www.elsevier.com/locate/scriptamat
Improvingcorrosionresistanceoftitanium-coatedmagnesiumalloybymodifyingsurfacecharacteristicsofmagnesiumalloy
priortotitaniumcoatingdeposition
GuosongWu,a,*KejianDing,bXiaoqinZeng,aXueminWangaandShoushanYaoaaNationalEngineeringResearchCenterofLightAlloyNetForming,ShanghaiJiaoTongUniversity,Shanghai200030,China
bInstituteofBiologicalScienceandTechnology,BeijingJiaoTongUniversity,Beijing100044,China
Received16February2009;revised30March2009;accepted30March2009
Availableonline5April2009
NitrogenionswereimplantedintoAZ31magnesiumalloypriortotitaniumcoatingdeposition.Theionimplantationinduceddensificationoftheoriginaloxidefilmonthealloysurfacewiththeformationofaluminumnitrideandmagnesiumnitride,aswellasanimprovementintheadhesionbetweenthecoatingandthesubstrate.Therefore,thecorrosionresistanceoftheTi-coatedAZ31alloywiththeabovepretreatmentwassignificantlyimprovedin3.5wt.%NaClaqueoussolution.Ó2009ActaMaterialiaInc.PublishedbyElsevierLtd.Allrightsreserved.
Keywords:Magnesiumalloy;Coating;Ionimplantation;Sputtering;Corrosion
Physicalvapordeposition(PVD)coatingshave
beenpaidincreasingattentioninrecentyearsasameanstoimprovethepoorwearresistanceandpoorcorrosionresistanceofmagnesiumalloysduetotheirexcellentproperties,suchastheirdensestructure,goodinterfacebondingandinnocuitytoenvironment[1–6].Unfortu-nately,pinholesandcracksareinevitablycreatedwhenPVDcoatingsarepreparedbythecurrentmethods[7],andlocalcorrosionoccursinsuchthrough-thicknessdefectsofthesecoatingswhenthecorrosivesolutionattacksthecoatedsample.
Inordertoreduceoreliminatetheeffectofthrough-thicknessdefects,suchascracksandpores,YuandUan[8]appliedtheprincipleofcathodicprotectionandpre-paredpuremagnesiumfilmsasprotectivecoatingsonAZ91Dmagnesiumalloysbythevacuumevaporationtechnique.However,thiscathodicprotectionwastooweakduetothesmallelectrochemicalpotentialdiffer-encebetweenpuremagnesiumandAZ91D.Anodicpro-tectionisanothertypicalprotectionmethodformetalmaterials.However,thereisnopassivationinthepro-cessofanodicpolarizationformagnesiumalloyinmostcorrosivesolutions,e.g.NaClaqueoussolution.Thus,it
seemsthatitisverydifficulttoapplytheideaofanodicprotectiononmagnesiumalloy.
Generally,foramagnesiumalloysubstratewithanoblecoating,thesubstratewillactastheanodeandthecoatingwillactasthecathodeingalvaniccellswhenthecoating(withdefects)isexposedtoacorrosiveenvi-ronment.Inthegalvaniccouple,thesubstrateisanodi-callypolarizedtothecoupledpotential[9].Ifthiscaseissimplified,theformulacanbeusedtodepicttheanodicpolarizationstateofthesubstrate:
VasÀVcorr
Ias¼Icorrexp
basHere,IcorrandVcorrarethecorrosioncurrentdensityandcorrosionpotentialofthesubstrateinthestateofnaturalcorrosion,respectively,basistheanodicTafelslopeofthesubstrate,andVasandIasarethecorrosionpotentialandcorrosioncurrentdensityofthesubstrateinthestateofanodicpolarization,respectively[10].IfIcorrandVasÀVcorraredecreasedandbasisincreased,Iaswilldecrease.Corrosioniscloselyrelatedtothesur-facecharacteristicsofmaterials.Thus,ifthecorrosionbehaviorofthesubstrateischangedreasonablybysur-facemodification,itwillbealsopossibleforustoimprovethecorrosionresistanceofthecoatedmagne-siumalloy.Inthisstudy,weselectedasputteredTicoat-ingasaprotectivecoatingonAZ31magnesiumalloy
*Correspondingauthor.Tel./fax:+8657486685036;e-mailaddresses:wugsjd@126.com,wuguosong@nimte.ac.cn
1359-6462/$-seefrontmatterÓ2009ActaMaterialiaInc.PublishedbyElsevierLtd.Allrightsreserved.doi:10.1016/j.scriptamat.2009.03.061
270G.Wuetal./ScriptaMaterialia61(2009)269–272
andinvestigatedtheeffectoftheintroductionofnitro-genionimplantationasthepretreatmentofthesubstratepriortocoatingdepositiononthecorrosionresistanceofthecoating/substratesysteminNaClaque-oussolution.
