Degradationbehaviorof Mg-based biomaterials containing different long-period stacking ordered phases

镁合金,LPSO

OPEN

SUBJECTAREAS:

MECHANICALENGINEERINGMETALSANDALLOYS

DegradationbehaviorofMg-basedbiomaterialscontainingdifferentlong-periodstackingorderedphases

QiumingPeng1,JianxinGuo1,HuiFu1,XuechengCai1,YananWang1,BaozhongLiu1,2&ZhigangXu3

12

Received18July2013Accepted

13December2013Published9January2014

StateKeyLaboratoryofMetastableMaterialsScienceandTechnology,YanshanUniversity,Qinhuangdao066004,China,SchoolofMaterialsScience&Engineering,HenanPolytechnicUniversity,Jiaozuo454000,China,3NSFEngineeringResearchCenterforRevolutionizingMetallicBiomaterials,1601EastMarketStreet,Greensboro,NC27411,USA.

Correspondenceandrequestsformaterialsshouldbeaddressedto

Q.M.P.

(pengqiuming@gmail.

com)

Long-periodstackingordered(LPSO)phasesplayanessentialroleinthedevelopmentofmagnesiumalloysbecausetheyhaveadirecteffectonmechanicalandcorrosionpropertiesofthealloys.TheLPSOstructuresaremostlydividedto18Rand14H.However,todatetherearenoconsistentopinionsabouttheirdegradationpropertiesalthoughbothofthemcanimprovemechanicalproperties.HereinwehavesuccessfullyobtainedtwoLPSOphasesseparatelyinthesameMg-Dy-ZnsystemandcomparativelyinvestigatedtheeffectofdifferentLPSOphasesondegradationbehaviorin0.9wt.%NaClsolution.Ourresultsdemonstratethatafinemetastable14H-LPSOphaseingraininteriorismoreeffectivetoimprovecorrosionresistanceduetothepresenceofahomogeneousoxidationfilmandrapidfilmremediationability.TheoutstandingcorrosionresistantMg-Dy-Znbasedalloyswithametastable14H-LPSOphase,coupledwithlowtoxicityofalloyingelements,arehighlydesirableinthedesignofnovelMg-basedbiomaterials,openingupanewavenueintheareaofbio-Mg.

M

agnesium(Mg)alloyshaveattractedagreatdealofattentionasorthopedicbiodegradableimplantmaterialsduetotheirpropermechanicalpropertiescomparabletonaturalboneandgoodbiocompat-ibility1.Incomparisonwithceramicsorpolymericmaterials,metallicMgmaterialsaremoresuitablefor

load-bearingapplicationsduetotheircombinationofhighmechanicalstrengthandfracturetoughness2.Moresignificantly,thedegradationproductisasoluble,non-toxicoxidesorhydroxideswhichcantemporarilyenhanceosteoblastactivityanddecreaseosteoclastnumberduringboneremodelling3.TheseintriguingcharacteristicshavemadeitpossibleforMgalloystobedevelopedintodegradablebio-Mgimplants.However,themainbottleneckofdegradablebio-Mgstentsliesinitsfastcorrosionratewhichleadstoweakmechanicalintegrity.Therefore,duringthehealingprocesstheycannotprovideenoughmechanicalsupport,resultinginbloodvesselrestenosis4.Consequently,thedevelopmentofnewbio-Mg-basedmaterialbecomesacriticalissue.

Recently,Mgalloyswithlong-periodstackingordered(LPSO)phaseshavebeenfoundinternaryMg-Zn-RE(RE5Y,Gd,Tb,Dy,Hoetc.)alloys5–7.ThemajortypesofLPSOstructuresinvolvethe18Rand14Hmodels.TherelationshipbetweenLPSOphasesandmechanicalpropertieshasbeenwellconfirmed.Suzukietal.8reportedthatasmallamountofZncanincreasethenumberofstackingfaultsinMg-Ybasedalloysbyintroducing18R-LPSOphase,whereinthemovementofdislocationscanbeprohibited,resultingintheimprovementoftensilestrength.Yamadaetal.9pointedoutthattheadditionofZnisattributedtotheformationof18R-LPSOphaseinMg-2.1Gd-0.6Y-0.2Zr(at.%)alloy,inwhichboththestrengthandplasticityareenhancedsignificantlyatlowtemperatures.Theseimprovedmechanicalpropertiesarewellelucidatedintermsoftheincrementofcriticalresolvedshearstressofbasalornon-basalslips10,theinterfacebetweenLPSOphaseandMgmatrix11andtheformationofkinkbands9,12–14.

Comparedtomechanicalproperties,theeffectofLPSOphasesoncorrosionpropertiesisunclear.Ingeneral,theLPSOphaseactsasaheterogeneousspot,andthenthepitting/galvaniccorrosionmightbeincreased15.However,Zhangetal.16recentlyreportedthattheas-extrudedMg-Gd-Zn-Zralloyexhibitalowanduniformcorrosionrateof0.17mm/yinHank’ssolutionowingtothepresenceof14H-LPSOphase.Zhaoetal.17verifiedthatthecorrosionpropertiesofMg-Zn-Yalloycontaining18R-LPSOphaseinsimulatedbodyfluidarebetterthanconventionalengineeringMgalloyssuchastheAZ31,WE43,ZK60andZX60alloys.Unfortunately,bothstudiesdidnotprovidedetailedmechanismswithrespecttotheselowdegradationrates.

