O
pen
A
rchive
T
OULOUSE
A
rchive
O
uverte (
OATAO
)
OATAO is an open access repository that collects the work of Toulouse researchers and
makes it freely available over the web where possible.
This is an author-deposited version published in :
http://oatao.univ-toulouse.fr/
Eprints ID : 16746
To link to this article : DOI:10.1016/j.mtcomm.2016.07.007
URL :
http://dx.doi.org/10.1016/j.mtcomm.2016.07.007
To cite this version :
Hentour, Karim and Marsal, Alexis and Turq,
Viviane and Weibel, Alicia and Ansart, Florence and Sobrino,
Jean-Michel and Chen, Yan Ming and Garcia, Julien and Cardey,
Pierre-François and Laurent, Christophe Carbon nanotube/alumina and
graphite/alumina composite coatings on stainless steel for
tribological applications. (2016) Materials Today, vol. 8. pp.
118-126. ISSN 1369-7021
Any correspondence concerning this service should be sent to the repository
administrator:
staff-oatao@listes-diff.inp-toulouse.fr
Carbon
nanotube/alumina
and
graphite/alumina
composite
coatings
on
stainless
steel
for
tribological
applications
Karim
Hentour
a,b,
Alexis
Marsal
a,b,
Viviane
Turq
a,
Alicia
Weibel
a,
Florence
Ansart
a,
Jean-Michel
Sobrino
b,
Yan
Ming
Chen
b,
Julien
Garcia
b,
Pierre-Franc¸
ois
Cardey
c,
Christophe
Laurent
a,∗aCIRIMAT,UniversitédeToulouse,CNRS,INPT,UPS,118,RoutedeNarbonne,31062,ToulouseCedex9,France
bCETIM(CentreTechniquedesIndustriesMécaniques),52,AvenueFélixLouat,CS80067,60304,SenlisCedex,France
cCETIM(CentreTechniquedesIndustriesMécaniques),7,RuedelaPresse,CS50802,42952,Saint-ÉtienneCedex1,France
Keywords:
Carbonnanotubes
Graphite Alumina
Solgelcoatings
Friction Wear
a
b
s
t
r
a
c
t
Carbon/aluminacoatingsonstainlesssteelarepreparedbyasol-gelroute,usingeithercarbonnanotubes (8wallsonaverage)orgraphiteflakes.Thefrictioncoefficientagainstasteelballisdecreasedbyafactor of4–5comparedtopurealuminaandwearisreducedbyafactorof2withgraphiteflakes.ARaman spectroscopystudyofselectedspecimensoutsideandinsidethewornsurfaceshowsthatthecarbon nanotubesarenotdramaticallydamagedwhereasthegraphiteflakesarebrokenintographenelayers. Thereasonswhygraphiteismoreeffectivethanthecarbonnanotubes,forthesamecarboncontent,to improvethetribologicalbehaviorarediscussed.
1. Introduction
Austeniticstainless steels arewidely used in theaerospace, energy,medical and foodindustries becauseoftheirresistance to corrosion but they are highly susceptible to adhesive wear (seizure), therefore leading to a significant loss of profitability and possiblytosomeenvironmentalimpact.Carbon-containing compositesareofparticularinterestasself-lubricatingmaterials showingahighresistancetofrictionandwear,preventingtheneed forliquidlubricants.Reportsonthetribologicalbehaviorof metal-[1–11]and ceramic-matrix[12–17] bulk compositescontaining carbonnanotubes(CNTs)areincreasinglyabundant.However,the comparisonoftheresultsreportedbydifferentgroupsishampered notablybecausedifferentCNTsareused,thepreparation routes markedlydifferandthetribologicaltestingconditions (counter-face,load,slidingdistance,relativehumidity,temperature)vary widely.Moreover,itiseitherundesirableorimpossibletousebulk compositesfor applicationssuchastheprotectionof austenitic steels.Therefore,compositecoatingsaretobepreferred,suchas CNT-metalelectrolesscoatings[3]andplasma-sprayedCNT/Al2O3
coatings[18,19].Balanietal.[18]reportedthattheslidingwear
∗Correspondingauthor.