TheextrudedAZ31magnesiumalloyplatesservedassubstratesinthisstudy.TheyweregroundwithSiCem-erypaperupto#2000andthenpolishedwithAl(averagesize1lm).Amulti-functionalion-plating2Opaste3machine,whichisequippedwithaDCmagnetronsourceandagasplasmasource,wasappliedtoperformthesurfacemodificationofmagnesiumalloyinthisstudy.Thespecimenswereultrasonicallywashedinpurealcoholfor5minandthensenttothevacuumchamber.Whenthebasepressuredecreasedto1.8Â10À3Pa,thespecimenswerefirstimplantedwithadoseof5Â1017N+cmÀ2atanionenergyof35keVtoobtainanimplantedlayerwithhighnitrogencontent.After1hofimplantationandabout15minofnaturalcooling,thespecimensweremovedtothemagnetronsputteringsourcewithoutbreakingvacuum.Subsequently,thedepositionwasperformedwithargonasthesputteringgastoobtainaTicoatingwithathicknessofabout2lm.Theexperimentalparametersareshownasfol-lows:workpressureof0.24Pa,sputteringvoltageof305V,sputteringcurrentof1.0Aanddepositiontimeof1h.
Atomicforcemicroscopy(AFM)andfieldemissionscanningelectronmicroscopy(FESEM)wereperformedtoobservethesurfaceandcross-sectionmorphologyoftheTi-coatedsamples.AFMwasalsousedtoobservethesurfacemorphologyofAZ31beforeandafterimplantation.Augerelectronspectrometry(AES)wasusedtoanalyzethechemicalcompositionsneartheinterfaceofcoatingandsubstrate.Accordingtothespe-cificrequirementsoftheAEStest,wepreparedathinnerTifilmontheN-implantedAZ31underthesameexper-imentalconditionsasthatdepictedintheabovesection.X-rayphotoelectronspectrometry(XPS)wasalsoap-pliedtoanalyzethechemicalstateofthoseelementsintheN-implantedAZ31.Aftersubtractingtheback-groundsignal,thespectrawerefittingwithGaussian–Lorentzianpeakshapes.ThedeconvolutedpeakswereidentifiedbyreferencetoanXPSdatabase[11].Animmersiontestwasappliedtoevaluatethecorrosionresistanceofthespecimensinthisstudy.After24hofimmersionina3.5wt.%NaClsolution,thespecimensweretakenoutandobservedbyeye.Thecross-sectionsofthecorrodedspecimenswereobservedbyscanningelectronmicroscopy(SEM).Polarizationcurveswerealsousedtoinvestigatethecorrosionbehaviorofallthespecimens.Theexposedareawas10Â10mm2.Ineachcase,threeexperimentswereconducted.Thepolar-izationtestswerecontrolledbythePARSTAT2273advancedelectrochemicalsystemusingtheconventionalthree-electrodetechnique.Thepotentialwasreferredtoasaturatedcalomelelectrodeandthecounterelectrodewasaplatinumsheet.Thetestswerecarriedoutat1mVsÀ1atroomtemperaturein3.5wt.%NaClsolution.
Figure1(a)and(b)showsthesurfacemorphologyoftheTi-coatedsamplewithoutandwiththepretreatmentofN-implantation,respectively.Figure1(c)showsthe
Figure1.SurfacemorphologyofthecoatedsamplesobservedbyAFM:(a)withoutand(b)withN-implantationpretreatment.(c)Cross-sectionofthecoating/substratesystemswith(top)andwithout(bottom)thepretreatment.(d)PinholesontheTicoatingobservedatamagnificationofÂ160,000.Theinsertshowsamicrodefectinthecoating.Here,thesamplehasbeentreatedbyionimplantationbeforeTideposition.
