Characterization of DLC coatings over nitrided stainless steel with and without nitriding pre-treatment using annealing cycles

(1)

w w w . j m r t . c o m . b r

Availableonlineatwww.sciencedirect.com

Original

Article

Characterization

of

DLC

coatings

over

nitrided

stainless

steel

with

and

without

nitriding

pre-treatment

using

annealing

cycles

Eugenia.

L. Dalibón

a,∗

_,

_Thierry

_Czerwiec

b

_,

_Vladimir

_J.

_{Trava-Airoldi}

c

_,

_Naureen

_Ghafoor

d

_,

Lina

Rogström

d

_,

_Magnus

_Odén

d

_,

_Sonia

_P.

_Brühl

a

a_Surface_Engineering_Group,_Universidad_Tecnológica_Nacional_(UTN-FRCU),_Ing._Pereira_676,_E3264BTD_Concepción_del_Uruguay,

Argentina

b_Institut_Jean_Lamour_UMR_7198,_CNRS,_Université_de_Lorraine,_Dept._CPSS,₅₄₀₁₁_Nancy,_France

c_Instituto_Nacional_de_Pesquisas_Espaciáis_(INPE),_Av._dos_Astronautas_1758,_12.227-010_São_José_dos_Campos,_SP,_Brazil

d_{Nanostructured}_Materials,_Department_of_Physics,_Chemistry_and_Biology_(IFM),_Linköping_University,_SE-581₈₃_Linköping,_Sweden

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received24June2018 Accepted1December2018 Availableonline12January2019

Keywords: DLCcoatings Thermalstability Duplextreatment Nitriding

a

b

s

t

r

a

c

t

Amorphoushydrogenateddiamond-likecarbon(DLC)coatingsweredepositedusingplasma assistedchemicalvapourdeposition(PACVD)onprecipitationhardening(PH)stainlesssteel. Plasmanitridinghasbeenusedaspre-treatmenttoenhanceadhesionandmechanical prop-erties.ChemicalandmechanicalpropertiesofDLCcoatingsaredependentonthehydrogen contentandsoontherelationbetweensp3_/sp2_bondings._The_bondings_and_the_structure

oftheDLCﬁlmchangewithtemperature.Inthiswork,astudyofthethermaldegradation andtheevolutionofthemechanicalpropertiesofDLCcoatingsoverPHstainlesssteelhave beencarriedout,includingtheeffectofanadditionalnitridedlayer.

Nitridedandnon-nitridedsteelsamplesweresubjectedtothesamecoatedinthesame conditions,andtheyweresubmittedtothesamethermalcycles,heatingfromroom tem-peratureto600◦Cinseveralsteps.

Aftereachcycle,Ramanspectraandsurfacetopographymeasurementswereperformed and analyzed.Nanohardnessmeasurements and tribologicaltests, using a pin-on-disc machine,werecarriedouttoanalyzevariationsinthefrictioncoefﬁcientandthewear resistance.

Theduplexsample,withnitridingaspre-treatmentshowedabetterthermalstability. Forduplexsample,thecoatingproperties,suchasadhesion,andfrictioncoefﬁcientwere sustainedafterannealingathighertemperatures;whereasitwasnotthecaseforonlycoated sample.

∗ _{Corresponding}_author..

E-mails: dalibone@frcu.utn.edu.ar(E.L. Dalibón), thierry.czerwiec@univ-lorraine.fr (T. Czerwiec),vladimir@las.inpe.br (V.J. Trava-Airoldi),naugh@ifm.liu.se(N.Ghafoor),linro@ifm.liu.se(L.Rogström),magod@ifm.liu.se(M.Odén),sonia@frcu.utn.edu.ar(S.P.Brühl).

https://doi.org/10.1016/j.jmrt.2018.12.002

2238-7854/©2018BrazilianMetallurgical,MaterialsandMiningAssociation.PublishedbyElsevierEditoraLtda.Thisisanopenaccess articleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).

(2)

1. Introduction

Despitetheiramorphouscharacteristicsandthepresenceof hydrogencarbonbondings,diamond-likecarbon(DLC) coat-ingshaveuniquepropertiessimilartodiamondsuchasvery highhardness,chemicalinertia,andaverylowfriction coef-ﬁcientcomparedtootherhardcoatings.

DLCcoatingscanbedepositedviaplasma-assisted tech-niques,from a solidcarbon target,such asplasmavapour deposition (PVD) or via hydrocarbon precursors by plasma assisted chemical vapour deposition (PACVD). In the latter case,thepresenceofhydrogenintheﬁlmcannotbeavoided, andthechemicalbondingsarenotonlyC CbutalsoC H, withdifferentproportionsofsp3_and_sp2_bondings,_depending

onplasmadepositionparameterssuchastime,temperature andgasflow[1–3].Inawiderangeofstructures,thesefilms havepropertiessimilartodiamondsuchashighhardness,a verysmoothsurfacewithalowcoefficientoffriction(below 0.1)andverygoodwearandcorrosionprotectionproperties. However,thesepropertiesarenotonlyaffectedbyfilm struc-turebutalsobythetribosystemproperties:hardnessprofile, adhesion,mechanicalsupportandstresses.Forthesereasons nitridinghasbeenusedaspre-treatment,especiallyoversoft substratessuchasstainlesssteel[4–6].

DuplextreatmentscombiningnitridingandDLCcoatings canbeusedtoincreasethemechanicalpropertiesand tribo-logicalperformancebyincreasingtheloadbearingcapacityof theso-formedcomposite.Suchduplextreatmentswith nitrid-ingareoftenappliedonmediumalloysteelsassubstrates[7,8]. Itwasreportedthatagoodwearresistancewasobtainedwith theDLCdepositedoverahardandhomogeneouscompound layer[8,9].Althoughotherauthorsfoundthatthebest load-bearingcapacitywasachievedforanitridedlayerprimarily constitutedbyadiffusionlayer[7].