镁合金,LPSO

objectivesintheMg-2Dy-0.5Zn(at.%)alloy,inwhichboth18R-LPSOand14H-LPSOphasesareproducedindependentlybytuningtheheattreatmentprocessing.TheinfluenceofLPSOphasesoncorrosionpropertiesin0.9wt.%NaClhasbeeninvestigatedandthecorrosionmechanismshavebeenclarifiedindetail.Theseresultsprovidesomevaluablecluesfortheexplanationofdegradationbeha-viorsofdifferentLPSOphases,aswellasforthedeterminationofsuitablethermo-mechanicalprocessinMg-Zn-RE-basedalloyscon-tainingLPSOphasesinthefuture.

Figure1|DSCcurvesoftheas-castMg-2Dy-0.5Znalloyataheatrateof66

C/min.

Inaddition,thepresenceofdifferentLPSOstructuresiscloselyassociatedwithsolidificationprocessandheattreatmentcondition.ThecorrosionrateofMg97.25Zn0.75Y2alloycontainingtheLPSOstructurepreparedbyrapidcoolingissignificantlydependentonphasemorphology18.An18R-LPSOphasecanbedetectedinas-castordeformedsamples,whilsta14H-LPSOphasefrequentlyformsafterisothermalheattreatmentatelevatedtemperatures19.Both18Rand14HLPSOphasesareofparticularimportancetoimprovethemechanicalpropertiesfromtheviewpointofpractice,therefore,itisverynecessarytounderstandthedifferenceincorrosionprop-ertiesbetween18R-LPSOand14H-LPSOphases.

Basedontheaboveanalysis,twoquestionsareproposed:(i)whetherthecorrosionmechanismisthesameforboth18R-LPSOand14H-LPSOphases?(ii)whichphaseisbettertoimprovethecorrosionpropertiesalthoughbothofthemplayanessentialroleonimprovingthemechanicalproperties?Inordertoanswerthesequestions,itisbettertoobtainthedifferentLPSOphasesinthesamealloysystem,whichcaneliminatetheeffectofdifferentalloyingelementsoncorrosionproperties.Inthiswork,weachievethese

Results

Phasecharacterization.Twopeaksareobservedat539uCand633uCfromtheDSCheatingcurve(Fig.1)oftheas-castMg-2Dy-0.5Znalloy(MDZ-C),whereinthesecondarypeak(633uC)isverywide.TheDSCcoolingcurvealsorevealsthesimilartrend.Themerediscrepancyliesinthevariationofpeakpositions,whichismostlyaffectedbythermalhysteresis.AccordingtoMg-Dybinaryphasediagram20,thesecondarypeakcorrespondstomeltingprocess.Inordertoclarifythefirstpeak,thesampleissolid-solutiontreatedat545uCfor4h(MDZ-545).

ItcanbeseenfromFig.2thattheMDZ-Csampleismostlycomposedofa-Mgmatrix,anet-likeeutecticphaseandasand-wichedblock-shapedprecipitatewithanapproximatecompositionofMg87.88Zn4.91Dy7.21(Fig.2c),paratively,theMDZ-545samplemainlycontainsa-Mg,aeutecticphaseandafinelamel-lar-shapedphasewithanapproximatecompositionofMg89.34Zn4.42Dy6.24(Fig.2f).

Basedonthebright-fieldTEMmicrographoftheMDZ-Calloy(Fig.3a),wecanseethatatypicalsandwichedstructureisdistributedingrainboundaries.Inthecorrespondingselectedareaelectrondiffraction(SAED)patternofgrayphase(Fig.3b)withtheincidenceofelectronbeamdirectionof[1000]Mg(Fig.3c),someextraweakerspotsaredetectedatpositionsn/6(wherenisaninteger)betweendirectspotand(0002)Mgdiffraction,indicatingthattheLPSOphaseis18R-typestructure13,22.Therectangleprecipitateisdistributedalonggrainboundaries(Fig.3d),whichisascribedtoMg24(Zn,Dy)5eutecticphasewithfcccrystalstructure.Thecalculatedavalueis1.023nm(Fig.3e),whichisslightlylessthanthatofMg24Dy

5

Figure2|TypicalSEMgraphsofMDZalloysunderdifferentstates;(a)and(b)theMDZ-Calloy;(c)representativeelementalcompositionof18R-LPSOphase;(d)and(e)theMDZ-545sample;(f)representativeelementalcompositionsof14H-LPSOphase.

