics,atleast16indents oneachsamplewereusedtodeterminethemechanicalproperties.Surfaceexaminationsas wellasthecrossͲsectioninvestigationswereperformedwithaZeissAurigaCrossbeamfield emissiongunSEM.CrossͲsectionsaswellasTEMlamellaswerepreparedwithanOrsay PhysicsCobraZͲ05FIBextension.EnergyͲdispersiveXͲrayspectroscopy(EDS)measurements wereperformedwithanEDAXApollo40+detector.AllTEMimaging,TEMEDSandenergy filteringTEM(EFTEM)elementaldistributioninvestigationswereperformedwith aFEI TecnaiF20TEMequippedwithafieldemissionguncombinedwithaquantumenergyfilter.

Theselectedareaelectrondiffraction(SAED)measurementswereconductedwithaFEIT12 TEMequippedwithaLaB6cathodeandanaccelerationvoltageof120kV.

(38)

34

3.Results

TheperformedballͲonͲdisktestsrevealthatthecoefficientoffrictionisdecreasingwith increasingtemperaturefrom0.8to0.5(seeFigs.1aandc).Figs.1banddindicatethatalso themaximumweartrackdepthisdecreasingfrom1.5to0.75Pm.Thiscorrespondstoa decreaseofthewearcoefficientofalmostoneorderofmagnitudefrom1.36u10^Ͳ14to 2.27u10^Ͳ15 m³/Nm. Decreasing friction and wear coefficients with increasing testing temperatureforTiAlNͲbasedcoatingswerepreviouslyalsoreportedbyKutschejetal.[15]

andPfeileretal.[7],respectively.Similarbehaviorwasreportedforothercoatingsystems likeAlSiTiN[16,17]whiletheseeffectsdonotappearwhenCrAlNorCrAlSiNistested[18].

However,withinthisworkspecialattentionispaidtotheresponseofdropletstomechanical loadingandtheircontributiontothetribologicalsystem.

Fig.1:Coefficientoffrictionasafunctionoftheslidingdistanceand2Dprofileofthewear trackobtainedfortribologicaltestingatRT(a,b)and700°C(c,d).

Toshedlightonthewearbehaviorandthustodeterminepossiblemechanismstriggering coatingdegradation,surfaceandcrossͲsectionalSEMinvestigationsontheweartracksof thetestedsampleswereperformed.AfterRTtesting,comettaillikefeaturesoriginating fromdropletswerefoundinsidetheweartracks.Arepresentativecomettailismarkedby

(39)

35

thewhitearrowinFig.2a,whileadetailedsurfaceviewofadropletwithintheweartrackon asampletestedat700°CisshowninFig.2b.Infront(withrespecttotheslidingdirection)of thedropletinFig.2b,cracksarevisible(indicatedbywhitearrows).Thosecracksarerelated totensilestressescausedbythefrictionforceactingduringtheballͲonͲdisktest.This corroboratestheassumptionthatdropletssignificantlycontributetothedegradationofthe coatingmaterialinslidingcontacts.

Fig.2:TopviewofaRTweartrack,slidingdirectionfromlefttoright(a).Acomettail, originatingfromadroplet,canbeseenandisindicatedbythewhitearrow.Ahigher magnificationofadroplet(b)exhibitssurface(tensile)crackswhicharemarkedbythewhite arrows.

Toilluminatethisdamagebehaviorfurther,thenatureofthedropletswithouttribological loadinghasbeencharacterized.Thetopviewofapristinecoatingsurface,i.e.outsidethe weartrackofasampletestedatRT,includingarepresentativedropletisshowninFig.3a.

Figure3bshowsacrossͲsectionalviewofthesamedropletwithasuperimposedEDS mapping.Fromboth,thetopͲaswellasthecrossͲsectionalview,adropletdiameterof a6Pmcanbeobtained.TheEDSmappinginFig.3brevealsthatthecoreregionofthe dropletmainlyconsistsoftitanium.Withinthismetalliccore,brighteranddarkerareasare visible,correspondingtolocaldifferencesinchemicalcomposition.Ontopofthecore, coating material can be seen which has overgrown the droplet during the ongoing depositionprocess.Ontheverytop,aplatinumlayerisvisible,whichwasdepositedto obtain a smooth FIB cut. Droplets with essentially metallic character and a chemical compositioncorrespondingtotheusedtargetmaterialwerepreviouslyreportedbyseveral authors[19Ͳ22].Theirshapeandsizehasbeenreportedtobedeterminedwhetherthey arriveatthesubstrateinamoltenorsemiͲmoltencondition[19Ͳ21].