In the case of stainless steels, where only a nitrogen-expandedphasearisesafterlowtemperaturenitriding,some paperswerepreviouslypublished[10–12].

DLCcoatingsarefrequentlyusedinapplicationsforwhich there may be high work temperature or localized heating causedbyfriction.Hencedetailedstudiesoftheevolutionof coatingpropertieswiththermal solicitationsare necessary, forexample,touseitinenginecomponents[13,14].Several studiesaboutthethermalstabilityofDLCwerefoundinthe literature,reportingdetrimentalstructuralchangescausedby thechangeinthecarbonbondings,from sp3 tosp2 [15,16]. Also,somepaperswerepublishedaboutimprovingthermal stabilityusingdopantelementsornanoparticlesincorporated totheﬁlm[17–19].

A preliminary work was published on the thermal behaviourof duplex coatings [20]. In this work adifferent thermaldegradationbetweenduplexandcoatedsampleswas detected.Thenitridingtreatmentimprovednotonlythe adhe-sionofthecoatingbutitalsoaffectedthesurfaceproperties.It wasdecidedtocontinuewiththeresearchinthistopic, ana-lyzingnewaspects.Themaingoalofthepresentworkisto studynotonlychangesinstructuralpropertiesbutalsothe tribologicalbehaviourofDLCcoatingsdepositedonnitrided andnon-nitridedprecipitationhardening(PH)stainlesssteels goingfurtherthaninthepreviouswork.Ramanspectrawere

performed,andacompletestudyofthefilmpropertiesand degradationafterthedifferentannealingcycleswascarried out.Topographymeasurementswereperformedtoimprove the understanding of the coating degradation and friction coefficientsmeasurementsarepresented.Theinfluenceofthe nitridedlayeronthethermalbehaviouroftheDLCcoatingis analyzed.

2. Experimental

PHmartensiticstainlesssteels(Corrax®fromUddeholm)were usedasbasematerial.Thechemicalcompositioninweight percentageofCorrax®is0.03%C,12%Cr,1.4%Mo,0.3%Mn, 0.3%Si,9.2%Ni,1.6%AlandFeasbalance.

Disc-shapedsampleswerecutfroma25mmindiameter barandtheywerehardenedbymeansofanageingprocessfor twohoursat530◦Caccordingtosupplierrecommendations. Onegroupofsampleswasnitridedinanindustrialreactor usingaDCpulsedischargeat390◦Cfor10hina25%N2+75%

H2gasmixture.

DLC coatings were deposited on all samples by PACVD usingmethaneasgasprecursor.Theprocesstotalgas pres-surewasof10Paat10ml/mingasﬂowduring2h.Structural properties of the nitrided layer and the DLC ﬁlm without annealing,soastheduplexsystemandcorrosionproperties have beenpublished before[10,11,21,22].Thegroup, which was notnitrided,but witha depositedcoating wasnamed “coatedsamples”andtheone,whichwaspreviouslynitrided, was named “duplex samples”. The coating thickness was about3.0±0.3␮mwithasiliconinterlayerof0.3␮mandthe nitridedlayerthicknesswasof14±1␮mapproximately.

The heat treatment post DLC deposition consisted of annealingstepsat200◦C,300◦C,400◦C,500◦Cand600◦Cfor 1h,countingfromthetimethetemperaturestepwasreached. Theheatingratewas20◦C/minandbothcoatedandduplex sampleswereheattreatedinavacuumchamber,following aprocedurereportedpreviouslyintheliterature[22].Ineach step,onegroupofsampleswastakenout.Soattheend,ﬁve groupsofsamplesannealedatdifferenttemperatureswere studied,eachgroupwithequalquantityofcoatedandduplex samples.

The DLC coatings were characterized by Raman spec-troscopy.TheRamanspectrawereﬁttedusingtwoGaussian lines. Theintensityforeachband wasdeterminedand the

ID/IG ratio was calculated.Thehydrogen content was

esti-matedusingtheslopeofthephotoluminescencebackground intheRamanspectratakingintoaccountthemodelproposed byCasiraghietal.[23].

Thesurface wasobserved byelectronicmicroscopy and whitelightinterferometry(WLI)using20×objective.Both sur-facemap andproﬁlewereobtainedbyMetropro© _software.

Moreover,focusedionbeam(FIB)cuttingswereperformedin differentregionsofthesurfacesamplesinordertoobserve thecrosssectionandtomeasuretheDLCthicknessafterthe annealing.Mechanicalpropertieswereassessedusing instru-mentednanoindentationwith15mNload.

Slidingweartests were carriedout using 2Nload,1800 cycles, 1mm ofamplitude and alumina balls (=6mm)as counterpart to evaluate the tribological behaviour of the

(3)

Duplex sample Coated sample 18 17 16 15 14 13 100 200 300 400 500 600 Temperature (°C) Nanohardness (GP a)

Fig.1–VariationofnanohardnessinDLCcoatingsafterthe annealingcyclesatdifferenttemperaturesforduplexand coatedsamples.

coatings.Three testswereperformedforeach sample.The weartrackswereobservedbySEMandWLI.Afterannealing, thedegradationofthesurfacecouldbeevaluatedfromthe valuesof the coefﬁcient offriction and the observation of thesurfacedamagealongandatthesidesoftheweartracks. The changes in ﬁlm properties, surface topography and tribological behaviour were compared after each thermal cycletodescribethecoatingdegradationwithtemperature.