镁合金,LPSO

Figure3|(a)typicalTEMgraphoftheMDZ-Csample;(b)and(c)SAEDpatternsof18R-LPSOphasealong[1000]directionandMgmatrixin(a),respectively;(d)eutecticphaseintheMDZ-Calloy;(e)SAEDpatternofeutecticphasealong[114]directionin(d);(f)TEMgraphoftheMDZ-545sample;(g)SAEDpatternof14H-LPSOphasealong[1000]directionin(f).

phaseduetothesubstitutionofDybyZn.Thesimilarfinelaminar-shapedconfigurationisalsoobservedintheMDZ-545alloy(Fig.3f).However,thewidthisreducedandtheaspectratioisincreased.AccordingtotheSAEDoffinelayer-shapedprecipitatealong[1000]Mgdirection(Fig.3g),itisfoundthatsmallperiodicdiffractionspotsattheintervalof1/14ofdistancebetweendirectspotand(0002)Mgreflection.Thespotsof(00014)and(0002)Mgarecoincid-entinthesameposition.Basedonthediffractionpeakappearancealongc-directionandthepreviousresults7,23,itcanbeconfirmedthatthefinelaminar-shapedLPSOphaseis14H-typestructure.Opencircuitpotentialandimmersiontest.Fig.4depictstheevolu-tionofcorrosionpotential(vs.SCE)ofimmersiontest.TheOCPoftheMDZ-Calloyisincreasedfromapproximately21750mVto21579mVin20min.Incontrast,theMDZ-545alloymerelytook3mintoreacharelativelystablevalueof21639mV.Additionally,notethattheactivestatedissolutionpotential(Ecorr)oftheMDZ-C

alloyfluctuatesgreatlyincontrasttotheMDZ-545alloy,whichismainlyascribedtotheruptureofoxidationfilm24.

Theeffectoftheimmersiontimeonmicrostructuresandsurfacemorphologiesofcorrosionproductsarestudiedtounderstandtheformationprocessoftheoxidationfilm.FortheMDZ-Csample,thecoarseprecipitatesingrainboundariesarestillidentifiedclearlyafterimmersingfor10minor30min(Fig.5a-1and5b-1),however,theconfigurationbecomesunclearafterimmersingfor60minand120min(Fig.5c-1and5d-1).Inaddition,itwashardlyobservedcracksonthesurfacefilmofcorrosionproductsafterimmersingfor10and30min.Nevertheless,somecrackscanbeseenbetweensec-ondaryphaseandthematrixafterimmersingfor60min(Fig.5c-2).Underhighmagnifications(Fig5a-3,5b-3and5c-3),ahoneycombcorrosionfilmwithcorrosionproductsisdetected.Moreover,acor-rosionfilmfirstlyformsonthesurfaceofMgmatrix(Fig.5b-3).Withincreasingimmersiontime,thecorrosionfilmgrowstowardsanoblesecondaryphase.Somecracksof,1mminwidthareobservedintheareasbetween18R-LPSOphaseandmatrix(whitearrows,Fig.5c-2).

Nevertheless,itisworthytonotethatthecracksarerecoveredafterimmersingfor120min(Fig.5d-2and5d-3),suggestingthattheremediationofoxidationfilmoccurs25.AnotherelucidationisthattheMDZ-Calloytakesabout120mintoformacompactoxidationfilm26.

AsimilarcorrosionprocessisalsoobservedintheMDZ-545sample.Theprecipitatesbecomevaguewhentheimmersiontimeis60min(Fig.6).Underhighermagnifications(Fig.6a-3,6b-3and6c-3),ahoneycombconfigurationfilmisobservedontheentiresurface.Astheimmersiontimeisextended,theoxidationfilmbecomesmoreandmorecompact.UnlikethediscontinuousoxidefilmonsurfaceoftheMDZ-Calloy,thecorrosionfilmformsonthesurfaceofbothMgmatrixandthe14H-LPSOphaseconcurrently(Fig.6a-3).

Figure4|OCPcurvesofMDZalloysunderdifferentstatesin0.9wt.%NaClsolution.

镁合金,LPSO

Figure5|TheinitiationanddevelopmentofthecorrosionproductsformedontheMDZ-Calloy,(a)afterimmersion10min;(b)30min,(c)60minand(d)120

min.

Cyclicpolarization.Acyclicpolarizationscanprovidesaqualitativeviewofpittingcorrosionmechanismanddeterminesthetrendofundergoingsurfacepittingwhenamaterialisplacedinacorrosivemedium27.ItisparticularlysuitableforMgalloyscontainingLPSOphases,inwhichpittingcorrosionoccursfrequently.ThecyclicpolarizationcurvesoftheMDZ-CandMDZ-545alloysafterdifferenttimeswereshowninFig.7.Thepolarizationcurveswereusedtoestimatethecorrosionpotential(Ecorr),pittingpotential(Epp),andcorrosioncurrentdensity(Icorr)bytheTafel

extrapolationofthecathodicbranches26.Theaveragecorrosionrate(ACR,mmy21)iscalculatedbasedonfollowingequation:

ACR~

IcorrKEw

d

ð1Þ

WhereKisconstantthatdefinestheunitsofthecorrosionrate(,3272mm/A-cm-y).

Ewisequivalentweight(,12g/equivalent).disdensity(,1.7g/cm3

).

Figure6|TheinitiationanddevelopmentofthecorrosionproductsformedontheMDZ-545alloy,(a)afterimmersion10min;(b)30minand(c)60min.

镁合金,LPSO

Figure8|Thesurfacepittingmorphologiesaftercyclicpolarizationtest(10min),(a)theMDZ-Calloy;(b)theMDZ-545alloy.

Figure7|CyclicpolarizationcurvesoftheMDZ-CandMDZ-545alloysatascanrateis2.5mV/sin0.9wt.%NaClsolutionafterimmersingfor10min,2hand4h,respectively.