cross-sectionsofthecoating/substratesystemswithandwithoutthepretreatment,whichwereobservedbyFESEMaftermechanicalpolishing.Bycomparison,thereisnodistinctdifferenceinthesurfacemorphologyandcross-sectionmorphologybetweenthesampleswithandwithoutthepretreatment.PinholesintheTicoat-ing,whichareformedinthedepositionduetothesha-doweffect,canbeclearlyobservedbyFESEMatÂ160,000magnification.Inadditiontothesepinholes,severalmicrodefectsarealsopresent,randomlydistrib-utedinthecoating,oneofwhichisshownintheinsetofFigure1(d).Wangetal.[12]evenreportedthatanoxidelayerexistedonthesurfaceoftheas-polishedAZ31magnesiumalloy.Figure2showsthesurfacemorphologyoftheAZ31substratebeforeandafterimplantation.Clearly,theoxidelayeronthemodifiedsurfacebecamemuchdenserthantheoriginalsurface.Figure3(a)showstheatomconcentrationinthecoat-ing/substratesystemasafunctionofsputteringtime.TheAESanalysisoftheelementaldistributionshowsthattheoxidelayerbetweentheTicoatingandthealloymatrixhasbeenfullyimplanted.ElementNexhibitsaGaussiandistributioninthisimplantedlayer.ElementAliscongregatedintheimplantedlayer,whichisiden-tifiedbycomparingtheAlcontentintheimplanted
Figure2.AFMimagesshowingthesurfacemorphologyofAZ31substrate(aandb)beforeimplantationand(candd)afterimplantation.
G.Wuetal./ScriptaMaterialia61(2009)269–272271
Figure3.Augerelectronspectraof(a)AugerdepthprofileoftheTi-coatedsamplewiththepretreatmentofN-implantation;(b)AugerlineshapesofAlKLLatdifferentdepths;and(c)AugerlineshapesofMgKLLatdifferentdepths.High-resolutionXPSspectraof(d)N1sand(e)O1satdifferentdepthsoftheN-implantedAZ31magnesiumalloy.
layerwiththatinthealloymatrix.Figure3(b)and(c)showstheAugerlineshapesofAlKLLandMgKLLatvariousetchingdepthscorrespondingtothoseposi-tionsdesignatedbythedashedlinesinFigure3(a).ThepositionofMgKLLmovedfrom1178to1186eVastheetchingdepthincreasingfromdepthAtodepthC.Bouvieretal.[13]proposedthatthechem-icalstatesofelementMgindifferentenvironmentscanbedistinguishedbyAESspectra.Here,theMgKLL
ofdepthBhasthesameelectronenergyasthatofdepthC.BasedonthefactthatdepthCisintheinnerofthealloymatrix,thisindicatesthatelementMginametallicstate.Thus,itcanbefurtherconcludedbycomparisonthatelementMgatdepthBisalsoinametallicstate.ThepositionofAlKLLofdepthBisdifferentfromthatofdepthC.
ThecontentofelementOisverysmallatthesamedepth,asshowninFigure3(a).ThismeansthatelementAlmustbondwithelementNtoformaluminumnitride.XPSwasappliedtodetectthevalencestatesoftheelementsinthreedifferentetchingdepthsoftheim-plantedAZ31.Withincreasingdepthinvestigated,thecontentofelementNincreasesandthatofelementOde-creases,incorrespondencewiththeresultoftheAESanalysis.AsshowninFigure3(d),theoriginalpeakcanbedeconvolutedintotwopeakswhosebindingener-giescorrespondtoMg3N2andAlN.Withincreasingdepth,theN1speakbecomesincreasinglycomposedofelementNfromAlN.Similarly,Figure3(e)showsthattheoccurrenceoftheO1speakisattributedtoelementOfromMgOandAl2O3.TheXPSanalysisindicatesthatMg3N2andAlNasnewphaseshavebeenformedintheoxidelayerduringionimplantation.
