Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are encouraged to visit:
http://www.elsevier.com/copyright
ContentslistsavailableatSciVerseScienceDirect
Materials Science and Engineering A
j o u r n al hom ep ag e :w w w . e l s e v i e r . c o m / l o c a t e / m s e a
Precipitation microstructure and their strengthening effects of an Mg–2.8Nd–0.6Zn–0.4Zr alloy with a 0.2 wt.% Y addition
J.H. Li
a,b,1, G. Sha
a,c,∗, W.Q. Jie
b, S.P. Ringer
a,caAustralianCentreforMicroscopyandMicroanalysis,TheUniversityofSydney,MadsenBuildingF09,Sydney,NSW2006,Australia
bStateKeyLaboratoryofSolidificationProcessing,NorthwesternPolytechnicalUniversity,Xi’an710072,China
cARCCentreofExcellenceforDesigninLightMetals,TheUniversityofSydney,NSW2006,Australia
a r t i c l e i n f o
Articlehistory:
Received3October2011
Receivedinrevisedform10January2012 Accepted11January2012
Available online 20 January 2012
Keywords:
Magnesiumalloy Precipitation Mechanicalproperties TEM
Atomprobetomography
a b s t r a c t
Amicro-alloyadditionofY(0.2wt.%)producedasignificantimprovementinthetensileyieldstrengthof anMg–2.8Nd–0.6Zn–0.4Zr(wt.%)alloyat200◦C,witha28%increasefrom143MPato183MPa.APT characterisationconfirmedthatsmallsoluteclustersinahighnumberdensitywerepresentinthe as-quenchedsample.TEMexaminationsrevealedthat-typeprecipitateshabitingonprism{1120}
planesweredominantinthesampleafteragedfor14hat200◦C,andtheyco-existedwith␥-typepre- cipitatesonthebasalplaneofMgmatrix.Theprecipitationsequenceof-typeprecipitatesissolute clusters→→→1beforethepeakhardness.TherewassignificantclusteringofsolutesintheY- containingalloy,butYdidnotclearlypartitionintoclustersorprecipitatesandremainedintheMgmatrix.
,and1weremeasuredwithstoichiometricMg9(Nd,Zn),Mg4(Nd,Zn)andMg2(Nd,Zn),respectively.
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
Magnesiumalloyshaveimportantapplicationsintheautomo- tiveandaerospaceindustriesbecauseoftheirhighspecificstrength fortheweightreductionandbetterfueleconomy[1].However, themechanicalpropertiesofconventionalMgalloysareoftennot suitableforhigh-temperatureapplications.Thelowermechanical properties,particularlythelowertensileyieldstrengthatelevated temperatures,hinderthewiderapplicationofMgalloys[2].The developmentofMg alloys forthehigh-temperatureapplication remainstobeachallengingresearchtargetfortheinternational materialscommunity.
Rare-earth(RE) elementaddition hasbeenconsideredto be effectivetopromoteprecipitationhardeningandtoimprovethe high-temperatureperformanceofMgalloysduetotheformation ofthermallystablenano-sizedprecipitatesinahighnumberden- sity[3–18].Indeed,somecommercialhigh-strengthheat-resistant Mgalloys(i.e.WE54/43,QE22,etc.)containatleasttwotypesof REelements(oneismainandtheotherisminor).Frequently,RE
∗Correspondingauthorat:AustralianCentreforMicroscopyandMicroanalysis, TheUniversityofSydney,MadsenBuildingF09,Sydney,NSW2006,Australia.
Tel.:+61290369050;fax:+61293517682.
E-mailaddress:gang.sha@sydney.edu.au(G.Sha).
1 Presentaddress:ChairofCastingResearch,TheUniversityofLeoben,A-8700 Leoben,Austria.
elementsareselectedfromNd,Y,sometimesScandGd,etc.[2]for furtherimprovingthehigh-temperatureperformanceofMgalloys.
MgalloyswithanNdadditionexhibitastrongage-hardening response and improvement in mechanical properties. The pre- cipitationof, andphasesandthemechanicalproperties ofsomeMg–Ndbasedalloyshavebeenextensivelyinvestigated [3–12].However,thehigh-temperaturemechanicalpropertiesof theseMg–Ndbasedalloysareoftennotidealforapplicationsabove 250◦C.Yisanimportantalloyingelementtoeffectivelyimprove themechanicalpropertiesofMgalloysatelevatedtemperatures [13–18].SeveralimportantcommercialMgalloys,suchasWE54 andWE43,weredevelopedhavingacombinedhigh-leveladdition ofYandNd[13–15].TheprecipitationphasesintheseMgalloysare
,,1and,proposedwithstoichiometryofMg3RE,Mg5RE, Mg3REandMg3RE,respectively[14,16].Theyprovidestrengthen- ingeffectstothesehigh-Y-containingMgalloys[13–17].