EISmeasurement.Fig.9ashowstheNyquistspots,measuredatcorrosionpotential,ontheMDZ-Calloyimmersedin0.9wt.%NaClaqueoussolutionsfordifferenttimes.AllEISdiagramsexhi-bitthesameconfigurations.TheBodespots(Fig.9b)demonstratesthatthesamplepresentsahighfrequencyresistivebehaviorfollowedbyahigh/medium(100–10Hz)capacitiveresponse.Additionally,adistortedinductanceloopisobservedatmedium/lowfrequencies(,1–0.1Hz).Thediameterofcapacitiveloopisreducedastheimmersiontimeisincreased.Bycomparison,itisseenfromtheNyquistspots(Fig.9c)oftheMDZ-545alloythattheEIScurvesof1hand4harecomposedoftwoloopsandtheothercurvescontainmerelyaloop.IncontrasttotheMDZ-Csamples,twodifferentcha-racteristicsareobserved,whereinthediameterofcapacitiveloopisincreasedwithincreasingimmersingtimeandtheinductanceloopishardlydetected.

TheSEMgraphsafterEIStestsareshowninFig.10,whereitrevealsthatbothoxidationfilmsarerupturedduringtheEIStest(Fig.10a-1and10b-1).However,underhighmagnificationsoftheMDZ-C(Fig.10a-2and10a-3),thenakedmicrostructureisiden-tifiedinthebottomofcracks.Conversely,anewhoneycomboxida-tionfilmisobservedintheMDZ-545sample,suggestingthattheabilitytoformanoxidationfilmoftheMDZ-545alloyisstrongerthanthatoftheMDZ-Coneaftertheruptureofoxidationfilm.Corrosionproducts.TheXRDresults(Fig.11a)revealthatthephasesintheun-immersedMDZ-CalloyscontainMgmatrixand18R-LPSOphase,whichisconsistentwiththeaboveSEMandTEMresults.ThelowvolumefractionandlimitationofXRDtechniqueisresponsiblefortheabsenceofeutecticphase.Somenewpeaksareobserved,whicharewellassignedtothefollowingcrystallinephase:Mg(OH)2,Dy(OH)3andMgCl2compounds.IntheMDZ-545alloy(Fig.11b),thepeaksareidentifiedasMgmatrixand14H-LPSOphase,whicharepresentinboththeun-immersedsampleandintheoneimmersedfor12h.Nevertheless,comparedwiththeMDZ-Csample,theMgCl2peaksarehardlydetectedintheMDZ-545sample,whileMg(OH)2andDy(OH)3compoundsarestilldetectable.

ThedetailedresultsaresummarizedinTable1.Basically,twoEcorrvaluesshiftforwardpositivelywithincreasingimmersiontime.FortheMDZ-Calloy,theEPPshowsthesametrendasEcorr.However,aconstantEPPof21303mVisobservedintheMDZ-545alloy.TheIcorroftheMDZ-Calloyisincreasedwithextendingimmersiontime,resultinginalargeACRvalue(24mmy21).However,theoppositetrendisachievedintheMDZ-545alloy.Thoughthebackwardscanbreakstheoxidationfilmlayerdown,therelativelystableIcorrandACRvaluesareobserved.TherepresentativeSEMmicrographsoftheMDZ-CandMDZ-545alloysaftercyclicpolarization(10min)areshowninFig.8.InregardstotheMDZ-Calloy,aporousbandalongthegrainboundariesisobserved,whichindicatesthattheoxidationfilmisabsolutelybrokendownduringtheEIStesting(alsoconfirmedinFig.10).Conversely,onlysomeisolatedpittingspotsaredetectedintheMDZ-545alloyaftercyclicpolarizationtest.

Discussion

Inpresentstudy,theprecipitatesaremostlydistributedingrainboundariesintheMDZ-Calloy,whichiscomposedofa-Mgand18R-LPSOphase.18R-LPSOphasedissolvesgraduallyand

14H-

Table1|Electrochemicalparametersofthesamplesderivedfrompolarizationtestsin0.9wt.%NaCl

SampleMDZ-CMDZ-545

Immersedtime10min2h4h10min2h4h

Ecorr(mV)216002154621507216052157121516

Icorr(mAcm22)0.10110.13081.04210.09510.08620.0461

EPP(mV)214532131021213213032130321303

ACR(mmy21)2.33443.020224.0622.19511.99041.0645

镁合金,LPSO

Figure9|Impedanceplotsofalloysimmersedin0.9wt.%NaClsolutionfordifferentimmersiontimes,(a)NyquistdiagramoftheMDZ-Csample;(b)BodeplotsoftheMDZ-Csample;(c)NyquistdiagramoftheMDZ-545sample;(d)BodeplotsoftheMDZ-545sample.

Figure10|SEMmicrographsoftheMDZ-C(a-1,a-2anda-3)andMDZ-545(b-1,b-2andb-3)samplesafterEIS

test.