Figure4(a)presentstheappearanceofallspecimensafter5and24hofimmersionin3.5wt.%NaClsolution.Clearly,lessdamagedregionsoccurredonthesurfaceoftheTi-coatedAZ31withN-implantationpretreatmentcomparedtothesurfaceofotherspecimens.ThismeansthattheTi-coatedspecimenwiththeN-implantationpretreatmenthadthebestcorrosionresistanceamongallspecimens.Figure4(b)–(d)showsthecross-sectionsofallsamplesafter24hofimmersion,whichexhibitthecorrosioncharacteristicsofallsamples.ForbothTi-coatedsamples,localcorrosionisthemainmodeoftheircorrosionfailure.AccordingtotheimagesshowninFigure4(c)and(d),theprotectedcoatedregionisintactwhiletheregionexposedtosolutionisseverelycorroded.Moreover,delaminationcanalsobeseenat
Figure4.(a)Appearanceofthespecimensaftertheimmersiontest.(b,candd)Cross-sectionsofbareAZ31,theTi-coatedAZ31withthepretreatmentandtheTi-coatedAZ31withoutthepretreatment,respectively,observedbySEMafter24hofimmersion.Theinsetof(c)showsoneofthescenesinotherregions.(e)PolarizationcurvesoftheN-implantedAZ31andbareAZ31.(f)PolarizationcurvesoftheTi-coatedAZ31withandwithoutN-implantationpretreatment.
272G.Wuetal./ScriptaMaterialia61(2009)269–272
thecoating/substrateinterfaceofthesamplewithoutthepretreatmentbutnotinthepretreatedsample.Thismeansthatthesamplewiththepretreatmenthadbettercoating/substratebondingthanthatwithoutthepretreatment.InFigure4(e),thepolarizationcurveoftheN-implantedAZ31isshowncomparedwiththatofbareAZ31.ThecorrosionpotentialoftheN-im-plantedAZ31isclosetothatofbareAZ31.Thecorro-siondensityoftheN-implantedÀ5AZ31isdecreasedfrom2.38Â10À5to1.01Â10AcmÀ2.Itisworthnotingthattheslopeofthepolarizationcurveintheanodicregionissteeperafterionimplantation.Thatis,theanti-dissolutionabilityoftheanodicpolarizationregionisimproved.Figure4(f)presentsthepolarizationcurvesoftheTi-coatedAZ31withandwithoutthepretreat-mentofN-implantation.TheyhavethesamecorrosionpotentialasthatofbareAZ31,butshiftedpositivelybyabout195mV.However,thereisasignificantdifferenceinthatthecurrentdensityofthecoating/substratesys-temwiththepretreatmentismuchsmallerintheanodicregionthanthatofthesystemwithoutthepretreatment.Magnesiumalloyispronetogalvaniccorrosionduetoitsnegativepotential[14].Whenathrough-thicknessholeoftheTicoatingisexposedtoacorrosiveenvi-ronment,agalvaniccellwillbeformedbetweenthecoatingandthesubstrate.Ionimplantationasapre-treatmentwasintroducedintothePVDcoatingprocessbyXuandLiu[15–18].Becausetheprocessinvolvescollision,thedirtysubstanceontheoriginalsurfaceofthesubstratewillbecleanedinthisprocess.Ionbombardcanalsoactivatethesurfaceatoms,whichisbeneficialtothebondingbetweenthesubstrateandthecoating.
Tothebestofourknowledge,theeffectofN-implan-tationonthecorrosionresistanceoftheTi-coatedAZ31hasnotbeenreported.Tianetal.[19]studiedthecorro-sionbehaviorofanN-implantedmagnesiumalloyinNaClsolutionsandhypothesizedthattheimprovementinthecorrosionresistanceofAZ31afterNionimplan-tationwasattributedtothecompactedoriginaloxidelayerscausedbythebombardmenteffect,buttheydidnotprovideanyproofofdensification.Asdepictedinourstudy,theoxidelayerwasdensifiedwiththeforma-tionofnewphasesafterionimplantation,asconfirmedbyAFMobservation.Whenthecorrosivesolutionpassesthroughtheporestotheinterfaceofthecoatingandthesubstrate,theimplantedregioncanactasabar-rierthatcanresisttheaggressiontosomeextent.Huetal.[20]reportedthatMgH3N2isverysensitiveto2OandreadilyformsMg(OH)2inanaqueoussolu-tion.Fortunately,duringN-implantationNionstendtobondwithAlatomsandformAlN,whichhasgoodthermalstabilityandgoodcorrosionresistance[21].Combiningtheresultofthepolarizationcurves,Iandbseentohavealreadyshiftedincorrascanbetheexpecteddirection.Thus,nitrogenionimplantationasapretreatmenthelpsenhancethecorrosionresistanceofTi-coatedAZ31duetotheabovecombinedsynergis-ticeffects.