The addition of Y in high quantities, even for WE54 (Y:
4.75–5.5wt.%) and WE43 (Y:3.7–4.3wt.%)alloy, hasthedraw- backstoincreasealloydensityandproductioncost.Consequently, alowerleveladditionofYbecomesmoreattractiveifitcanpro- duceeffectiveimprovementsinthemechanicalpropertiesofMg alloys. Recently, a micro-alloy additionof only 1–2wt.% Yinto Mg–3Nd–0.5Zn–0.4Zr(wt.%)alloyhasbeenreportedtoremark- ably increase itstensile properties and creepresistance[18].It isgenerallybelievedthattheimprovementofthealloystrength is correlatedwithanenhancedprecipitationbytheaddition of Y. To date, there is a lack of detailed information about the 0921-5093/$–seefrontmatter© 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.msea.2012.01.043
evolutionofprecipitatesmicrostructure,and thepartitioningof solutesintoprecipitates during ageinglow-Y-containing alloys.
Better understanding precipitates microstructure formation is of importance for establishing the relationship between the microstructureandmechanicalpropertiesofthelow-Y-containing Mg–Ndbasedalloys.
Thispaperaimstoreportthemechanicalpropertiesandpre- cipitatesmicrostructureofanMg–2.8Nd–0.6Zn–0.4Zr(wt.%)alloy witha0.2wt.%Yaddition.Byunveilingcomprehensivestructural andchemicalinformationofsolute-richfeaturesformedinthealloy usingcarefulTEMcharacterisationsandquantitativeatomprobe dataanalyses,this investigation hasobjectivestoelucidate the precipitationsequence,toaddressthesolutepartitioningbehav- iorsduringageing,andtounderstandthestrengtheningeffectand mechanismsofthelow-Y-containingMg–Ndbasedalloy.
2. Experimentalmaterialandprocedures
Mg–2.8Nd–0.2Y–0.6Zn–0.4Zr alloy (wt.%) wasprepared with high purity Mg (99.9%), Zn (99.9%), Nd (99.9%), Mg–28Y and Mg–33Zrmasteralloysinanelectricresistancefurnaceunderthe protectionofananti-oxidizingflux(containing54–56%KCl,14–16%
BaCl2,1.5–2.5%MgO,27–29%CaCl2),andcastedintoasandmould.
Thechemicalcompositionsoftheexperimentalalloyweredeter- minedbyusinganinductively coupledplasma atomicemission spectrum(ICP-AES) apparatus.The solutiontreatmentof speci- menscutfromthealloyingotwasconductedat525◦Cfor18hina saltbath.Theageingofwater-quenchedspecimenswasperformed inanoilbathforvariousageingtimeat200◦C.TheVickershardness testingwasundertakenonLECOHardnessTester(LV700AT)with 1kgloadand10sdwellingtime.Eachdatapointreportedinthis paperrepresentedanaverageofatleast10measurements.Theten- siletestswereperformedusingstandardtensiletestingmachine (Instron1195)atroomtemperature(RT),200◦C,250◦C,300◦Cand 350◦C, respectively,witha crossheadspeedof5mm/minand a strainrateof2.0×10−3s−1.A5minholdingwasappliedtoeach sampletobalanceitstemperaturebeforeeachhigh-temperature tensiletest.Eachdatapointreportedinthispaperwasanaverage ofatleast3testsamples.
TEMfoilspecimenswerepreparedbytwin-jetelectro-polishing inasolutionof25%HClO4 and75%methanolat−40◦Cand20V, andthenusinglow-energyionbeamthinningforsurfacecleaning.
TheTEMobservationswereperformedinaPhilipsCM12operat- ingat120kVandahighresolutionTEM(JEOL-3000F)operating at300kV.Thesamplesforatomprobeanalysiswerepreparedby two-stageelectro-polishingofblankswithasizeofapproximately 0.5mm×0.5mm×15mm.Thefirststepwasconductedusingan electrolyteof25%perchloric acidinacetic acidat 15Vatroom temperatureandthesecondstepwasin2%perchloricacidin2- butoxyethanolat20V.Atomprobeanalyseswereperformedusing aImagoLEAPTM3000SI,operatingataspecimentemperatureof 20K,20%voltagepulsefractionandunderultrahighvacuumcon- ditions(about1.8×10−11Pa).
Atom probe data sets werecarefully reconstructedusingan approachoutlinedrecentlybyMoodyetal.[19].Themaximum separationalgorithmwasemployedtoidentifysolute-richfeatures [20,21],inwhichNdandZnwereselectedasclusteringsolutes,a solute-separationdistanceof0.8nmandtheminimumsizeof15 soluteatomswereusedintheidentificationofsolute-richfeatures, suchasclustersandprecipitates.Solute-richfeaturescontaining
<15soluteswereneglectedintheanalysisbecausesuchsmall-size featureshighlyexistedinthevolumewithsolutesinrandomdistri- bution[11,12].Inordertoestimatetheprecisecompositionoflarge precipitates,selection-boxanalysiswasemployedtomeasurethe compositionfromthecoreregionofeachprecipitateinorderto
Fig. 1.Age-hardening responses of Mg–2.8Nd–0.2Y–0.6Zn–0.4Zr (wt.%) and Mg–2.8Nd–0.6Zn–0.4Zralloysagedat200◦C.
reducetheeffectoftrajectoryoverlapbetweenprecipitateandthe Mgmatrixonthemeasurement.