E-mailaddress:laurent@chimie.ups-tlse.fr(C.Laurent).
volumelossofa8wt.%CNT/Al2O3coatingagainstaZrO2pin(dry
conditions, normal load 48N) was49 times lower than for an Al2O3coating.Keshrietal.[19]reporteda72%increaseinwear
resistanceagainsta WCball(298K,ball-on-disktribometer)for a8wt.%CNT/Al2O3 coatingcomparedtoAl2O3.Note thatthese
resultsarepartlyattributedtoindirecteffectsofthepresenceof CNTs,such asalocally enhanced densificationofthewear sur-faceandahighertoughnessofthecoatingthroughCNTsbridging betweenthesplatsand/orAl2O3grains.Graphite/Al2O3bulk
com-positeswereshown[20]tohaveafrictioncoefficienthalfofthat ofpureAl2O3.Alowshearstrengthintheslidingdirectionand
highcompressionstrengthinthedirectionoftheload(i.e. perpen-diculartotheslidingdirection)arebeneficialtolowerthefriction coefficient.Laminatedgraphite/Al2O3 compositeswerereported
[21]toshowsignificantlybetterfrictionandwearbehaviorsthan monolithicgraphite/Al2O3composites(againstanAl2O3ball,dry
conditionsroomtemperature)becausegraphiteparticlescanbe easilydraggedontothefrictionsurfacefromthegraphitelayers. Changesinthelayerspacingandinthegraphitevolumefraction affectthe formation oflubricating and transferring films, load-bearing capacities and wear mechanisms of thematerials. The aimsofthepresentworkaretoshapeCNT/Al2O3coatingsontoan
austenitic304-Lstainlesssteelsubstrate,tocomparetheir tribo-logicalbehaviortothatofpureAl2O3andgraphite/Al2O3coatings
Fig.1.TEMimagesoftheCNTs(aandb)andFESEMimage(c)andTEMimage(d)ofgraphiteflakes.
andtoexplainwhygraphiteismoreeffectivethanCNTstoimprove thetribologicalbehavior.
2. Experimentalmethods 2.1. Rawmaterials
ACNTsampledescribedindetailelsewhere[10]waspurchased fromNanocyl(Belgium).TEMimages(Fig.1aandb)showthatthe CNTsaremostlynotbundledandcontainafairnumberofdefects alongtheirlengths.CNTswith3–22wallsareobserved.CNTswith 8wallsaredominant(30%)withCNTswith7and9walls(both16%) thesecondmostabundantpopulations.
Theaveragenumber ofwallsis equal to8.5(rounded to8). TheaverageCNTouterdiameteris10.2nmandlengthisbelow 1.5mm.Thespecificsurfaceareaofthesampleis242m2g−1,in
agreementwithcalculationsfromgeometricaldata[22].TheCNTs werecarboxyl-functionalizedwithanitricacidsolution(3molL−1)
as described elsewhere[23]. Graphite platelets (Fig. 1c and d) werepurchasedfromAbcr(Germany).Flakesabout15nmthick areagglomeratedintoplateletsabout1.5mminsize.Thespecific surfaceareaofthesampleis20m2g−1.
Anaqueousaluminumchloridehexahydrate(AlCl3·6H2O)
solu-tion(0.13molL−1)waspouredintoanexcessofammoniasolution
Fig.2. Measured carboncontent versustheexpectedcarbon content forthe
CNT/Al2O3(N)andgraphite/Al2O3coatings( ).Thesolidlineshowstheexpected
carboncontent(i.e.1for1).
(5molL−1)undermagneticstirringatroomtemperature.The
so-obtainedboehmite(AlOOH)precipitatewasfiltered,washedwith deionizedwaterandoven-driedovernight,producingaboehmite
powder.Aknownamountofthispowderwasdispersedinto deion-izedwaterandthesolutionwaspeptizedwitha3.1vol.%solution ofaceticacid(100%)andstirredfor24h,producingaboehmite col-loidalsol.Carbonwasaddedindifferentproportions(6,8,10,12 and14gL−1 forCNTsand5,10,25and50gL−1forgraphite),as
requiredforthestudy.Thecarbon-boehmitesolsweresonicated (VibraCell75042sonotrode),1hforCNTsand0.5hforgraphite.
Stainless steel disks (AISI 304-L, diameter 30mm, thickness 5mm,arithmeticaverageroughnessRa0.6mm)wereusedas sub-strates.Thediskswerefirstpre-treatedusingaroutineinvolving alkalinedegreasing,acidpicklingandnitricacidpassivation[24]. 2.2. Preparationofthecoatings
Al2O3andcarbon-Al2O3coatingswerepreparedbydip-coating
followedbythermaltreatments[24].Thestainlesssteelsubstrates weredippedintotheboehmiteorcarbon-boehmitecolloidalsols andwithdrawnatacontrolledspeed(300mmmin−1),resultingin
thedepositionofasollayerontothesubstrate.Afterremovingof theexcessliquid,dryinginair(80◦C,2h)provokedsolvent
evapo-rationandtransformationofthesolinaxerogel.Aheat-treatment inN2(500◦C,25min,heatingrate100◦Ch−1,naturalcooling)then
resultedintheformationoftheAl2O3andcarbon-Al2O3coatings.
Thecoatingswillbereferredtohereafterbyaletter(TforCNTsorG
forgraphite)andanumberdenotingitsmeasuredcarboncontent (weight%),i.e.T18orG21.