镁合金,LPSO

ThisexplanationisconsistentwiththeOCPresultsthattheMDZ-Calloytakeslongertimetoreachingbalancepotentialanditspoten-tialfluctuatessignificantlyduetotheruptureofoxidationfilm.Ontheotherhand,theoxidationfilmremediationabilityalsoplaysanimportantroleincorrosionproperties.Itiswell-knownthattheoxidefilmonthesurfaceofMgalloysisnotcompact,andthenthesolutioncanpenetratethroughtheoxidationlayer.Takingintoaccountthereleaseofhydrogenduringcorrosionprocess,theoxidationfilmwillbebrokendownwithextendingimmersiontime.ThisisthereasonforthecontinuousdecompositionofMgalloys25,28.Therefore,inviewoftheformationfreshsurface,theoxidationfilmremediationabilityiscrucialtodeterminecorrosionrate.BasedonEISresults(Fig.9),thecapacitiveresistanceisreducedwithincreas-ingimmersiontestintheMDZ-Calloy,whileitisenhancedintheMDZ-545onemonotonously.Themainreasonliesinthedifferentoxidationfilmremediationability(Fig.10).AfterEIStesting,thepristineoxidationfilmisbrokendownandtheoxidationfilmreme-diationoccursimmediatelyintheMDZ-545alloy(Fig.10d-10f).Anewcompactoxidationfilmformsinthecracks,whicheffectivelyprohibitsthecontactbetweensolutionmediaandthematrix.Itisconsistentwiththeimmersiontest,whereMDZ-Ctakesmoretime(Fig.5d)toformacompleteoxidationfilm.Thisstrategyofimprov-ingcorrosionpropertiesbyformationofseparatedlayerwasemployedinMg-AlalloybySonget.al.28,29.

Theoretically,thesurfacefilmonMgalloysismainlycomposedofaqueoussolutionmagnesiumoxideormagnesiumhydroxide.Thecorrosionreactionsinneutralaqueousmediaarepresentedbythefollowingequations30–33.

Mg?Mg2zz2e{2H2Oz2e{?H2z2OH{

Figure11|XRDpatternsofalloysbeforeandafter12hofimmersionin0.9wt.%NaClsolution,(a)MDZ-C;(b)MDZ-545.

ð2Þð3Þð4Þð5Þ

Mg2zz2OH{<Mg(OH)2Dy3zz3OH{?Dy(OH)3

LPSOphasegenerateswhenitissolid-solutionheattreatedat545uC.Bothmorphologyanddistributionof14H-LPSOphasearedifferentfromthoseof18R-LPSOphase.Thewidthof14H-LPSOphaseisreducedanditismainlydistributedinsidegrains.TheseoccurrencesareconfirmedbyDSCresult,togetherwiththeSEMandTEMresults.Itisbelievedthatthe18R-typestructureisnotthermodyna-micallystableat/above500uC.Itisbecausethatitcontainsfoura-Mgatomicfaultlayerswhichprovideasimpletransitionalpathwaytoform14H-typeLPSO13,whichisconsistentwiththethermalcalcula-tionsofLPSOphasetransformations7.

AccordingtotheaboveSEMandEISresults,itcanbeconfirmedthatthecorrosionmechanismsaredifferentbetweentheMDZ-C(18R-LPSO)andMDZ-545(14H-LPSO)alloys.Specifically,thedetailedschematicdiagramsforcorrosionprocessexplanationwerepresentedinFig.12.IncontrasttotheMDZ-C(18R-LPSO)alloy,theMDZ-545(14H-LPSO)onepossesseslowercorrosionrate.Therearetwomainfactorsresponsibleforthedecreaseofcorrosionrate.Ontheonehand,differentfromtheMDZ-Csample,itisfaciletoformahomogeneousandcompactoxidationfilmattheinitialcorrosionstageintheMDZ-545one(Fig.5andFig.6),whichisascribedtodifferentformingsequencesofoxidationfilms.FortheMDZ-Calloy,theoxidefilmgeneratesonthesurfaceofMgmatrixfirstandthen

expandstowardthesecondaryphase.Thus,thethicknessoffilmisdifferent,whichresultsintensilestress27.Withincreasingtheimmer-siontime,thetensilestressisincreased,resultingintheruptureoftheoxidationfilmcontainingcorrosionproductsintheinterfacebetween18R-LPSOphaseandMgmatrix.Asaresult,theformedcracksprovidechannelsforthecorrosivesolutiontopenetratethroughandreachthefreshsurface,aswellasforhydrogentorelease.Consequently,thecorrosionprocessoccurssequentially.

However,thefollowingreactionpossiblyoccurswhentheenvir-onmentcontainschlorideion:

Mg2zz2H2Oz2Cl{?2Mg(OH)2Cl2?MgCl2z2H2Oð6ÞAccordingtocorrosionproducts,itcanbeconfirmedthatequation5ismorefaciletoproceedintheMDZ-CalloythanintheMDZ-545one.Chlorideionsareveryaggressivetomagnesium.ChlorideionsintheinterfacetransformMg(OH)2toeasilysolubleMgCl2.Asaresult,thethermodynamicalequilibrium(equation3)isreadilybro-kendownduetotheexistenceofchlorideions,whichdeterioratesthecorrosionpropertiesofthealloys5.ThiscorrosionexplanationagreeswellwiththecurrentopinionthatthecorrosionresistanceofMgalloysisdeterminedbyapartiallyprotectivesurfacefilmwherethecorrosionreactionoccursprincipallyatthebreaksofthefilm.Theseresultsarealsoconsistentwiththetendencyofchlorideioncausingfilmbreakdown,whichacceleratesthecorrosionofMgalloy34.However,thedirectevidencethatthechlorideioniseasiertobeabsorbedonthesurfaceofalloycontaining18R-LPSOphaseisnotdemonstratedinthisworkandfurtherresearchworksarenecessarytoclarifytheirdiscrepancies.