3. Results
3.1. Agehardeningresponseandtensilepropertiesof Mg–Nd–Zn–Zrbasedalloys
Fig.1showstheage-hardeningresponseoftwoMg–Nd–Zn–Zr alloyswith/withouta0.2wt.%Yadditionduringageingat200◦C.
TheMgalloywithanadditionof0.2wt.%Yexhibitedanenhanced strengtheningeffectincomparisonwiththeY-freealloy.Thehard- nessoftheas-quenchedsamplesincreasedfrom43.6±2.0HVto 58.0±2.0HV.Afterageingfor3hat200◦C,thehardnessofboth alloysexhibitedstrongincreasesof∼18.8HVintheY-containing alloyand∼23HVintheY-freealloy.ThepeakhardnessoftheY- containingalloywas79.4±2.4HVafter14hageing.Furtherageing beyond14hledtoprogressivelydecreaseintheirhardness.
Fig.2showsthetensilepropertiesofaY-containingMgalloy andaY-freeMgalloywithdifferentthermalhistory,includingas- cast,as-quenchedafterasolutiontreatment,agedfor3hand14h at200◦C.Thetensilestrengthsofthetwoalloys(Fig.2a)arecor- relatedwellwiththeagehardeningresponsesofthetwoalloys (Fig.1).Anadditionof0.2wt.%Yprovidedaclearimprovement intheyieldstrengths(YS)ofthealloyinallconditions.However, theirultimatetensilestrengths(UTS)andductilitybetweenthetwo alloysshowednosignificantdifference.Theyieldstrengthoftheas- quenchedsamplesexhibitedanimprovementfrom100±6.4MPa to133±7.2MPa(Fig.2a).Withtheincreaseinageingtimefrom 3hto14h,boththeirUTSandYSincreased,buttheirductilityval- uesdecreased,asshowninFig.2a.Importantly,theY-containing alloyinthepeak-agedcondition(T6,14h)exhibitedanimproved YSoverthatoftheY-freealloyinthetemperaturerangefromRTto 350◦C(Fig.2b).Inparticular,theYSoftheY-containingMgalloy at200◦Cis183±7MPa,28%higherthan143±4MPa,thatofthe Y-freealloy.Inordertogaininsightintomicrostructuresproducing theinterestingstrengtheningeffecttotheY-containingalloy,sam- plesindifferentagedconditionswereselectedtoperformfurther TEMexaminationandAPTanalysis.
3.2. TEMcharacterisationofprecipitatesinanMg–Nd–Zn–Zr–Y alloy
Fig. 3 shows a representative bright field (BF) [0001] TEM micrographandthecorrespondingselectedareadiffractionpat- terns(SADP)ofanas-quenchedsample.Noprecipitatesareevident inTEMimage,asshowninFig.3a.ThediffractionsintheSADPwere
Fig.2. TensilepropertiesofMg–2.8Nd–0.6Zn–0.4Zr(wt.%)alloyswithorwithout0.2wt.%Yaddition,(a)indifferentageingconditions,and(b)measuredatdifferent temperatures.
fromthe␣-Mgmatrix,asshowninFig.3b.TheTEMresultsindi- catedthatnoprecipitationtookplaceintheas-quenchedsample.
Afteragedfor14hat200◦C,plate-likeprecipitateswereevidentin TEMimages,asshowninFig.4.Theplate-likeprecipitateshabited on{1120}␣,andabout16±4.3nminlengthand1.5±0.7nmin widthwithanaspectratio(lengthtowidth)ofapproximately10:1, asobserved in[0001]BF image.Theweakdiffraction spotsat 1/2{011 0}␣inthe[0001]SADP,asshowninFig.4b,werefrom
precipitates,withD019structure(a=b=0.64nmandc=0.52nm) [14,17].Thiswasfurtherconfirmedbyweakdiffractionsobserved at 1/2{0110}␣ and 1/2{2114}␣ in the {2110}␣ and [0110]␣ SADPs,asshowninFig.4dandfrespectively.Weakdiffractionsat 1/2(2 ¯1 ¯1 2)␣inthe{0110}␣SADP,asmarkedwithawhitesolid arrow inFig.4f,suggested that precipitateswerepresent in themicrostructureofthealloy.The  phaseisknowntohave abase-centredorthorhombicunitcella=0.640nm,b=2.223nm, c=0.521nm[14,17],andhasalamella-likemorphologywithalon- gitudinalaxisparallelto[0001]␣.