2.3. Characterization
The viscosity of the sols was measured using a viscometer (LAMYRM100)inaTaylor-Couetteconfigurationatashearrange of966–2900s−1.Grazing incidenceX-raydiffraction(XRD)was
carriedoutonthesubstrateandonthecoatingsusingaSiemens D5000diffractometerwithCuKaX-raysource.XRDpatternswere collectedatroomtemperatureby0.02◦(2u)scanningstepsover
the10–100◦range.Thecarboncontentinthecarbon-Al
2O3
coat-ingswasobtainedusinganelectronmicroprobe(EPMA,Cameca SXFive).Thesurfaceofthecoatingswasobservedby field-emission-gunscanningelectronmicroscopy(FESEM,JEOLJSM6700F)and interferentialrugosimetry(ZygoNewView100).Thethicknessof thecoatingswasdeterminedoncross-sectionalSEMimages(SEM JEOL35CF).Crosssectionswerepreparedbycryo-fracturationon notchedsamples.The weartrackswereobserved by3Doptical profilometry(SENSOFARSneox)andopticalmicroscopy(Keyence VHX-1000E). Selected samples were studied by Raman spec-troscopy(Horiba800spectrometerusing633nmlaserexcitation). Thespectrarepresentresultsobtainedfromthreedifferentareas inagivensample.
Fig.3.FESEMimagesshowingthesurfaceoftheCNT/Al2O3(18wt.%)(a)andgraphite/Al2O3(21wt.%)(b)coatingsandhighermagnificationimagesshowingcracksmade
Fig.4. FESEMimagesshowingcross-sectionsoftheCNT/Al2O3(18wt.%)coating(a)(whitearrowsindicateCNTs)andgraphite/Al2O3(21wt.%)(b)coating;higher
magnifi-cationimagesshowCNTs(c)andgraphiteplatelets(d),respectively.
2.4. Nano-indentationandtribologicaltesting
Instrumented indentation tests were conducted (in air at roomtemperature,maximumnormalload600mN,load/unloading rate 800mNmin−1, dwell 60s, maximum penetration depth
100–300nm) witha Berkovich nano-indenter(CSM Instrument UltraNanoindenter)inordertorecordtheloadversus penetra-tiondepthandtoevaluatetheYoung’smodulusandhardnessof theraw(i.e.un-polished)coatings.Frictiontestswereperformed usingaball-on-diskgeometryinrotarymode(CSMTribometer)in compliancetotheASTMG99internationalstandard.Tests(normal load2N,rotatingspeed10cms−1andtotalslidingdistance250m)
wereperformedatroomtemperatureinambientairwitha40–60% relativehumidity.Theun-polishedsurfaceswererubbedagainsta 316-Lsteelball10mmindiameter.Thefrictionalforcetransferred toaloadcellwasrecordedthroughoutthetest.Allfrictiontests wererepeatedthreetimes,showingidenticalresults.
3. Resultsanddiscussion
3.1. Compositionandmicrostructureofthecoatings
Theviscosityofthesolincreasesmarkedlyupontheincreasein CNTcontent,from11mPas(pureboehmitesol)to28mPasforthe solwithacarbonproportionof14gL−1.Bycontrast,theviscosity
ofthegraphite-boehmitesolsremainsconstantforallthe stud-iedcarbonconcentrationrange(5–50gL−1)andisonlymarginally
higherthanforpureboehmite(13vs11mPas).Thiscouldreflect thehigheraspectratio(length/diameter)ofthe8CNTs(150) com-paredtothatofgraphite(1).Thedifferentcarbonproportions(6,8, 10,12and14gL−1for8CNTsand5,10,25and50gL−1forgraphite)
inthestartingsolsshouldyieldcarboncontentsinthecoatings equalto12,14,16,18,20wt.%for8CNTsand7.9,14.6,30,60wt.% forgraphite.Theactualmeasuredcarboncontentis7–27%lower thanexpectedfortheCNT/Al2O3coatings(Fig.2),butthe
discrep-ancytendstodiminishforhighercontents,andis62–68%lower thanexpectedforthegraphite/Al2O3coatings(Fig.2),indicating
thatcarbonspecies,inparticularthegraphiteparticles,tendtobe drainedbygravityduringthewithdrawingstepofdip-coatingif theviscosityofthesolistoolow.
AnalysisoftheXRDpattern(notshown)fortheun-coated304-L steelsubstraterevealstheaustenite(111)and(200)peaks(major phase)andaweakferriticormartensiticpeak(minorphase).For theAl2O3 coating,thesteelpeaksarelessintenseandnopeaks
correspondingtocrystallizedAl2O3aredetected.Forthe
carbon-Al2O3coatings,veryweaksteelpeaksaredetectedinadditiontoa
strongCNT-orgraphite(002)peak.Again,nopeakscorresponding toAl2O3aredetected,showingthatAl2O3hereiseitheramorphous
Fig.5.LoadversusdisplacementplotsmeasuredwithaBerkovichnano-indenter
(a),Young’smodulus(b)andhardness(c)versuscarboncontent.