3. Results

3.1. Changesinﬁlmpropertieswiththermalcycles

ThehardnessfortheDLCfilmsandforthenitridedlayerwere 16±1GPaand11±1GParespectivelyasitwasreportedina previouspublication[11].Themeanvaluesfornanohardness measuredaftereachannealingcyclearepresentedinFig.1, comparing coated and duplex samples. Twenty measure-mentsindifferentplaceswerecarriedoutandthemagnitude oftheerrorwasthestandarddeviationofthemeasured val-ues.Astheindenterpenetrationislessthan10%ofthefilm thickness, thishardness valuecanbe consideredas afilm propertywithouttheinfluenceofthesubstrate.

Inthistest,somedifferencesbecamevisible.Fortheduplex sample,thenanohardnessvaluesdidnothavean apprecia-blevariation,upto500◦C.Onlyat600◦C,somechangeswere observed,whichwereprobablycausedbystructural modiﬁca-tion.

Ontheotherhand,ahardnessdecreasewasnoticedinthe coatedsampleat300◦C.Probably,adegradationthatcannot beobservedoccurredattheinterfacebetweentheDLCcoating andthesubstrate.However,nochangeswereobservedneither insurfaceimagesnorintheRamanspectrum,asitwillbe mentionedbelow.

Athighertemperatures, the dispersion increasedinthe coatedsampleasitcanbeobservedinFig.1.Itispointless toindicateameannanohardnessvaluewhentheannealing wasperformedat600◦Cbecausethedispersionisveryhigh. Thenanohardnesschangesfromonepointtoanotheronthe surfacesamplewerebetween4GPaand14GPa.

Thisfactcanbeexplainedintwoways.Thecoatingwas detached partiallyor the film suffereda structureor com-position change. The first one could be easilyobserved in some regionsofthe surface and thesecond one was ana-lyzedwithRamanspectroscopyandtheresultsarepresented below.WhengraphitizationoccursinDLCfilms,thehardness decreases[18].

TheRamanspectraallowedthepossibilitytoevaluatethe ﬁlmstructure.ItcanbeseeninFig.2thatbothgroupsof sam-plespresentedsimilarfeaturesafterannealingat200◦Cand 300◦C.TheDLCwithoutthermalposttreatmentwasanalyzed aswellanditwasdepictedforcomparisoninblack.Inthis case,theRamanspectrumpresentedtheclassicaltwo over-lappingbandsDandG,which,asitwasalreadypublished, were well positioned, in 1379cm−1 and 1554cm−1 [15,16]. However,inthecoatedsampleat400◦C,itcouldbenoticed thattheﬁlmtransformationstartedbecausethe characteris-ticGandDbandsofDLCcoatingscouldnotbedistinguished (Fig.2).Ontheotherhand,intheduplexsample,thisdidnot occuruntilreachingatemperaturevalueof500◦C(Fig.2).

Afterannealingat200◦Cand300◦C,aslightshiftinthe DandGpeakpositionswasdetected,asitisshowninFig.2

andTable1.Forthesetemperatures,inbothsamples,itwas possibletomeasuretheDandGbandsposition,theintensity ratiobetweentheDandGbandsandtheDandGbandswidth, indicatingonlyaverysmallgraphitedomainforbothsamples (seeTable1).

Forhigherannealingtemperatures,itwasnotpossibleto perform an accurate analysis of the DLC coatings. A very strong ﬂuorescence appeared at 400◦C,revealing degrada-tion ofthe structural propertiesof the ﬁlm in the caseof the coated sample (Fig. 2). On the contrary, inthe duplex sample,theanalysisoftheRamanspectrumcorresponding to400◦C annealingcould stillbecarriedout.The degrada-tion starts at 500◦C, as it can be observed in the Raman SpectrumofFig.2,wheretheGandDbandscannotbe dif-ferentiated.Finally,at600◦C,therewasnocharacteristicDLC coatingstructureintheRamanspectrumforbothgroupsof samples.

3.2. Changesinsurfacetopographyandﬁlmintegrity

Surface topography did not change after the annealing at 200◦Cand300◦Cinbothcoatedandduplexsamples. More-over, this latter did not present changes in the surface topographyaftertheannealing at400◦C(Fig.3a).However, differentfeaturesindicatingafilmtransformationappeared whenthesurfaceofthecoatedsampleswasobservedafter the 400◦C cycle. Infact, part ofthe film wasdelaminated; bubbles and holes were formed, as it is shown in Fig. 3b. Moreover,theexistenceofholeswasconfirmedbythe differ-entheightsdetectedonWLIImages(Fig.4aandb).Itshould be mentionedthat the presenceofthe coating,even thin-ner,wasconfirmedbyEDSanalysisinallregionsofinterest (notshown).

Intheduplexsamples,somechangescould beobserved onthecoatingsurfaceattheendoftheannealingprocessat 500◦C,asitisshowninFig.5a.Thissamplepresentedsome cracksonthesurface(Fig.5a)buttherewasnotdetachment ofthe coatinginany region. Afterannealing at600◦C,the

(4)

Without annealing Without annealing 200 °C 300 °C 400 °C 500 °C 600 °C 200 °C 300 °C 400 °C 500 °C 600 °C 60000 50000 40000 30000 20000 10000 0 60000 50000 40000 30000 20000 10000 0 800 800 1000 1200 1400 1600 1800 1000 1200 1400 1600 1800 Duplex sample Coated sample Intensity (a.u.) Intensity (a.u.) Raman shift (cm-1)

Fig.2–Ramanspectrumoftheduplexandcoatedsamplesafterannealingatdifferenttemperatures.

Table1–AnalysisofRamanspectra.