The{2110}␣and[0110]␣TEMBFmicrographs,asshownin Fig.4canderespectively,revealedthatthinprecipitates(marked withawhitesolidarrow)habitingonthebasalplaneoftheMg matrixwerepresentinthemicrostructure,andtheyareperpen- dicularto-typeprecipitates(markedwithablacksolidarrow).
No clearstreaks of thebasal precipitatesobserved in [2110]␣ and[0110]␣SADPs(Fig.4d andf)indicatedthatthebasalpre- cipitateswereprobablyatanextremelylowvolumefraction.The precipitatesonbasal planeof Mg alloyshave beenreported to
be␥-typephase(␥,MgZnREcontaining,hexagonal,a=0.55nm, c=0.52nm[4]).Similar␥-typeprecipitateswereobservedinan Mg–Nd–Gd–Zn–Zralloyagedat330◦Cfor90min[9].
CarefulTEMexaminations,asshown inhigh-resolutionTEM imagesinFig.5,revealedthat1 phasestartedtoappearinthe microstructureofthealloyagedfor14h.The{002},{111}and {220}latticeplanesof aprecipitate, asshownin Fig.5b,were clearly resolvedin thehigh-quality latticefringeimage.The d- spacing of{002},{111}and{220}latticeplanes(Fig.5)were measuredtobe0.36, 0.416and 0.255nm, respectively,in good agreementwiththeidentificationas1phase,withanfccunitcell (a=0.72nm),whichisveryclosetoa=0.74nmreportedbyNieetal.
[14].The1precipitateswereprobablyataverylownumberden- sityinthemicrostructure,becausenoclearreflectionsof1were observedintheSADPsofthealloy,asshowninFig.4.
3.3. APTcharacterisationofprecipitatesinanMg–Nd–Zn–Zr–Y alloy
Fig. 6 shows the APT elemental maps of Nd, Y and Zn obtainedfromanMg–2.8Nd–0.2Y–0.6Zn–0.4Zr(wt.%)alloysam- plequenchedincoldwaterafterasolutiontreatment(for18hat 525◦C).AstrongclusteringofNdwasobservedintheNdatom map, as shown in Fig.6b. Afterremoving solute atoms in the matrix,finesoluteclustersinahighnumberdensity,asshownin Fig.6a,wereevidentintheanalysedvolume.Nootherclearstruc- turalfeature(i.e.dislocation,precipitates,grainboundary[22])was
Fig.3. TEMbrightfieldimagein[0001]␣zoneaxisandcorrespondingSADPofanMg–2.8Nd–0.2Y–0.6Zn–0.4Zr(wt.%)alloysamplequenchedafterasolutiontreatment.
Fig.4.TEMbrightfieldimagesandSADPsoftheMg–2.8Nd–0.2Y–0.6Zn–0.4Zr(wt.%)alloyagedat200◦Cfor14h.(a,b)B//[0001]␣;(c),(d)B//[1120]␣;(e,f)B//[0110]␣.
observedintheanalysedvolume.After3hageing,plateletprecip- itatesenrichedwithNd,ZnandZrandhabitedon{2110},with anaveragelengthof10±3nmalong0110,wereevidentwithin theanalysedvolume,asshowninFig.7.Interestingly,thedistri- butionofYatomsremainedtobeuniformandwassimilartothat observedintheas-quenchedcondition.Afteragedat200◦Cfor14h, theaveragelengthofprecipitatesalong0110directionswere measuredto be15±4nm,in goodagreementwith16±4.3nm observedbyTEMexaminationin Fig.4a,and Yremainedtobe uniformly distributed in the ␣-Mg matrix, as shown in Fig. 8.
Examinationsofprecipitatesindifferentviewdirectionsconfirmed that most precipitates were elongated with their longitudinal axisparallel to[0001]␣, which is in agreementwithour TEM
observationsof, and1 precipitates.Suchprecipitatesare believedtobeeffectivetohamperthebasaldislocationmovement [23,24].
Thequantitativeanalysisresultsofatomprobedataareshown inFig.9.Thenumberdensityofthefinesoluteclustersaftersolu- tiontreatmentis5.0±0.2×1023m−3(whichwasestimatedonthe basisofthetotalnumberofsoluteclustersidentifiedintheanalysed volume),asshowninFig.9a.Thenumber densityofsolute-rich featuresreachedthepeakof7.6±0.6×1023m−3after1hageing, thendecreasedsignificantlyto3.4±0.6×1023m−3(3h)andfur- therdownto2.4±0.3×1023m−3(14h)withincreasinginageing time.ThesoluteconcentrationsofNd,ZnandZrinthe␣-Mgmatrix decreasegreatlyduringtheageingupto3h,inparticularthecase
Fig.5. HRTEMimages(a,b)oftheprecipitates(1)intheMgmatrixofanMg–2.8Nd–0.2Y–0.6Zn–0.4Zr(wt.%)alloyagedat200◦Cfor14h.B//[0001]␣.