Fig.6.Frictioncoefficientagainstasteelballversustheslidingdistanceforthe
un-coatedsubstrateandtheAl2O3,CNT/Al2O3(18wt.%)andgraphite/Al2O3(21wt.%)
coatings.Thetestloadisequalto2N.
Fig.7.Steady-statefrictioncoefficientagainstasteelballversuscarboncontentfor
theAl2O3( ),CNT/Al2O3(N)andgraphite/Al2O3( )coatings.
FESEMobservationsofthesurfaceofthecoatingsreveala homo-geneousdistributionoftheCNTs(Fig.3a)andgraphiteplatelets (Fig.3b, back-scatteredelectronimageinchemical composition mode).Cracks(Fig.3candd),weregeneratedonpurposeby per-forminghigh-loadindentations.CNTsareobservedbridgingthe crack(Fig.3c),ascouldbeexpectedfrompreviousresults[23,25], whereastheshorterplateletsdonot(Fig.3d).
Cross-sectionsimagessuchasthoseshownforT18andG21 sug-gestthattheTcoatings(Fig.4a)arelessroughthantheGcoatings (Fig.4b),asconfirmedbywhite-lightinterferometry.Notethatthe crackbetweensubstrateandcoating(Fig.4a)wasformedbecause thecrosssectionwaspreparedbycryo-fracturation.TheCNTsare homogenouslydispersedwithinthecoating(Fig.4c).Thegraphite plateletsmostlyformmicrometer-sizedagglomerates(Fig.4b)but individualplateletsnotthickerthan50nmareobservedina higher-magnificationimage(Fig.4d).Thethicknessofthecoatingswas evaluatedonsimilarimages.Itincreasesupontheincreasein car-boncontentfrom1mmforAl2O3toabout7mmforT18andabout
5mmforG21.
Aboveacarboncontentof10wt.%,thehigherthicknessofthe TcoatingscomparedtothatoftheGcoatingscouldbeattributed tothehigherviscosityofthecorrespondingsols.Thearithmetic average roughness (Ra) calculated from white-light
interferen-tial rugosimetry images (not shown) is equal to about 0.4mm for both the steel substrate and the Al2O3 coating, 0.8mm for
theGcoatingsandonly0.2mmfortheTcoatings,inagreement with the FESEM observations where the T18 coating (Fig. 4a) seemstobemorelevelingtowardstheroughnessofthesteel sub-strate.
3.2. Nano-indentation
Typicalload-penetrationdepthplotsmeasuredwithaBerkovich nano-indenterareshown in Fig.5a. Themaximum penetration depthwaskeptlowerthanthetenthofthecoatingthicknessforall samples,inordertoneglecttheinfluenceofthesubstrateonthe calculatedmechanicalproperties.Analysisofsuchcurvesbythe OliverandPharrmethod[26]showsthattheYoung’smodulusof thecoatings(Fig.5b)steadilydecreasesupontheincreasein car-boncontent,from41GPaforAl2O3to9GPaforG21.Thehardness
Fig.8.Opticalimagesshowingtheweartrackfortheun-coatedsubstrate(a)andtheAl2O3(c),CNT/Al2O3(18wt.%)(e)andgraphite/Al2O3(21wt.%)(g)coatingsandthe
correspondingsteelballs(b,d,fandh,respectively).
andthendecreasessharply,reachingonly200MPaforG21.FESEM observationofthesurfacesdonotindicateacorrespondinglylower densification,thusthiscouldreflectalowercrystallizationstate ofAl2O3,lesscohesivegrainboundariesoraweakcarbonmatrix
interface.
3.3. Tribologicaltesting
Typicalcurvesshowingthefrictioncoefficientagainstasteelball versustheslidingdistanceareshowninFig.6.Thecurvesforthe un-coatedsteelandAl2O3coatingareveryunstable,whichcould
Fig.9.WidthoftheweartracksversuscarboncontentfortheAl2O3( ),CNT/Al2O3
(N)andgraphite/Al2O3( )coatings.
(itwasneitherquantifiednorimaged).FortheT18coating,there isarunning-inperiodofabout20m, withamarkedincreaseof thefrictioncoefficient,andafterwhichitisstabilized.Thiscould revealthatthecoatingisseparatedfromthesubstrate(reflecting totalwear)veryearlyinthetest,possiblyreflectingaweak coat-ing/substrateinterface.Theso-produceddebriswouldstayinthe trackandseeminglyshowalubricatingrolethatpureAl2O3does
not,whichcouldbeinagreementwithreportsbyotherauthors [18,19].FortheG21coating,thereisasmallbutregularincrease uptoaslidingdistanceofabout200m.