Samples Dposition Gposition ID/IG Dwidth Gwidth

Duplexsamplewithoutannealing 1379 1554 0.66 283 132

Duplexsample200◦C 1390 1570 0.88 289 121

Duplexsample300◦C 1395 1573 0.83 285 104

Duplexsample400◦C 1388 1570 0.76 276 109

Coatedsamplewithoutannealing 1383 1557 0.71 283 127

Coatedsample200◦C 1388 1561 0.74 284 121

Coatedsample300◦_C ₁₃₉₄ ₁₅₇₂ _0.83 ₂₈₄ ₁₀₃

Fig.3– SEMimageofthesamplessurfacesafterannealingat400◦C:(a)duplexsampleand(b)coatedsample.

a

b

0.000 mm 0.282 -0.20000 +0.00000 +0.20000 +0.40000 +0.60000 0.000 0.050 0.100 0.150 0.200 0.250 0.000 mm 0.212 -1.15330 +2.02723 µm Distance (mm) Height (mm)

(5)

Fig.5–SEMimageoftheduplexsamplesurfacesafterannealingat(a)500◦Cand(b)600◦C.

Fig.6–SEMimageofthecoatedsamplesurfacesafterannealingat(a)500◦Cand(b)600◦C.

samplehadmorecracksasitcanbeseeninFig.5b,indicating amajordegradationofthecoating.However,thedegradation ofthe duplex coatings ismuch less than that of the only coatedsamples.

Indeed,inthecoatedsamplessomecracksareobservedin regionswherethecoatingwaspartiallydetachedattheend oftheannealingcycleat500◦C(Fig.6a).Thedegradationof thecoatingwasgreateraftertheannealingat600◦C(Fig.6b). However,thecoatingwasnotcompletelypeeledoffbutsome surfacelayersweregone.

In the coated sample surface, three FIB cuts were per-formedandthe locationisindicatedinFig.7.Thefirstone coveredaregionofaholedefectfromtheoutsidetotheinside (regionnamed1),thesecondone,ontheedgeofa protuber-ancewithadelaminatedpart(named2),andthethirdone, overaprotuberance(named3).Inregion1,Fig.7c,astepcan beobserved. Theright part ofthe coatinghas athickness similartotheoriginalcoating,showingthatthefilmhasnot beendamagedthere.Intheleftpartofthefilm,thicknessis only15%lessthantherightpart,showingthatthefilmwas nottotallyremoved.Inregions2and3,Fig.7bandd,some delaminatedlayersofthecoatingcanbeobservedinthecross section.Itcanbeseenthattheminimumfilmthicknesswas 1.5␮m,indicatingthatpartofthefilmwasstilltherebut thin-nerthanthenotannealedone(2.7␮m).Itcanbeconcluded thatseverallayersofthecoatingweredegradedanddetached, i.e.thecoatingdelaminationoccurredbysheets,revealinga cohesivefailure.

3.3. Tribologicalbehaviour

Theevolutionofthefrictioncoefﬁcient()isshowninFig.8

asafunctionoftimefortwoannealingtemperatures.Other tribological results (final values and wear damage) are presentedin Table2 forduplex and coated samples.After annealingat200◦Cand300◦C,thesteadyvalueofforboth sampleswaslowandsimilar,about0.2.Thisvalueisinthe rangeofthosethathavebeenreportedforthiskindof coat-ingsintheliterature[24].Thepresenceofthecoatingstrongly reducedthefrictioncoefficientincomparisonwithsteel,with orwithoutthenitridedlayer,whichisabout0.7.Thisisdue totheformationofagraphitetransferlayerwhichhasa self-lubricatingeffect[24,25].At200◦Cand300◦Cthetrackwas almostundetectable,indicatingthatthefilmwasintact.After highertemperaturecycles,theweartrackswere visiblebut veryirregularinwidthanddepth.Forthisreason,awear vol-umelosscouldnotbecalculated.Inaddition,thecoefficientof frictionbecomeshigherandirregularduetothesurface inho-mogeneitiesandareaswithdifferentcoatingthicknessand flaking.

The frictioncoefficient increasedfor the coatedsample aftertheannealingcycleat400◦C.Itreachedvaluesthat cor-respondtotheuntreatedsteelfrictioncoefficient(0.8–0.9in thesametestconditions).Ontheotherhand,fortheduplex samples,thefrictioncoefficientdidnotraisethatmuchafter annealing atthe same temperature; it beginsat0.2 and it reachedafinalvalueof0.47,asitcanbeobservedinFig.8.

(6)

a

c

d

1

3

2 b

Platinum layer Platinum layer Platinum layer 1.53 µ m (cs) 1.56 µ m (cs) 343.7 n m (cs) 2.72 µ m (cs) 2.42 µ m (cs) 2.37 µ m (cs) 50 µm 2µm 2µm 2 µm

Fig.7–(a)SEMimageofasurfaceregion,wheretheﬁlmwasdegradedinthecoatedsampleafterannealingat400◦C, showingthethreeregionswereFIBcutswereperformed.(b)Crosssectioninregion2,aprotuberancethatwasinpart peeledoff,canbeobserved.(c)Thestepinthecrosssectionofregion1.(d)Thesubsurfacestructureofaprotuberance, correspondingtoregion3.

Table2–Summaryofthetribologicalresultsforduplexandcoatedsamples.

Annealing(◦C) Duplexsample Coatedsample

(steadyorﬁnal) Weardamage (steadyorﬁnal) Weardamage

200 0.20±0.01 Undetectabletrack 0.20±0.01 Undetectabletrack

300 0.20±0.01 Undetectabletrack 0.20±0.01 Undetectabletrack

400 0.47±0.08 Detectabletrack 0.86±0.08 Coatingdetachment

500 0.77±0.07 Coatingdetachment 0.90±0.09 Coatingdetachment.