Insummary,paredwiththealloycontaining14H-LPSOphase,itispronetoformsomecracksintheinterfacesbetween18R-LPSOandMgmatrixatthebeginningofcorrosion,whichismostlyrelatedtotheinhomogeneous

镁合金,LPSO

Figure12|Thedetailedschematicdiagramsofcorrosionprocess,a1-a4graphscorrespondtotheMDZ-Calloy;b1-b4graphscorrespondtotheMDZ-545alloy.Itrevealsthatbothrapidformfilmabilityandremediationabilityaretwoimportantreasonsinenhancingthecorrosionproperties.

distributionoftensilestress.Inaddition,asdegradationprocessispreceded,bothsurfaceoxidationfilmsarebrokendown.However,incontrasttoas-castalloycontaining18R-LPSOphase,thealloywith14H-LPSOphaseexhibitsrapideroxidationfilmremediationability,whicheffectivelyinhibitsthecontactbetweensolutionandthematrix.

Methods

Samplepreparation.AnormalcompositionMg-2Dy-0.5Zn(at.%)alloyhasbeenpreparedbychillcastingtechnology.Thealloyhadbeenpreparedinatantalum-crucibleunderacovergasmixtureofCO2andSF6.Aftermixingat720uCfor1h,thealloywascasttothemouldpreheatedat600uC.Thefilledmouldwasheldat670uCfor1hundertheprotectivegas.Then,thewholetantalumcruciblewiththemeltingalloywasimmersedtotheflowingwaterat600ml/s.Whenthebottomoftantalumcrucibletouchedthewater,itstoppedfor2second.Assoonastheliquidleveloftheinsidemeltwasinalignmentwiththeheightofoutsidewater,thesolidificationprocesswasfinished.Thediameterandlengthoftheingotwere70mmand80mm,respectively.ThechemicalcompositionofMg-1.91Dy-0.32Zn(indexedasMDZ-C)wasmeasuredbyX-rayfluorescencespectroscopy.Theelementcompositionwaslowerthanthenominalcompositionowingtotheelementburningloss.Moreover,thetotalcontentofmainimpurities(Cu,FeandNi)islowerthan0.015wt.%.Microstructureandphaseidentification.Themicrostructuralinvestigationswereperformedusingscanningelectronmicroscopywithafieldemissiongun(SEM-FEG,JEOLJSM-7001F).Thestandardmetallographicprocedureswereapplied,includinggrinding,polishingandetching.Thesampleswereetchedinapicralsolutiontorevealgrainboundaries.Theelementalconcentrationinthematrixandphaseswere

investigatedbySEMequippedwithenergydispersiveX-rayanalysis(EDX)(INCAfromOxford).Theaveragevalueswereobtainedbasedonatleastfiverandomspots.Calorimetricresponseoftheas-castalloywasmeasuredusingdifferentialscanningcalorimetry(DSC,STA449C).Aheatingrateof6uC/minwasemployedunderargonpurgeat35ml/min.Thephaseshavebeenidentifiedbytransmissionelectronmicroscopy(TEM,JEM-2010).ThinfoilsamplesfortheTEMobservationwerepreparedusinganIonPolishingSystem(RES101).

Electrochemicaltest.Electrochemicalbehaviorsweretestedusingapotentiostat/frequencyresponseanalysissystem(Bio-logic,VSP).Experimentswerecarriedoutinathree-electrodeelectrochemicalcell,inwhichasaturatedcalomelelectrode(SCE)asthereferenceelectrode,aplatinummeshascounterelectrodeandthespecimenunderinvestigationastheworkingelectrode.Theexperimentswerecarriedoutin0.9wt.%NaClaqueoussolutions.InordertoreducethefluctuationofpH,theratioofsolutiontosurfaceis80mlto1mm2.Duringthewholetest,thetrayswereplacedinanincubator.Thetemperatureintheincubatorwasmaintainedat25uC.Opencircuitpotential(OCP)measurementsweremadebetweentheworkingelectrodeandthereferenceelectrodewithoutcurrentbeingpassedtothecounterelectrodes.This

measurementshowedthepotentialatwhichtheanodicandcathodicreactioncurrentsattheworkingelectrode/solutioninterfacewerebalanced.TheOCPtestswerebegunimmediatelyafterthespecimenwasimmersedinthesolutionandweremeasuredfor6hduration.

Thecyclicpolarizationswereperformedafterimmersingfor10min,2hand4h.Theforwardscannedpolarizationcurvesstartedfromacathodicpotentialof

21800mV,wherethesurfacesofthealloyswerenotyetcorroded,andextendedtothevertexpotentialof250mV.Theforwardscanwasfollowedbyareversescanbacktothefinalpotential21950mV.Thescanratewas2.5mV/s,thestepheightwas1mV.Thesurfacemorphologyaftercyclicpolarizationtestafterimmersingfor10minwasdriedandobservedbySEMimmediately.