AsshowninFigure4,localcorrosionisthemaintypeoffailureinthesamplewiththecombinedtreatment.
TheoxidelayerofAZ31isdensifiedafterionimplanta-tion,butitcanbespeculatedfromtheexperimentalphenomenonthattherestillexistseveralweakregionsonthesurfaceofAZ31.Oncethecorrosivesolutionpen-etratestheporesinthecoatingandthedefectsintheoxidelayer,thealloymatrixwillbecorroded.Thus,thelocalcorrosioninthiscaseissimilartothatwithoutpretreatment.However,theapplicationofthispretreat-mentcanstillreducetheprobabilityofpitoccurrenceandenhancetheadhesionofthecoatingandsubstratetoimpedetheenlargementofcorrosionpits.Duetothecombinedsynergisticeffectsoftheanti-dissolutionabilityinanodicpolarizationofthesubstratestrength-eningandthecoating/substrateadhesionstrengthening,thiscombinedtreatmentisapotentialtechniqueforimprovingthepoorcorrosionresistanceofmagnesiumalloy.Iftheimplantedionsspeciesisoptimizedtoim-provetheanti-dissolutionabilityinanodicpolarizationofthesubstrateorotherprotectivecoatingsareselectedtoreducethedifferenceinthegalvanicseriesbetweenthecoatingandthesubstrate,itshouldbepossibletofurtherenhancethecorrosionresistanceofthecoating/substratesystem.
Insummary,nitrogenionimplantationwasintro-ducedasapretreatmenttechniqueinthefabricationoftheTicoating/AZ31substratesystem.Theionimplan-tationinduceddensificationoftheoxidelayerandtheformationofaluminumnitridenearthesurfaceoftheAZ31,bothofwhichimprovethecorrosionresistanceofAZ31inNaClaqueoussolution.Consequently,theTi-coatedsamplepretreatedwithN-implantationhadbettercorrosionresistancethanthatwithouttheabovepretreatment.
[1]J.E.Gray,B.Luan,J.AlloysCompd.336(2002)88.[2]B.Navinsek,P.Panjan,I.Milosev,Surf.Coat.Technol.116–119(1999)476.
[3]FrankHollstein,RenateWiedemann,JanaScholz,Surf.Coat.Technol.162(2003)261.
[4]E.Zhang,L.Xu,K.Yang,ScriptaMater.53(2005)523.[5]H.Hoche,C.Rosenkranz,A.Delp,Surf.Coat.Technol.193(2005)178.
[6]J.Senf,E.Broszeit,Adv.Eng.Mater.1(1999)133.[7]H.Altun,S.Sen,Surf.Coat.Technol.197(2005)193.[8]B.Yu,J.Uan,ScriptaMater.54(2006)1253.
[9]G.Song,CorrosionandProtectionofMagnesiumAlloy,ChemicalIndustryPress,Beijing,2006.
[10]C.N.Cao,PrincipleofCorrosionElectrochemistry,
ChemicalIndustryPress,Beijing,2004.[11]http://srdata.nist.gov/xps/.
[12]X.Wang,X.Zeng,G.Wu,J.AlloysCompd.437(2007)87.[13]Y.Bouvier,B.Mutel,J.Grimblot,Surf.Coat.Technol.
180–181(2004)169.
[14]G.Song,A.Atrens,Adv.Eng.Mater.1(1999)11.
[15]M.Xu,X.Cai,Q.Chen,J.Appl.Phys.101(2007)053520.[16]Y.Liu,L.Li,M.Xu,ThinSolidFilms493(2005)152.[17]Y.Liu,L.Li,M.Xu,Surf.Coat.Technol.200(2006)2672.[18]Y.Liu,L.Li,M.Xu,Mater.Sci.Eng.A415(2006)140.[19]X.B.Tian,C.B.Wei,S.Q.Yang,Surf.Coat.Technol.198
(2005)454.
[20]J.Hu,Y.Bando,J.Zhan,NanoLett.6(2006)1136.[21]T.Reier,S.Simson,J.W.Schultze,Electrochim.Acta43
(1998)149.
因篇幅问题不能全部显示，请点此查看更多更全内容