Greaterdamagethanin

theduplexsample.

600 0.90±0.09 Coatingdetachment 0.90±0.09 Coatingdetachment

Afterannealing at600◦C,thefriction coefﬁcientrapidly reached the value of the base material for both samples (Table2),indicatingthatthecoatinghasdetachedduringthe testandtheweardamageoccurredonthesubstrate.Infact, theweartrackdepthwasgreaterthanthecoatingthickness forbothsamples.

As a summary, the difference in tribological properties betweenbothgroupsofsamplesoccurredafterannealingat 400◦C.The valuesremainedratherlowafterthe anneal-ingprocessonlyintheduplexsamples.Itcouldbeconﬁrmed notonlythatthe coatingwas stillpresent,but alsothatit hasmaintaineditsproperties.Asitwasalreadymentionedin

Section 3.2,thecoating isthinneranddidnotdetach com-pletely,somepartsofitremainedonthesurface.

Fig.9presentsSEMpicturesofpartsoftheweartracksat theendofthetribologicaltestforannealedsamplesat400◦C. Ahighmagniﬁcationwasselected,withtheintentiontoshow topographyfeatureswithinandaroundthetrack.InFig.9a, whichcorrespondstotheweartrackonthecoatedsample, thecoatingdetachmentisclearlyvisible;infact,chemical ele-mentscorrespondingtothesteelsubstrateandaluminiumof thecounterpartcouldbedetectedbytheEDXanalysis. More-over,somedarkparticleswerealsoobservedinthetrack.Since the EDXanalysisshowsthattheyarecontainingoxygen,it

(7)

1,2 1,0 0,8 0,6 0,4 0,2 0,0 0 200 400 600 800 100012001400 160018002000 Time (s)

Coated sample after annealing at 500°C

Coated sample after annealing at 400°C

Duplex sample after annealing at 400°C

Duplex sample after annealing at 500°C

F

riction coefficient

µ

Fig.8–Frictioncoefﬁcientversustimeintheweartestfor duplexandcoatedsamplesannealedat400◦Cand500◦C.

canbededucedthattheyareoxidizeddebris.Themeantrack widthwas380␮minthiscase.

In Fig. 9b, which corresponds tothe wear track on the duplexsample,onlysomegroovesinthedirectionof move-mentandsomesmalldamagedregionscanbeobserved.The wearmechanism could be classified as mildabrasive. The coatingwasstilloralwayspresentinthewearscarandaround it,eveninthedamagedareas,asitcouldbeconfirmedbythe EDXspectrum,inwhichSiandCweredetected.Withless mag-nification,thewholescarcouldbeobserved,andthemean widthwasabout150␮m.

Attheendoftheweartestsonsamplesafterannealingat 500◦C,someﬂakingcouldbedetectedintheweartracksas itcanbeseeninFig.10aandb.Intheseﬁgures,onlypartsof thetracksarepresented,becausefocuswasmadeontheedge

ofthewearscartoobservethedamagefeatures.Outsidethe track,theﬁlmsurfacecanalwaysbeobservedanditisclear,by comparingbothﬁgures,thatthedamagewasalreadypresent beforethetest.Itcanalsobeobservedthatthedegradation wasmoreextendedinthecoatedsamplethanintheduplex one.

Looking inside the wear track in the coated sample (Fig.10a),adhesivewearclearlyoccurred,andlargeplate-like wear debriscould befoundas other authors alsoreported

[26]. Itiswell knownthat bondingoccurs betweensurface asperities, andwhenthesejunctions arebroken,newones areformedaslargelumpsandtransferofmaterialscomesasa result.Alargeamountofplasticdeformationisusually associ-atedwiththismechanism[24,26].Thewearwassevereinthis case;theelementsfromthesubstrate andalsoAlfrom the counterpartweredetectedbyEDXanalysis.Onthecontrary, intheduplexsample,thedamagecouldbedescribedasmild abrasivewear. Soft grooveswithsomedarkregions, which correspond tooxidizedparticles,can beobserved(Fig. 10b)

[27].

4. Discussion

AnexplanationaboutthedifferentbehaviourofDLCcoated PH sampleswith and without nitriding pre-treatment can beproposedand,relatedtothispoint,somehypothesesare describedbelow.

ThefirstoneisthechemicalmodificationoftheDLCfilm withanincreaseintemperature.ItisknownthatDLCfilms transform toanaromatic ring structureandthen to disor-deredgraphiticringstructureduringannealingattemperature

Fela Feka O ka EDS zone 4 Fela Auma Sika EDS zone 2 Alka C ka Sika crKa Feka Nika 2 4 6 8 keV 2 4 6 8 keV EDS zone 3 crKa Nika AlKa AuMa Sika Cka Sika O ka C ka CrkaFeka EDS zone 1 2 4 6 8 keV 2 4 6 8 keV

a

b

O ka C ka 3 Particles Damaged regions Detachment coating Grooves Particles

Small damaged regions

Mild abrasive wear 1

2

30 µm

Fig.9–SEMimagesoftheweartrackandEDXanalysisattheendofthetribologicaltestforannealingat400◦C:(a)duplex sampleand(b)coatedsample.

(8)

Fela FeKa EDS zone 2 AuMa O ka O ka CrKa C ka C ka Sika Sika EDS zone 1 Alka Nika Feka Crka 2 1 2 3 4 5 6 keV 4 6 8 keV

a

b

Oxidation

Flaking

Plate - like wear debris

Adhesive wear

20 µm

Fig.10–SEMimagesoftheweartrackandEDXanalysisattheendofthetribologicaltestafterannealingat500◦C:(a) duplexsampleand(b)coatedsample.

higherthan400◦C[28].Withincreasingannealing tempera-ture,thestructuraldisorderdecreasesandthesizeofgraphitic domainsincrease[28,29].