Fig.6.APTelementalmapsofNd,YandZnofanMg–2.8Nd–0.2Y–0.6Zn–0.4Zr(wt.%)alloysamplequenchedafterasolutiontreatment(T4).(a)AcombinedmapofNd (green),Y(blue)andZn(red)afterremovingsolutesinthematrix;(b)Nd(green)map;(c)Y(blue)map;(d)Zn(red)map.(Forinterpretationofthereferencestocolorin thisfigurelegend,thereaderisreferredtothewebversionofthearticle.)
Fig.7.APTelementalmapsofNd,Y,ZnandZrofanMg–2.8Nd–0.2Y–0.6Zn–0.4Zr(wt.%)alloysampleagedat200◦Cfor3h.(a)Nd(green);(b)Y(lightblue);(c)Zn(red);(d) Zr(darkblue).(Forinterpretationofthereferencestocolorinthisfigurelegend,thereaderisreferredtothewebversionofthearticle.)
Fig.8. APTelementalmapsofNd,Y,ZnandZrofanMg–2.8Nd–0.2Y–0.6Zn–0.4Zr(wt.%)alloysampleagedat200◦Cfor14h.(a)Nd(green);(b)Y(blue);(c)Zn(red);and (d)Zr(darkblue).(Forinterpretationofthereferencestocolorinthisfigurelegend,thereaderisreferredtothewebversionofthearticle.)
forNd,asindicatedinFig.9b.ThedecreaseoftheNd,ZnandZr soluteconcentrations ofthematrixisdue tothepartitioningof theseelementsinto theprecipitate, asshown in Fig.9c and d.
Theaveragesoluteconcentrationofprecipitatesincreasedwiththe increasingageingtime,asshowninFig.9c.Theeffectofageingtime onpartitioningratioofsolutes(calculatedbytheaveragesolute
concentrationofprecipitatesoverthatofthematrix)isshownin Fig.9d.Ndwasobservedhavingamuchstrongerpartitioningthan Znduringthe3hageing.
Fig.10showstypicalcompositionprofilesmeasuredfromthree different precipitates.The thin precipitates,widely observedin shortly agedsamples, containeda lowlevel of solutesandhad
Fig. 9.Quantitative APT analysis results about the evolution of structure and chemical compositions of the solute-rich features and the Mg matrix in Mg–2.8Nd–0.2Y–0.6Zn–0.4Zr(wt.%)alloyduringshortageingat200◦Cupto14h.(a)numberdensity;(b)soluteconcentrationofthe␣-Mgmatrix;(c)soluteconcentration ofthesolute-richfeatures;and(d)thepartitioningratioofsolutesbetweensolute-richfeaturesandtheMgmatrix.
Fig.10.TypicalsolutecompositionprofilesmeasuredacrossthecoreofthreedifferentprecipitatesinMg–2.8Nd–0.2Y–0.6Zn–0.4Zr(wt.%)alloyagedat200◦Cfor14h.
a stoichiometry of Mg9(Nd,Zn), as shown in Fig. 10a. The thin precipitateslikely are.The compositionprofiles ofa slightly thickerprecipitate,asshowninFig.10b,giveastoichiometryof Mg4(Nd,Zn).Suchprecipitateswereoftenobservedinsampleaged formorethan3h.Theyarelikely.Athickprecipitatewasmea- suredhavingacorechemistryclosetoMg2(Nd,Zn),asshownin Fig.10c.Itisreasonabletobelievethattheprecipitateis1,given that1hasbeenpositivelyidentifiedbyTEMcharacterisationof a14-hageingsample(Fig.5).Thedifferentchemicalcompositions observedamongtheseindividualprecipitates(Fig.10)areconsis- tentwiththetrendobservedinFig.9c,inwhichtheaveragesolute concentrationofsolute-richfeaturesincreaseswithincreasingage- ingtime.Itisworthnotingthattheprecipitatechemistryunveiled bythisworkisdifferentfromMg3RE,Mg5RE(Mg12NdY),Mg3RE (Mg14Nd2Y)for, and1respectivelyreportedinadifferent Mgalloy[14].Itisunclearifthedifferenceisduetotheinfluenceof thealloycomposition,ageingcondition,ordifferentanalysistech- niquesused.Thinbasal␥-typeprecipitateswerenotresolvedby theAPTanalysis.Thisisprobablyduetothehighlocalmagnifica- tioneffectoftheprecipitatesandtheAPTdetectionefficiencyof
∼55%.