Theelectrochemicalimpedancespectroscope(EIS)wascarriedoutatopen

potentialwithanappliedsignalof10mVrms.Thescanningfrequencyrangedfrom100kHzdownto0.1Hz.Thesampleswereimmersedin0.9wt.%NaClsolutionsfordifferenttimes,viz.2h,4h,6h,8hand12h,toinvestigatethecorrosionprocess.AfterEIStests,theMDZ-CandMDZ-545samplesweredriedandobservedbySEMimmediately.

Immersiontestandproductanalysis.Theimmersiontestsweretoinvestigatethecorrosionmechanismandtorevealthedevelopmentandgrowthofcorrosion

productsofMDZalloysinNaClaqueoussolutions.Tothispurpose,threecylinder(Ø1035)specimenscontaining18R-LPSOor14HLPSOphaseswereimmersioninNaClaqueoussolutionsatdifferenttimesviz.10min,20min,60minand120min.Theywerecleanedbyrinsingwithpurewater,andthenrinsedbyethanolanddriedinairpriortoSEM-FEGobservation.

Thephasecompositionofcorrosionproductsonthesamplesimmersedupto12hinthetestsolutionswereconfirmedbyx-rayelectrondiffractiondirectly(XRD,Rigaku).TheXRDanalysiswasperformedbyscanningfrom20uto80uwithastepsizeof0.02uwithCuKaradiation.Thefilamentcurrentandaccelerationvoltagewere30mAand40kV,respectively.

1.Staiger,M.P.,Pietak,A.M.,Huadmai,J.&Dias,G.Magnesiumanditsalloysasorthopedicbiomaterials:Areview.Biomaterials.27,1728–1734(2006).

2.Serre,C.M.,Papillard,M.,Chavassieux,P.&Boivin,G.Invitroinductionofacalcifyingmatrixbybiomaterialsconstitutedofcollagenand/orhydroxyapatite:anultrastructuralcomparisonofthreetypesofbiomaterials.Biomaterials.14,97–106(1993).

3.Janning,C.etal.Magnesiumhydroxidetemporarilyenhancingosteoblastactivityanddecreasingtheosteoclastnumberinperi-implantboneremodelling.ActaBiomater.6,1861–1868(2010).4.Zberg,B.,Uggowitzer,P.J.&Lo¨ffler,J.F.MgZnCaglasseswithoutclinicallyobservablehydrogenevolutionforbiodegradableimplants.NatureMater.8,887–891(2009).

5.Yamasaki,M.,Hayashi,N.,Izumi,S.&Kawamura,Y.CorrosionbehaviorofrapidlysolidifiedMg-Zn-REelementalloysinNaClsolution.Corr.Sci.49,255–262

(2007).

镁合金,LPSO

6.Yokobayashi,H.etal.EnrichmentofGdandAlatomsinthequadrupleclosepackedplanesandtheirin-planelong-rangeorderinginthelongperiodstacking-orderedphaseintheMg-Al-Gdsystem.ActaMater.59,7287–7299(2011).7.Zhu,Y.M.,Morton,A.J.&Nie,J.F.The18Rand14Hlong-periodstackingorderedstructuresinMg-Y-Znalloys.ActaMater.58,2936–2947(2010).

8.Suzuki,M.,Kimura,T.,Koike,J.&Maruyama,K.StrengtheningeffectofZninheatresistantMg-Y-Znsolidsolutionalloys.Scr.Mater.48,997–1002(2003).9.Kentaro,Y.etal.AlloydevelopmentofhightoughnessMg-Gd-Y-Zn-Zralloys.Mater.Trans.47,1066–1070(2006).

´s,G.etal.EffectofmicrostructureoncreepbehaviorofcastMg97Y2Zn110.Garce

(at.%)alloy.Mater.Sci.Eng.A539,48–55(2012).

11.Zhang,J.etal.Researchonlong-period-stacking-orderedphaseinMg-Zn-Y-Zralloy.J.AlloysCompd.558,195–202(2013).

12.Shao,X.H.,Yang,Z.Q.&Ma,X.L.StrengtheningandtougheningmechanismsinMg-Y-Znalloywithalongperiodstackingorderedstructure.ActaMater.58,4760–4771(2010).

13.Zhu,Y.M.,Morton,A.J.&Nie,J.F.Growthandtransformationmechanismsof18Rand14HinMg-Y-Znalloys.ActaMater.60,6562–6572(2012).

14.Tong,L.B.,Li,X.H.&Zhang,H.J.Effectoflongperiodstackingorderedphaseonthemicrostructure,textureandmechanicalpropertiesofextrudedMg-Y-Znalloy.Mater.Sci.Eng.A563,177–183(2013).

15.Ambat,R.,Aung,N.N.&Zhou,W.EvaluationofmicrostructuraleffectsoncorrosionbehaviourofAZ91Dmagnesiumalloy.Corr.Sci.42,1433–1455(2000).16.Zhang,X.etal.BiocorrosionbehaviorandcytotoxicityofaMg-Zn-Yalloywithlongperiodstackingorderedstructure.Mater.Lett.86,42–45(2012).