Suchphenomenonisassociatedwithplasticdeformation, whichreducesthe internalstressand inturn thehardness oftheDLCcoating.Accordingtotheresultspresentedabove, frompropertiessuchashardnessandtheobservationof sur-facemorphology,itcanbestatedthatthetransformationof theDLCcoatingingraphitewasproducedatahigher tem-peraturein the duplex samplethan in the coated sample. Thiscouldbeconfirmedbythe Ramananalysis, wherethe modification and the shift of the classical DLC peaks is a signofcoatingdegradation.Suchdegradationwasdetectedin bothgroupsofsamplesat600◦C.However,thisphenomenon occurredatahighertemperatureinthecaseoftheduplex samplesascomparedtotheonlycoatedsamples.Itishighly probablethatintheduplexsample,therateofgraphitization inthecoatingwaslowerthaninthecoatedsamples. There-fore,thefilmtransformationdidnothavesomuchinfluence onthecoatingpropertiesanditwasalsolessnoticeable.This changeingraphitizationisprobablyrelatedtoachangeinthe thermalconductivityaswillbediscussedlater.

The difference between both groups of samples is the nitridingpre-treatment.Theinfluencethatithasonthefilm adhesion,whichwasbetterintheduplexsamplethaninthe coated sample, wasreported in aprevious paper [11]. The nitridedlayerraisestheloadbearingcapacityofthesystem. Thislayergeneratesagradedinterfacebetweenthecoating andsubstrate,increasesthehardnessofthesubstrate,reduces thestressesandimprovestheadhesionasitwaspublished alsobyotherauthors[11,30,31].Moreover,inbothsamples,a siliconinterlayerwasdepositedprevioustothecoating.Asa result,thereisachemicalaffinitybetweenthisSilayerand thenitrogenofthenitridedlayer.So,intheduplexsample, SireactswithNformingsiliconnitridebonds,aspointedout bysomeoftheauthors,whenXPSanalysiswereperformed inathinSilayerdepositedontonitridedsteel[32,33].Since hardnessandadhesionarehigherfortheduplexcoating, con-sequentlyitismorestableathighertemperatures.

Regardingthermalconductivity,thesiliconadhesionlayer has a thermal coefﬁcient (3.2×10−6K−1) comparable to those from DLC(2.3×10−6K−1)andfrom thenitridedlayer (7.6×10−6K−1)butlowerwhencomparedtotheoneofbare stainless steel(14.5×10−6K−1).As a consequence,thermal

(9)

stressislowerintheduplexsamplethaninthecoated sam-ple[34,35].Ontheotherhand,thethermalconductivityofDLC isverylow;itcanbehundredtimeslowerthanthethermal conductivityofsteel.Therefore,iftheadhesionishigherfor theduplexcoating,theheatdissipationishigheraswelland thefilmresistsmoretothethermaleffect.Athinfilm with-outgoodadhesionduetothelowthermalconductivityheats rapidlymodifyingitsstructure,i.e.goingfastertoastructure closetothegraphiteortoagraphite-likefilm.Asthe dura-tionofthethermalcycleisonlyonehour,adelaycanexist beforetheestablishmentofastabilizedtemperaturebetween nitridedandnon-nitridedsamples.Thus,somebubblescould beformed inthe film withworse adhesion caused bythe releaseofbondedandunbondedhydrogenduringthe anneal-ing,asitwasreportedbyotherauthors[34,36].However,the bubbleformationwasnotobservedinwelladherentfilms,e.g. theduplexsample.Thiscouldbeduetoamoreefficientheat dissipation.

Arelationbetweennanohardnessmeasuresandadhesion canthusbeestablished.Thatis,wheneverathinfilmthatis welladhered, thehardnessmeasurementismoretrustable becausethewell-adheringfilmhasthesubstrate asamore efficientsupport.Ontheotherhand,thenon-adherentfilm is“loose”onthesubstrate,notallowingthehardness mea-surementtobereal,aswellasbeingclosertothegraphite structure. In this work, it was quite clearthat the disper-sionofthenanohardnessmeasurementsinthecoatedsample observedinFig.1wasdirectlyrelatedtothedispersionofthe adhesionthroughtheinterface,causedbythefilm degrada-tion.

ThefrictioncoefficientremainedsimilartotheDLCcoating withoutthermaltreatmentaftertheannealingat200◦Cand 300◦Cforbothsamples.Aftertheannealingat400◦C,the fric-tioncoefficientwaslow(0.2)atthebeginning.After800sthe valueincreasedforbothsamples,butattheenditwashigher forthe coatedsamplethan theduplexsample.Probably as thehardnesswaslowerforthecoatedsamplecomparedtothe duplexsampleandtheoriginalcoating,thepenetrationofthe counterpartwasdeeper,sotherealcontactareaincreasedand consequentlythefrictioncoefficienttoo[26,37].Thefriction coefficient increasedwhenthe annealing temperature was higher.Thiscouldbeduetothefactthattheannealingathigh temperaturesproducestheeffusionofhydrogenand deteri-oratethecoatingstribologicalproperties,asitwasreported intheliterature[38].Moreover,thecoatingproperties (hard-ness,topographyandthickness)effectivelychangedafterthe annealingatthistemperature,asitwaspresentedinthe previ-oussections.Consequently,themechanicalresistanceofthe systemdecreasedandthe wearmechanismschanged.Itis clearthattheweardamagebecamegreaterwiththe increas-ingannealingtemperaturebut,inallcases,thedamagewas worseinthecoatedsamplethanintheduplexone.