Therefore,forallspecimens,theaveragesteady-statefriction coefficient(m)wascalculatedforthelast50mofthetests.The Al2O3 coating (=0.95) doesnot provideany improvementon
frictionbehavior over un-coatedsteel (=0.80).By contrast,m decreasesupontheincreaseincarboncontent(Fig.7),very progres-sivelyfortheTcoatings(reaching0.26forT18)andmoreabruptly fortheGcoatingsforwhichaplateauatabout=0.17isreached forG5.
Post-testoptical observationsof the weartracksof selected coatingsandthecorrespondingsteelballsarepresentedinFig.8. The wear tracks of the composite coatings appear to be less wide and partially covered with a black film-like layer. This tribologically-formedfilmwastransferredtothesteelball, sug-gestingthattheobservedlubricatingeffectduringslidingmaybe atleastpartiallyrelatedtothesmearingofatribofilmoverthe con-tactarea.Thewidthoftheweartracks(Fig.9),measuredonsimilar opticalimages,islowerforthecompositecoatingsthanforthe Al2O3coating(>1000mm).Thevalueisconstant(about650mmfor
theTsamples)anddecreasingdownto300mmfortheGspecimens. Weartracksofselectedsamples(T18andG21)wereobserved bynon-contactopticalprofilerimaging (Fig.10aand c)andthe correspondingprofilesperpendiculartothetrack(Fig.10bandd) wereusedtomeasuretheirhalf-widthanddepth.NotethatforT18 (Fig.10b),thedepthoftheprofileisclosetothecoatingthickness (7mm,Fig.7)whichcouldconfirmtotalwearasnotedabove.The wearvolume(Vw)wascalculatedforahalf-ellipsetrackaccording
toEq.(1):
VW
!
mm3=1/2.prd.2pL (1)with r=track half-width (mm), d=track depth (mm), L=outer radiusofthecirculartrack(10mm).
Fig.10.White-lightinterferometryimagesoftheweartracksandthe
correspond-ingprofilesfortheCNT/Al2O3(18wt.%)coating(aandb)andthegraphite/Al2O3
(21wt.%)coating(candd).
ThecalculatedwearvolumeforT18andG21isequalto0.20and 0.06mm3,respectively.G21isthemostwear-resistantcomposite
coatingcomparedtoun-coatedsteel(0.30mm3)orAl
2O3coating
(0.12mm3).
4. Discussion
Althoughadetailedinvestigationofthewearmechanismsas performedbyotherauthors[18,19]iswellbeyondthescopeofthis workandwarrantsfurtherstudies,itwasfoundofparticular inter-esttostudytheweartracksbyRamanspectroscopy.TypicalRaman spectraforT18andG21,insideandoutsidethewornsurface,are showninFig.11.TheratiobetweentheintensitiesoftheDbandand theGband(ID/IG)wascalculatedfromthespectra.Ahigherratio
isgenerallyattributedtothepresenceofmorestructuraldefects (moresp3carbon).Interestingly,thespectraforT18(Fig.11a)are
remarkablysimilartoeachotherandtheID/IGratioaresimilartoo
(1.83outsideand1.90inside).Moreover,itisalsosimilartotheID/IG
ratio(1.86and1.88)reported[23]fortherawandfunctionalized CNTs(sameprovider,samebatch),respectively.Theseresults indi-catethattheCNTsdonotsufferdrasticdeterioration.Itisprobable thatthemoredefectiveCNTs,thosewithahighnumberofwalls, aredestroyedbecauseofthestrongwear,veryearlyinthefriction test,resultingintheformationofacarbon/CNT/Al2O3tribofilm,the
subsequentbehaviorofwhichdoesnotfavorCNTdestructionand delamination.Bycontrast,theID/IGratioforG21(Fig.11b)increases
significantlyfrom0.42(outside,and0.40forrawgraphite)to1.22 (inside),whichcouldindicatesomedestructionofthegraphite
par-Fig.11.TypicalRamanspectraoutsideandinsidetheweartrackfortheCNT/Al2O3
(18wt.%)(a)andgraphite/Al2O3(21wt.%)(b)coatings.
ticlesduringthefrictiontest.Theso-generateddebriswouldform alubricatingfilm,loweringtheshearingresistance.