4. Discussion
4.1. Soluteclusteringanditsstrengtheningeffectduringthe early-stageprecipitationinMg–Ndbasedalloys
Thesoluteclusteringoftenoccursduringtheearly-stageage- ing prior to the formation of precipitates (such as  phase [11]), and can affect age-hardening response of various alloys [11,12,25,26]. Our APTcharacterisation unveiled that Ndwasa strong-clusteringalloyingelementandcontributedtheformation offinesoluteclusters(consistingofNdandZn)intheas-quenched Y-containing Mg alloy after a solution treatment,as shown in Fig.9c.TheclusternumberdensityintheY-containingalloywas 5.0±0.2×1023m−3,muchhigherthan0.2±0.22×1023m−3inan as-quenchedMg–Gd–Nd–Zn–Zralloy[11,12].Thisindicatesthat theYadditionhaspromotedtheclusteringofsolutesintheMg
alloy.Interestingly,noclearYclusteringwasdetectedduringthe earlyclusteringofsolutesasshowninFig.6c.Theprecisereason fortheenhancedNdclusteringintheMgalloywithaYadditionis unclear,whichisprobablycorrelatedwiththeeffectofYaddition onthesolubilityofNdinMgandtheenhanceddiffusionofNddue toitshighbindingenergywithvacancies(0.25eV)[27].Itisworth- notingthatalthoughthenumberdensityofsoluteclustersinthe MgalloyisclosertothatobservedinsomeAlalloys[22,25,26],it shouldnotbeasurprisethatthestrengthofthisMgalloyismuch lowerthanthoseoftheAlalloyssincethedeformationmechanisms ofMgalloysaregreatlydifferentfromthoseofAlalloys.Inaddition, notallsoluteclustersareabletoprovidethesamestrengthening effecteveninAlalloys[25].Thesignificantenhancedstrength(from 43HVto58HVinhardness,andfrom100MPato133MPainYTS) oftheY-containingalloyabovethatoftheY-freealloy,asshownin Figs.1and2,clearlyindicatedthatthesoluteclustersdidprovide effectivestrengtheningeffecttotheY-containingMgalloy.
4.2. Precipitatesandtheirstrengtheningeffectinthe Mg–Nd–Y–Zn–Zralloy
Theprecipitationsequenceof-typeprecipitatesinthelow- Y-containingMg–Ndbasedalloyis:soluteclusters→→→1
duringtheageingupto14haccordingtoourTEMandAPTchar- acterisations.␥-Typeprecipitateswereobservedinthealloy.The basal␥-typeprecipitatesareknowntobelesseffectivetohinder theslidingofdislocationsonbasalplanes,andarelesseffectiveto providestrengtheningeffectthanprime-typeprecipitatesinMg alloysatroomtemperature[23].Withincreasingageingtime,the sizeand chemicalcompositionof -typeprecipitates,asshown in Figs.7–9,increased.Meanwhile, there wasfurtherenhanced partitioningofNdandZnintoprecipitateswiththeincreasingage- ingtimefrom1hto14h,asshowninFig.9candd,whichwere correlatedwellwiththeformationofmoreand1precipitates containinghigherconcentrationsofNdandZn,asshowninFig.10.
Withtheincreaseinageingtime,thenumberdensityofsolute- reachfeaturesreachedthepeakvalueinthealloyat1hageingand decreasedprogressivelythereafter.Theincreaseinnumberdensity
ofsolute-rich features(including clustersand smallprecipi- tate)providedfurtherstrengtheningtothealloy,withahardness increaseof15HV,asshowninFig.1.Adecreaseofthenumber densityofsolute-rich features (Fig.9a), butanadditional hard- nessincreaseof7HV(Fig.1)inthealloyduringfurtherageingup to14h,indicatedthatvolumefractionofsolute-richfeaturesare importantforproducing strengtheningeffect.More information abouttheeffectofprecipitatevolumefractionontheprecipitation- strengtheningcanbefoundinRef.[28].
ManyMgalloyshavebeenobservedtoformprecipitatesina highnumberdensity,similartothatobservedinourY-containing Mgalloys,butdifferentage-hardeningresponse.For example,a hot-rolledMg–12Gd–1.9Y–0.69Zralloyagedat200◦Cwasreported toproducesolute-richfeaturesintheorderof1024m−3inanum- berdensity[16],higherthanthatobservedinourcastY-containing Mgalloy.Thewroughtalloywasreportedtohavethepeakhard- nessof130HV,muchhigherthanthatofourcastalloy.Thisis likelydue tothehighcontentof alloyingelementsandrefined grainsizeofthewroughtMgalloy.Incontrast,WE43alloywas reportedtohaveprecipitatenumberdensityof1.5×1024m−3and thepeakhardnessof100HV[29],whichismuchclosetothatof ourY-containingMgalloy.Inliteratures,someMgalloys,suchas AZ91alloy[30],werereportedhavingsimilarhardeningresponse asourY-containingMgalloy,butmuchlowernumberdensityof precipitates(1.09×1019m−3)thanthat of ourY-containingMg alloy.Thelow-number-densitymeasurementcouldoftenfoundto berelatedwiththedifferentanalysistechniqueusedintheirmea- surement.OurTEMandAPTcharacterisationoftheas-quenched samples(Figs.3and6,respectively)clearlydemonstratedthatfine soluteclusterseasilyidentified by APTare hard tobe resolved byTEMexaminations.Asaresult,TEMexaminationscouldeas- ilyunderestimatethenumber densityof solute-richfeatures in themicrostructure.Inordertomakeasensiblecomparisonamong resultsobtainedbydifferentanalysistechniques,thelimitations ofeach analysistechnique havetobetaken intoconsideration.