5. Summary

and

conclusions

Asasummaryofresults,itcanbestatedthatatthesame tem-perature,thecoatingdegradationisalwayslowerintheduplex samplethanintheonlycoatedone.Moreover,thedegradation orthemorphologychanges aredifferentinthetwogroups

ofsamples.Intheduplex,cracksappearontheﬁlmsurface andnodelaminationwasobserved,meanwhileinthecoated sample,theﬁlmhascohesivefailuresanditdetachedlayerby layer.Changesinstructurepropertiesappearat400◦Cinthe coatedsampleandat500◦Cintheduplexone.

Itwasshownthatthelowfrictioncoefﬁcientandgood tri-bologicalpropertiesoftheDLCcoatingremainedunchanged intheduplexsamplesafteranannealingprocessat400◦C. Itwasalsoobservedthat thenanohardnessvaluesdidnot decreaseafterthermaltreatments,evenathightemperatures. Itwasproved thatanitridingtreatmentprevious tothe coatingdepositionimprovestheDLCthermalstabilitythrough thethermalannealingcycles,ifcomparedtothecoatingover thesamestainlesssteelwithoutanynitridingpre-treatment. Itisafactthatabetteradhesionisobtainedinthe duplex samples,andthereisalsoabettermatchingofthethermal coefﬁcients and the mechanical properties ofthe nitrided layerwiththe DLCcoating,allowinginsomewaythat the nitridedlayerdelayedthethermaldegradationofthecoating depositedoverit.

Inthispaper,itisshownthattheadhesionandthe hard-ness are related and that can be evaluated from a heat treatment. This is a new result, which the authors can speculate asbeing ofmuchimportancetoany measureof nanohardnessinthinﬁlms.

Furtherstudiesarenecessarytorelatethecoatings struc-ture to the properties of the interface with the nitrided substrateandtheearlystagesofgrowing,andtheirbehaviour atdifferenttemperatures.

Conﬂicts

of

interest

Theauthorsdeclarenoconﬂictsofinterest.

Acknowledgements

The authors would like to thank students of the Surface EngineeringGroup(UTN)fortheircollaborationinthe experi-mentaltasks.AlsotoDipl.Eng.FedericoLasserreatthattime atSaarland University (Germany), forthehelp inSEM and WLI experiments.PIDUTI 4716 TC UTN,2018-2021. SUMA2 Network,EUFP7-PEOPLE-2012-IRSES,ProjectNr.318903

r

e

f

e

r

e

n

c

e

s

[1]ChoyK.Chemicalvapourdepositionofcoatings.ProgMater Sci2003;48:57–170.

[2]RobertsonJ.Diamond-likeamorphouscarbon.MaterSciEng RRep2002;37:129–281.

[3]Trava-AiroldiVJ,BonettiLF,CapoteG,SantosLV,CoratEJ.A comparisonofDLCﬁlmpropertiesobtainedbyr.f.PACVD, IBAD,andenhancedpulsed-DCPACVD.SurfCoatTechnol 2007;202:549–54.

[4]ChicotD,Puchi-CabreraES,DecoopmanX,RoudetF,LesageJ, StaiaMH.Diamond-likecarbonﬁlmdepositedonnitrided 316Lstainlesssteelsubstrate:ahardnessdepth-proﬁle modeling.DiamRelatMater2011;20:1344–52.

[5]SnydersR,BousserE,AmireaultP,Klemberg-SapiehaJE,Park E,TaylorK,etal.Tribo-mechanicalpropertiesofDLC

(10)

coatingsdepositedonnitridedbiomedicalstainlesssteel. PlasmaProcessPolym2007;4:S640–6.

[6] AzziM,PaquetteM,SzpunarJA,Klemberg-sapiehaJE, MartinuL.TribocorrosionbehaviourofDLC-coated316L stainlesssteel.Wear2009;267:860–6.

[7] ShiogaJPHT,BinderC,HammesG,KleinAN,deMelloDB. Effectsofdifferentplasmanitridedlayersonthetribological performanceofDLCcoatings.MaterRes2016;19:1180–8.

[8] KovaciH,BaranÖ,YetimAF,BozkurtYB,KaraL,C¸elikA.The frictionandwearperformanceofDLCcoatingsdepositedon plasmanitridedAISI4140steelbymagnetronsputtering underairandvacuumconditions.SurfCoatTechnol 2018;349:969–79.

[9] PodgornikB,Vi ˇzintinJ,WänstrandO,LarssonM,HogmarkS, RonkainenH,etal.Tribologicalpropertiesofplasmanitrided andhardcoatedAISI4140steel.Wear2001;249:254–9.

[10]DalibonEL,CharadiaR,CaboA,Trava-AiroldiVJ,BrühlSP. EvaluationofthemechanicalbehaviourofaDLCﬁlmon plasmanitridedAISI420withdifferentsurfaceﬁnishing. SurfCoatTechnol2013;235:735–40.

[11]DalibonEL,Trava-AiroldiVJ,PereiraLA,CaboA,BrühlSP. WearresistanceofnitridedandDLCcoatedPHstainless steel.SurfCoatTechnol2014;255:22–7.

[12]ForsichC,HeimD,MuellerT.Inﬂuenceofthedeposition temperatureonmechanicalandtribologicalpropertiesof a-C:H:Sicoatingsonnitridedandpostoxidizedsteel depositedbyDC-PACVD.SurfCoatTechnol2008;203: 521–5.

[13]KovalevaMG,KolpakovAJ,PoplavskyAI,GalkinaME.Effect ofvacuumannealingontribologicalbehaviorofnanosized diamondlikecarboncoatingsproducedbypulsevacuumarc method.JFrictWear2013;34:481–6.