Asmentioned above,theCNTsdo notappear tobestrongly damaged duringthefrictiontest andtherefore thetotal carbon contentinthesamplesishigherthanthecontributing,or “use-ful”,carbon.Therelativeweightofeachwallwascalculated[27] anda plot(similar totheoneinFig.7)ofthesteady-state fric-tioncoefficientversusthecorrected carboncontentwasdrawn (Fig.12a),consideringthattheinnermostwallisinactive(i.e.one countsonly7walls),thenthatthetwoinnermostwallsareinactive (i.e.onecountsonly6walls),andsoonuntilonecountsonlythe outerwall,whichcontributesonly16.5%tothetotalweightofthe CNT.Thecorrespondingcurvesgraduallyshifttotheleftand inter-estinglythedatawhenconsideringonlytheouterwall(labelled 1WinFig.12a)fallreasonablyinlinewiththoseofthegraphite curve,formingwhatweconsidertobea“master-curve”.This sug-geststhatinsteadofusingweightorvolumiccarboncontents,the appropriateunitfortribologicalbehaviorcouldbethesurfacearea developedbytheinvolvedcarbonspecies.FortheCNT/Al2O3
spec-imens,thecontributingsurfaceareaisthusestimated,admittedly roughly,asthatcalculatedfortheouterwallonly,consideringthe geometriccharacteristicsoftheCNTsdescribedinSection2.1and nobundling.Assumingthatthecontributingsurfaceareais con-stantduringthefrictiontest,foragivensample,thecorresponding plotoffrictioncoefficientversussurfaceareaformsthebasisof
Fig.12.(a)Steady-statefrictioncoefficientversuscorrectedcarboncontent,
con-sideringonlythenumberofwallsconsideredactivefortheprocess(seetextfor
details).Curvesfor8,4,2and1wallsareshown.Thearrowshowsthegradualshift
totheleft;(b)steady-statefrictioncoefficientversusthesurfaceareaofthecarbon
speciesinvolvedinthefrictionphenomena(seetextfordetails)fortheCNT/Al2O3
(triangles)andgraphite/Al2O3( )coatings.Theerrorbarshavebeenomittedfor
clarity.
another“master-curve”(Fig.12b).For thegraphite/Al2O3
speci-mens,thecarbonflakeswerefirstdescribedascylinders1.5mmin diameterand1.5mmthick(whichobviouslyisfartoothick)toget athestartingdataforthecalculation:theinitialnumberofcarbon flakesforagivenweightcontentandthecorrespondingsurfacearea (theverticallateralareaofthecylinderwasneglected).Then,the cylinderswerehorizontally“sliced”inhalf,thuseachtimedoubling thesurfaceareadevelopedbythecarbon,untilareasonablefitwith the“master-curve”(Fig.12b)wasobtained.Thisoccurredafter slic-ing11timesandcorrespondstoacarbonlayerabout730pmthick, i.e.closetotwolayersofgraphene,whichreasonablysupportsthe findingthatitisthegraphene(graphite)surfaceavailablethatis importantforthereductionoffrictionandwearandthatitsinitial form(eitherCNTsorgraphiteflakes)isun-important.Therefore, ourresultssuggestthatselectinggraphiteflakesismore conve-nientbecause,firstitiseasiertodisperseintotheboehmitesol andseconditismorereadilydelaminatedandbrokenintothinner flakes(alsocalledfew-layered-graphene)duringthetest,therefore providingmoredesirablelubricatingsurfacearea.
5. Conclusions
Carbon/Al2O3coatingson304-Lstainlesssteelpreparedusing
eitherCNTs(averagenumberofwalls=8)orgraphiteflakespresent alowerfrictioncoefficient(reducedbyafactorof4–5)andlower wear(reducedbyafactorof 2forgraphiteflakes)againststeel compared to a pureAl2O3 coating. The CNTs and
micrometer-sizedgraphiteagglomeratesarehomogenouslydispersedwithin thecoatings,thethicknessofwhichisinthe1–7mmrange.There isnodramaticdeteriorationofthestructureoftheCNTsduringthe frictiontestwhereasthegraphiteflakesarepartiallydestroyed. TheCNT/Al2O3coatingisseparatedfromthesubstrateveryearly
inthetest,reflectingtotalwear,andtheso-produceddebriswould showa lubricating rolethat pureAl2O3 doesnot,but however
lessefficientlythanforgraphite/Al2O3coatings.Theso-generated
debris forma lubricatingfilm,which is transferredtothesteel ball,suggestingthattheobservedlubricatingeffectduringsliding maybeatleastpartiallyrelatedtothesmearingofatransferred filmoverthecontactarea.Somesimplemodelingshowsthatonly theouterwallof theCNTscontributes tothesliding, in agree-mentwiththeobservedabsenceofmajordeterioration,andthat thetotalavailablecarbonsurfaceareainthespecimenisthe rele-vantparameter.Therefore,itisshownthatgraphiteflakesaremore efficientthanCNTs,atleastintheseexperimentaltribological con-ditions,becausetheyarereadilydelaminatedintothinnerflakes (few-layered-graphene)duringthetest,providingmoredesirable lubricatingsurfacearea.Theseresultsprovideimportantguidelines forthedesignofself-lubricatingcarbon/Al2O3coatingsandfuture
workdirectionsinclude:thepreparationofgraphite/Al2O3coatings
withthinnergraphiteflakes,andanin-depthstudyofthe influ-enceofthechangeinthecoatingsthickness,Young’smodulusand hardnessevolutionsthatwereshowntooccurupontheincreasein carboncontent.