This indeed creates some difficulties to correlate with results obtainedbydifferentanalysistechniques.Tobetterunderstanding age-hardeningresponseofAZ91alloy,somedetailedAPTcharac- terisationwillbeuseful.
The Y-containingalloyagedfor 14hat200◦C wasobserved havingtheco-existence of-typeprismatic precipitatesand␥- typebasalprecipitatesinthemicrostructure,asshowninFig.4.
Singletypeprecipitates,either-typeor␥-type,werefrequently observedinmanyMgalloys.Forexample,-typeprecipitateswere observedintheMg–REalloys[9,14,17].The␥-typethinbasalpre- cipitateshavebeenobservedinMg–8Y–2Zn–0.6Zr(wt.%)alloy[31], Mg–1Gd–0.4Zn–0.2Zr (at.%) alloy [32], Mg–2.4Nd–0.4Zn–0.6Zr (wt.%)alloy[5]andMg–2.8Nd (wt.%)alloywiththeadditionof 1.3wt.%Zn[3,4].However,toourknowledge,thesimultaneousco- existenceof-typeprecipitatesand ␥-typeprecipitateshasnot beenreportedinpreviousstudiesofMg–Ndbasedalloys,although co-existenceofprecipitateslyingeitherparalleltothebasalplane oratanangletobasalplanehavebeenreportedinAZ91alloy[30].
Itisworthnotingthatthetwotypesofprecipitatescanproduce differentstrengtheningeffects.Thebasal␥-typeprecipitatesare effectivetoserveasobstaclesfornon-basalslippingofdislocations, andhenceimprovethehigh-temperaturemechanicalproperties of Mg alloys [33]. The formation of ␥-type precipitates in the Mg–2.8Nd–0.2Y–0.6Zn–0.4Zr(wt.%)alloy(Fig.4)wasconsistent withtheimprovedyieldstrengthofthealloyatelevatedtemper- atures(Fig.2b).Theprismatic-typeprecipitatesinthealloyare believedtoplayanimportantroletoenhancealloy’smechanical propertiesatroomtemperature.TEMobservation(Figs.3–5)and atomprobetomography(Figs.7and8)revealedahighnumber densityofplate-shapeseriesprecipitatesformedonprismatic planesofthe␣-Mgmatrixduringageingtreatment.Todate,itis
unclearwhytwotypesofprecipitatesformedsimultaneouslyin thealloy.Morefundamentalresearchisnecessarytounveilfactors controllingtheformationoftwotypesofprecipitates.Thiswillbe importantfordesignanddevelopingnewadvancedMgalloyswith excellentperformanceatroomtemperatureandelevatedtemper- atures.
5. Summary
1.AnewMg–2.8Nd–0.6Zn–0.4Zr(wt.%)alloywith0.2wt.%Yaddi- tionshowedasignificantimprovementonhardnessandtensile propertiesduringageingat200◦Cforupto14h.Thealloy,in particular,exhibitedenhancedtensileyieldstrengthinthetem- peraturerangeof200–350◦C.
2.The low-level Y addition has promoted the significant clus- tering of Nd and Zn, and subsequently the formation of high-number-density-typeprecipitateswithNd,Znandslight Zrenrichments.
3.ThepartitioningofNdinto-typeprecipitatesinthealloywas muchstrongerthanotheralloyingelements.Interestingly,the distributionofYatomsremainedinthealloywithincreasingin ageingtime,withoutexhibitinganysignificantpartitioninginto precipitatesinthealloyduringageingat200◦C.
4.Theimprovementofmechanicalpropertiesatelevatedtemper- atureswascorrelatedwellwithanenhancedprecipitationof
-typeprecipitates,including,and1,andtheco-existence withbasal␥-typeprecipitatesinthemicrostructureofthealloy.
5.The stoichoimetry of ,  and 1 was measured to be Mg9(Nd,Zn),Mg4(Nd,Zn),andMg2(Nd,Zn)respectivelyintheMg alloy.
Acknowledgments
Theauthorsaregratefulforscientificandtechnicalinputand supportfromtheAustralianMicroscopy&MicroanalysisResearch Facility (AMMRF) node at the University of Sydney. Jiehua Li alsowishestothanktheChinaScholarshipCouncilfor financial support.ThisworkispartlysupportedbytheDoctorateFounda- tion ofNorthwestern PolytechnicalUniversity under Grand No.