[14]ItoH,YamamotoK,MasukoM.ThermalstabilityofUBM sputteredDLCcoatingswithvarioushydrogencontents. ThinSolidFilms2008;517:1115–9.

[15]LengYX,ChenJY,YangP,SunH,WanGJ,HuangN. Mechanicalpropertiesandthermomechanicalstabilityof diamond-likecarbonﬁlmssynthesizedbypulsedvacuum arcplasmadeposition.SurfCoatTechnol2003;173:67–73.

[16]ChoiJ,MiyagawaS,NakaoS,IkeyamaM,MiyagawaY. Thermalstabilityofdiamond-likecarbonﬁlmsdepositedby plasmabasedionimplantationtechniquewithbipolar pulses.DiamRelatMater2006;15:948–51.

[17]MoolsradooN,AbeS,WatanabeS.Thermalstabilityand tribologicalperformanceofDLC-Si-Oﬁlms.AdvMaterSci Eng2011.

[18]HatadaR,BabaK,FlegeS,EnsingerW.Long-termthermal stabilityofSi-containingdiamond-likecarbonﬁlms preparedbyplasmasourceionimplantation.SurfCoat Technol2016;305:93–8.

[19]DaiW,GaoX,LiuJ,WangQ.Microstructure,mechanical propertyandthermalstabilityofdiamond-likecarbon coatingswithAl,CrandSimulti-doping.DiamRelatMater 2016;70:98–104.

[20]BrühlSP,DalibonEL,GhafoorN,RogströmL,Trava-AiroldiVJ, OdénM.Carbonbasedcoatingsdepositedonnitrided stainlesssteel:studyofthermaldegradation.In:MeyersM,

etal.,editors.Proc.3rdPanAm.Mater.Congress.The minerals,metals&materialsseries.Springer;2017.

[21]GuitarA,Trava-airoldiVJ,MuF,BrühlSP.Plasmanitriding andDLCcoatingsforcorrosionprotectionofprecipitation hardeningstainlesssteel.AdvEngMater2015;18:826–32.

[22]ZhangS,LamX,LiX.Thermalstabilityandoxidation propertiesofmagnetronsputtereddiamond-likecarbonand itsnanocompositecoatings.DiamRelatMater2006;15:972–6.

[23]CasiraghiC,FerrariAC,RobertsonJ.Ramanspectroscopyof hydrogenatedamorphouscarbons.PhysRevB2005;72:1–13.

[24]DonnetC,ErdemirA.Tribologyofdiamond-likecarbonﬁlms. fundamentalsandapplications.USA:Springer;2008.

[25]HolmbergK,MatthewsA.Coatingstribologyproperties, mechanisms,techniquesandapplicationsinsurface engineering.UK:Elsevier;2009.

[26]WangL,NieX,HuX.Effectofthermalannealingon

tribologicalandcorrosionpropertiesofDLCcoatings.JMater EngPerform2013;22:3093–100.

[27]BayerRG.Wearanalysisforengineers.NewYork:HNB Publishing;2002.

[28]PeterS,GüntherM,GordanO,BergS,ZahnDRT,SeyllerT. Experimentalanalysisofthethermalannealingofhard a-C:Hﬁlms.DiamRelatMater2014;45:43–57.

[29]ShengR,LiL,SuD,PengJ,GaoJ,ZhaoK,etal.Effectof unbondedhydrogenonamorphouscarbonﬁlmdepositedby PECVDwithannealingtreatment.DiamRelatMater 2018;81:146–53.

[30]PodgornikB,Vi ˇzintinJ,LeskovˇsekV.Tribologicalproperties ofplasmaandpulseplasmanitridedAISI4140steel.Surf CoatTechnol1998;108–109:454–60.

[31]BorgesCFM,PfenderE,HeberleinJ.Inﬂuenceofnitridedand carbonitridedinterlayersonenhancednucleationof diamondonstainlesssteel304.DiamRelatMater 2001;10:1983–90.

[32]LiD,GuruvenketS,AzziM,SzpunarJA,Klemberg-SapiehaJE, MartinuL.Corrosionandtribo-corrosionbehaviorof a-SiCx:H,a-SiNx:Handa-SiCxNy:HcoatingsonSS301 substrate.SurfCoatTechnol2010;204:1616–22.

[33]ChenCW,HuangCC,LinYY,ChenLC,ChenKH.Theafﬁnity ofSi NandSi Cbondinginamorphoussiliconcarbon nitride(a-SiCN)thinﬁlm.DiamRelatMater2005;14:1126–30.

[34]WeiCh,YenJY.Effectofﬁlmthicknessandinterlayeronthe adhesionstrengthofdiamondlikecarbonﬁlmsondifferent substrate.DiamRelatMater2007;16:1325–30.

[35]LiXY,SunY,BellT.StabilityofthenitrogenS-phasein austeniticstainlesssteel.IntJMaterRes1999;90:901–7.

[36]MaheswaranR,ThiruvadigalDJ,GopalakrishnanC.Thermal instabilityofDLCﬁlmsurfacemorphology–anAFMstudy. AIPConfProc2012;1447:789–90.

[37]CloutierM,HarnageaC,HaleP,SeddikiO,RoseiF,Mantovani D.Long-termstabilityofhydrogenatedDLCcoatings:effects ofagingonthestructural,chemicalandmechanical properties.DiamRelatMater2014;48:65–72.

[38]LiH,XuT,WangC,ChenJ,ZhouH,LiuH.Annealingeffect onthestructure,mechanicalandtribologicalpropertiesof hydrogenateddiamond-likecarbonﬁlms.ThinSolidFilms 2006;515:2153–60.