Acknowledgements
ThisworkwasperformedwithintheframeworkofCETIMAT, ajointlaboratorybetweenCETIMandCIRIMAT.K.Hentourand A. Marsal thank CETIM for the doctoral thesis grants. FESEM observationsandelectronmicroprobeanalyseswereperformedat theCentredemicrocaractérisationRaimondCastaing—UMS3623, Toulouse.TheauthorsalsothankDr.Ph.DeParsevalforassistance intheelectronmicroprobeanalyses.
References
[1]J.P.Tu,Y.Z.Yang,L.Y.Wang,X.C.Ma,X.B.Zhang,Tribologicalpropertiesof
carbonnanotube-reinforcedcoppercomposites,Tribol.Lett.10(2001)
225–228,http://dx.doi.org/10.1023/A.1016662114589.
[2]S.R.Dong,J.P.Tu,X.B.Zhang,Aninvestigationoftheslidingwearbehaviorof
Cu-matrixcompositereinforcedbycarbonnanotubes,Mater.Sci.Eng.A313
(2001)83–87,http://dx.doi.org/10.1016/S0921-5093(01)00963-7.
[3]W.X.Chen,J.P.Tu,L.Y.Wang,H.Y.Gan,Z.D.Xu,X.B.Zhang,Tribological
applicationofcarbonnanotubesinametal-basedcompositecoatingand
composites,Carbon41(2003)215–222,
http://dx.doi.org/10.1016/S0008-6223(02)00265-8.
[4]B.Lim,C.Kim,B.Kim,U.Shim,S.Oh,B.Sung,J.Choi,SeunghyunBaik,The
effectsofinterfacialbondingonmechanicalpropertiesofsingle-walled
carbonnanotubereinforcedcoppermatrixnanocomposites,Nanotechnology
17(2006)5759,http://dx.doi.org/10.1088/0957-4484/17/23/008.
[5]K.T.Kim,S.I.Cha,S.H.Hong,Hardnessandwearresistanceofcarbonnanotube
reinforcedCumatrixnanocomposites,Mater.Sci.Eng.A449–451(2007)
46–50,http://dx.doi.org/10.1016/j.msea.2006.02.310.
[6]S.R.Bakshi,D.Lahiri,A.Agarwal,Carbonnanotubereinforcedmetalmatrix
composites—areview,Int.Mater.Rev.55(2010)41–64,http://dx.doi.org/10.
1179/095066009X12572530170543.
[7]C.B.Lin,Z.-C.Chang,Y.H.Tung,Y.-Y.Ko,Manufacturingandtribological
propertiesofcoppermatrix/carbonnanotubescomposites,Wear270(2011)
382–394,http://dx.doi.org/10.1016/j.wear.2010.11.010.
[8]K.Rajkumar,S.Aravindan,Tribologicalstudiesonmicrowavesintered
copper-carbonnanotubecomposites,Wear270(2011)613–621,http://dx.
doi.org/10.1016/j.wear.2011.01.017.
[9]C.Guiderdoni,C.Estournès,A.Peigney,A.Weibel,V.Turq,C.Laurent,The
preparationofdouble-walledcarbonnanotube/Cucompositesbyspark
plasmasintering,andtheirhardnessandfrictionproperties,Carbon49(2011)
4535–4543,http://dx.doi.org/10.1016/j.carbon.2011.06.063.
[10]C.Guiderdoni,E.Pavlenko,V.Turq,A.Weibel,P.Puech,C.Estournès,A.
Peigney,W.Bacsa,C.Laurent,Thepreparationofcarbonnanotube
(CNT)/coppercompositesandtheeffectofthenumberofCNTwallsontheir
hardness,frictionandwearproperties,Carbon58(2013)185–197,http://dx.
doi.org/10.1016/j.carbon.2013.02.049.
[11]Z.Xu,Q.Zhang,X.Shi,W.Zhai,Q.Zhu,Comparisonoftribologicalproperties
ofNiAlmatrixcompositescontaininggraphite,carbonnanotubes,or
graphene,J.Mater.Eng.Perform.24(2015)1926–1936,http://dx.doi.org/10.
1007/s11665-015-1482-5.
[12]J.-W.An,D.-H.You,D.-S.Lim,Tribologicalpropertiesofhot-pressed
alumina-CNTcomposites,Wear255(2003)677–681,http://dx.doi.org/10.
1016/s0043-1648(03)00216-3.
[13]D.-S.Lim,D.-H.You,H.-J.Choi,S.-H.Lim,H.Jang,EffectofCNTdistributionon
tribologicalbehaviorofalumina-CNTcomposites,Wear259(2005)539–544,
http://dx.doi.org/10.1016/j.wear.2005.02.031.