(CX200705).
References
[1] I.J.Polmear,Mater.Sci.Technol.10(1994)1.
[2]B.L.Mordike,Mater.Sci.Eng.A324(2002)103–112.
[3]T.J.Pike,B.Noble,J.LessCommonMet.30(1973)63.
[4] P.A.Nuttall,T.J.Pike,B.Noble,Metallography13(1980)3–20.
[5]D.H.Ping,K.Hono,J.F.Nie,ScriptaMater.48(2003)1017–1022.
[6]R.Wilson,C.J.Bettles,B.C.Muddle,J.F.Nie,Mater.Sci.Forum.41(267)(2003) 9–422.
[7]L.R.Gill,G.W.Lorimer,P.Lyon,Adv.Eng.Mater.9(2007)784.
[8]P.J.Apps,H.Karimzadeh,J.F.King,G.W.Lorimer,ScriptaMater.48(2003) 1023–1028.
[9]G.Sha,J.H.Li,W.Xu,K.Xia,W.Q.Jie,S.P.Ringer,Mater.Sci.Eng.A527(2010) 5092–5099.
[10]J.H.Li,W.Q.Jie,G.Y.Yang,Trans.NonferrousMet.Soc.China18(2008)s27–s32.
[11]G.Sha,J.H.Li,W.Q.Jie,S.P.Ringer,in:K.U.Kainer(Ed.),Proceedingsof8th InternationalConferenceonMagnesiumAlloysandtheirApplications,Wiley- VCHVerlagGmbH&Co.KGaA,Weinheim,2009,pp.40–45.
[12]J.H.Li,G.Sha,T.-Y.Wang,W.Q.Jie,S.P.Ringer,Mater.Sci.Eng.A(2011), doi:10.1016/j.msea.2011.10.092.
[13]J.F.Nie,B.C.Muddle,ScriptaMater.10(1999)1089–1094.
[14]J.F.Nie,B.C.Muddle,ActaMater.48(2000)1691–1703.
[15] C.Antion,P.Donnadieu,F.Perrard,A.Deschampa,C.Tassin,A.Pisch,ActaMater.
51(2003)5335–5348.
[16]T.Honma,T.Ohkubo,K.Hona,S.Kamado,Mater.Sci.Eng.A395(2005)301–306.
[17]T.Honma,T.Ohkubo,S.Kamado,K.Hono,ActaMater.55(2007)4137–4150.
[18]J.H.Jun,B.K.Park,J.M.Kim,K.T.Kim,Adv.Mater.Res.26-28(2007)137–140.
[19]M.P.Moody,B.Gault,L.T.Stephenson,D.Haniel,S.P.Ringer,Ultramicroscopy 109(2009)815–824.
[20] J.M.Hyde,C.A.English,in:R.G.E.Lucas,L.Snead,M.A.J.Kirk,R.G.Elliman(Eds.), MicrostructuralProcessesinIrradiatedMaterials,MRS2000FallMeetingSym- posium,Boston,MA,2001,p.27.
[21]D.Vaumousse,A.Cerezo,P.J.Warren,Ultramicroscopy95(2003)215–221.
[22]G.Sha,R.K.W.Marceau,X.Gao,B.C.Muddle,S.P.Ringer,ActaMater.59(2010) 1659–1670.
[23]J.F.Nie,ScriptaMater.48(2003)1009–1015.
[24]C.L.Mendis,C.J.Bettles,M.A.Gibson,C.R.Hutchinson,Mater.Sci.Eng.A435-436 (2006)163–171.
[25]R.K.W.Marceau,G.Sha,R.Ferragut,A.Dupasquier,S.P.Ringer,ActaMater.58 (2010)4923–4939.
[26]R.K.W. Marceau,G. Sha, R.N.Lumley, S.P.Ringer, Acta Mater.58(2010) 1795–1805.
[27]T.Gorecki,in:F.J.Kedves,D.L.Beke(Eds.),Proc.Int.Conf.OnDiffusionin MetalsandAlloys(DIMETA-82),TransTechPublications,Switzerland,1983, p.365.
[28]S.Esmaeili,D.JLloyd,ActaMater.53(2005)5257.
[29]G.Barucca,R.Ferragut,F.Fiori,D.Lussana,P.Mengucci,F.Moia,G.Riontino, ActaMater.59(2011)4151–4158.
[30]S.Celotto,ActaMater.48(2000)1775–1787.
[31] Y.M.Zhu,A.J.Morton,M.Weyland,J.F.Nie,ActaMater.58(2010)464–475.
[32]J.F.Nie,K.Oh-ishi,X.Gao,K.Hono,ActaMater.56(2008)6061–6076.
[33] A.Galiyev,R.Kaibyshev,G.Gottstein,ActaMater.49(2001)1199–1207.