(40)

36

Fig.3:Topviewofatypicaldropletintheasdepositedstateoutsidetheweartrack(a).The crossͲsectionofthesamedropletincludinganEDSmappingisshownin(b)indicatingthe metallicTicore,thecoatingmaterialandaPtlayerwhichispromotingasmootherFIBcut.

Within the wear track, droplets with similar structure, shape and size were found.

Dependingonthegeometryandappearanceofthedroplets,threedifferentpossibilitiesof theireffectoncoatingwearcouldbedistinguished.Inthefirstcase,(i),thedropletisstable andnoevidenceofdropletͲinitiatedcoatingdegradationcanbefound.Thesecondcase,(ii), isillustratedinFigs.4aandb.ThetopviewSEMimage(Fig.4a)revealsnoevidenceof obviouscoatingdamagewiththeexceptionofmaterialaccumulationinfrontofthedroplet withrespecttotheslidingdirection.Incontrast,inthecrossͲsectionalviewofthesame droplet(Fig.4b)crackscanbeseenbehindthedroplet(againwithrespecttothesliding direction).AcloserviewofthecrackisprovidedbyFig.4c,wherethedirectionofthecrack suggeststhatitpropagatesduetoshearstresses.Thethirdcaseofcoatingdegradation,(iii), illustratedinFig.4d,isrelatedtodroplets,whicharemechanicallynotstableandcollapse duetotheloadappliedbytheballͲonͲdisktest.Inthiscase,fragmentsofthecollapsed dropletarereleasedintotheweartrackformingweardebris.Thisincreasedamountofwear debrisinsidetheweartrack,actingasabrasivemedium,isexpectedtosignificantlyincrease coatingwear.Thedropletscollapsingwithinthecoatingmatrixgeneratecavities,whichare subsequentlyfilledwithaccumulatedandcompactedweardebris(seethefeaturelessarea abovetheremainingpartofthedropletinFig.4d).Thesamematerialaccumulationcanbe foundwithinsurfacecavitiesforallthreecasesorifobstaclesarepresentwhichhinderthe flowofweardebriswithintheslidingcontact.

(41)

37

Fig.4:TopviewofaRTsampledropletinsidetheweartrack(slidingdirectionfromleftto right)includingmaterialaccumulationinfrontofthedroplet(a).ThecrossͲsectionofthe samedroplet(b)showstheaccumulationinthefrontandacrackattheendofthedroplet.A closerviewofthesubsurface(shear)crackisshownin(c)andacollapseddropletfilledwith accumulatedmaterialisshownin(d).

Tocharacterize the material accumulations (seeFig.4d) and the areas with differing brightnesswithinthedropletcore(seeFig.3b),TEMlamellasthroughtheweartrackofa coatingsamplewithadroplettestedatRTandat700°C,respectively,wereprepared.

Figure5ashowsabrightͲfield(BF)TEMmicrographofadropletwithintheweartrackofthe RTsample.Figure5btodpresentthecorrespondingEFTEMelementaldistributionsfor titanium, nitrogen and oxygen. The BFmicrograph inFig.5a shows,beneath the top platinumlayer,thecoatingwhichhasovergrownthedropletcore.Thisareaischaracterized by compositional variations originating from substrate rotation during the deposition process[23],whichisbestseeninthetitaniumandnitrogenmappingsinFigs.5bandc.Also theTiNadhesionlayerclosetothecementedcarbidesubstrateisclearlyseen.Thetitanium andnitrogenmappingsinFigs.5bandcconfirmthatthedropletcoremainlyconsistsof titanium.Thenitrogendistribution(Fig.5c)indicatesthattheareaswithdifferingbrightness withinthedropletcorecorrespondtoareaswithmoreorlessnitrogen.Theyareassumedto originfromthedifferentnitridingbehaviorofliquidorsemiͲliquiddropletsinthereactive gasatmosphereduringtheirtransportfromthetargettothesubstrate.Partiallynitrided