[14]Z.H.Xia,J.Lou,W.A.Curtin,Amultiscaleexperimentonthetribological
behaviorofalignedcarbonnanotube/ceramiccomposites,Scr.Mater.58
(2008)223–226,http://dx.doi.org/10.1016/j.scriptamat.2007.09.039.
[15]G.Yamamoto,M.Omori,K.Yokomizo,T.Hashida,K.Adachi,Structural
characterizationandfrictionalpropertiesofcarbonnanotube/alumina
compositespreparedbyprecursormethod,Mater.Sci.Eng.B148(2008)
265–269,http://dx.doi.org/10.1016/j.mseb.2007.09.013.
[16]I.Ahmad,A.Kennedy,Y.Q.Zhu,Wearresistantpropertiesofmulti-walled
carbonnanotubesreinforcedAl2O3nanocomposites,Wear269(2010)71–78,
http://dx.doi.org/10.1016/j.wear.2010.03.009.
[17]A.Kasperski,A.Weibel,D.Alkattan,C.Estournès,V.Turq,C.Laurent,A.
Peigney,Microhardnessandfrictioncoefficientofmulti-walledcarbon
nanotube-yttria-stabilizedZrO2compositespreparedbysparkplasma
sintering,Scr.Mater.69(2013)338–341,http://dx.doi.org/10.1016/j.
scriptamat.2013.05.015.
[18]K.Balani,S.P.Harimkar,A.Keshri,Y.Chen,N.B.Dahotre,A.Agarwal,
Multiscalewearofplasma-sprayedcarbon-nanotube-reinforcedaluminum
oxidenanocompositecoating,ActaMater.56(2008)5984–5994,http://dx.
doi.org/10.1016/j.actamat.2008.08.020.
[19]A.K.Keshri,A.Agarwal,Wearbehaviorofplasma-sprayedcarbon
nanotube-reinforcedaluminumoxidecoatinginmarineand
high-temperatureenvironments,J.Therm.SprayTechnol.20(2011)
1217–1230,http://dx.doi.org/10.1007/s11666-011-9669-2.
[20]C.-Y.Yu,B.J.Kellett,Tribologyofalumina–graphitecomposites,in:J.B.WJr.
(Ed.),Proc.20thAnnu.Conf.Compos.Adv.Ceram.Mater.Struct.Ceram.Eng.
Sci.Proc.,JohnWiley&Sons,Inc.,1996,pp.220–227,http://dx.doi.org/10.
1002/9780470314821.ch26http://onlinelibrary.wiley.com.
[21]J.Song,Y.Zhang,Y.Su,Y.Fang,L.Hu,Influenceofstructuralparametersand
compositionsonthetribologicalpropertiesofalumina/graphitelaminated
composites,Wear338–339(2015)351–361,http://dx.doi.org/10.1016/j.wear.
2015.07.017.
[22]A.Peigney,C.Laurent,E.Flahaut,R.R.Bacsa,A.Rousset,Specificsurfacearea
ofcarbonnanotubesandbundlesofcarbonnanotubes,Carbon39(2001)
507–514,http://dx.doi.org/10.1016/S0008-6223(00)00155-X.
[23]A.Kasperski,A.Weibel,C.Estournès,C.Laurent,A.Peigney,Multi-walled
carbonnanotube—Al2O3composites:covalentornon-covalent
functionalizationformechanicalreinforcement,Scr.Mater.75(2014)46–49,
http://dx.doi.org/10.1016/j.scriptamat.2013.11.015.
[24]A.Marsal,F.Ansart,V.Turq,J.P.Bonino,J.M.Sobrino,Y.M.Chen,J.Garcia,
Mechanicalpropertiesandtribologicalbehaviorofasilicaor/andalumina
coatingpreparedbysol–gelrouteonstainlesssteel,Surf.Coat.Technol.237
(2013)234–240,http://dx.doi.org/10.1016/j.surfcoat.2013.06.037.
[25]A.Peigney,F.L.Garcia,C.Estournès,A.Weibel,C.Laurent,Tougheningand
hardeningindouble-walledcarbonnanotube/nanostructuredmagnesia
composites,Carbon48(2010)1952–1960,http://dx.doi.org/10.1016/j.carbon.
2010.01.063.
[26]W.C.Oliver,G.M.Pharr,Measurementofhardnessandelasticmodulusby
instrumentedindentation:advancesinunderstandingandrefinementsto
methodology,J.Mater.Res.19(2004)3–20,http://dx.doi.org/10.1557/jmr.
2004.0002.
[27]C.Laurent,E.Flahaut,A.Peigney,Theweightanddensityofcarbonnanotubes
versusthenumberofwallsanddiameter,Carbon48(2010)2994–2996,