17.Zhao,X.,Shi,L.I.&Xu,J.Mg-Zn-Yalloyswithlong-periodstackingorderedstructure:Invitroassessmentsofbiodegradationbehavior.Mater.Sci.Eng.C.33,3627–3637(2013).

18.Izumi,S.,Yamasaki,M.&Kawamura,Y.RelationbetweencorrosionbehaviorandmicrostructureofMg-Zn-Yalloyspreparedbyrapidsolidificationatvariouscoolingrates.Corr.Sci.51,395–402(2009).

19.Yin,D.D.etal.EffectsofheattreatmentsonmicrostructureandmechanicalpropertiesofMg-11Y-5Gd-2Zn-0.5Zr(wt.%)alloy.J.AlloysCompd.509,1696–1704(2011).

20.Avedesian,M.M.&Baker,H.ASMSpecialtyHandbook:MagnesiumandMagnesiumAlloys.MaterialsPark,OH:ASMInternational,1999.

21.Bi,G.etal.AnelevatedtemperatureMg-Dy-Znalloywithlongperiodstackingorderedphasebyextrusion.Mater.Sci.Eng.A.528,3609–3614(2011).

22.Abe,E.,Kawamura,Y.,Hayashi,K.&Inoue,A.Long-periodorderedstructureinahigh-strengthnanocrystallineMg-1at%Zn-2at%Yalloystudiedbyatomic-resolutionZ-contrastSTEM.ActaMater.50,3845–3857(2002).

23.Hagihara,K.,Yokotani,N.&Umakoshi,Y.PlasticdeformationbehaviorofMg12YZnwith18Rlong-periodstackingorderedstructure.Intermetallics.18,267–276(2010).

24.Atrens,A.etal.Stresscorrosioncrackingandhydrogendiffusioninmagnesium.Adv.Eng.Mater.8,749–751(2006).

25.Song,G.&Atrens,A.Understandingmagnesiumcorrosion-aframeworkforimprovedalloyperformance.Adv.Eng.Mater.5,837–858(2003).

26.Song,G.etal.Theelectrochemicalcorrosionofpuremagnesiumin1NNaCl.Corr.Sci.39,855–875(1997).

27.Guo,X.etal.InvestigationofcorrosionbehaviorsofMg-6Gd-3Y-0.4ZralloyinNaClaqueoussolutions.Electrochim.Acta.52,2570–2579(2007).

28.Song,G.&StJohn,D.H.Corrosionofmagnesiumalloysincommercialenginecoolants.Mater.Corr.56,15–23(2005).

29.Song,Y.L.etal.EffectofceriumadditiononmicrostructureandcorrosionresistanceofdiecastAZ91magnesiumalloy.Mater.Corr.58,189–192(2007).30.Liu,M.etal.Theinfluenceofyttrium(Y)onthecorrosionofMg-Ybinaryalloys.Corr.Sci.52,3687–3701(2010).

31.Pardo,A.etal.Corrosionbehaviourofmagnesium/aluminiumalloysin3.5wt.%NaCl.Corr.Sci.50,823–834(2008).

32.Sudholz,A.D.etal.ElectrochemicalbehaviourandcorrosionofMg-Yalloys.Corr.Sci.53,2277–2282(2011).

33.Zhao,M.C.,Liu,M.,Song,G.L.&Atrens,A.InfluenceofpHandchlorideionconcentrationonthecorrosionofMgalloyZE41.Corr.Sci.50,3168–3178(2008).34.Dhanapal,A.,Rajendra,S.&Balasubramanian,V.InfluenceofpHvalue,chlorideionconcentrationandimmersiontimeoncorrosionrateoffrictionstirweldedAZ61Amagnesiumalloyweldments.J.AlloysCompd.523,49–60(2012).

Acknowledgments

ThisresearchissupportedbyNSFC(51101142,50821001and51102206),NewCenturyExcellentTalentsinUniversityofMinistryofEducationofChina(NCET-12-0690),ScienceFoundationfortheExcellentYouthScholarsofHebeiProvince(Y2012019)andScienceSupportingProjectofHebeiProvince(13961002D).

Authorcontributions

Q.P.designedtheexperimentsandcontributedtotheexperimentalanalysisand

interpretationofdata,andwrotethemanuscript.J.G.contributedtotheexperimentsandanalysis.X.C.contributedtotheexperimentsandanalysis.Y.W.contributedtothe

experimentsandanalysis.B.L.contributedtotheinterpretationoftheexperimentaldataandcontributedtothewritingofthemanuscript.Z.X.contributedtotheinterpretationoftheexperimentaldataandcontributedtothewritingofthemanuscript.Allauthorsreviewedthemanuscript.

Additionalinformation

Competingfinancialinterests:Theauthorsdeclarenocompetingfinancialinterests.Howtocitethisarticle:Peng,Q.M.etal.DegradationbehaviorofMg-basedbiomaterialscontainingdifferentlong-periodstackingorderedphases.Sci.Rep.4,3620;DOI:10.1038/srep03620(2014).

ThisworkislicensedunderaCreativeCommonsAttribution-NonCommercial-ShareAlike3.0Unportedlicense.Toviewacopyofthislicense,

visit

/licenses/by-nc-sa/3.0

本文来源：https://www.bwwdw.com/article/hr74.html