E1(镁合金黑色氧化耐热涂层的研究)

Thermalstudiesonelectrodepositedblackoxidecoating

onmagnesiumalloys

A.K.Sharmaa,*,R.UmaRania,S.M.MayannabbThermalSystemsGroup,ISROSatelliteCentre,Bangalore560017,India

DepartmentofChemistry,CentralCollegeCampus,BangaloreUniversity,Bangalore560001,India

Received11October2000;receivedinrevisedform17April2001;accepted19April2001

aAbstract

Thermalbehaviorofblackanodiccoatingsonmagnesiumalloy,AZ31Bandmagnesiumlithiumalloy,MLA9hasbeeninvestigated.Thechemicalnatureofcoatingischaracterizedbyinfraredspectralstudies.Thethermoanalyticalinvestigationshavebeencarriedoutusingthermogravimetry(TG),derivativethermogravimetry(DTG),anddifferentialscanningcalorimetry(DSC).Thedecompositionproceedsthroughthreestepsviz.,dehydration,decompositionofchromiumhydroxideandsulphateanddecompositionofmagnesiumchromatetooxide.Measurementofhemisphericalemittanceofcoatingsversestemperaturewasinvestigatedusingcalorimetricmethods.Thestudiesrevealedthatthethermalemittanceofcoatingsincreaseswithtemperature.#2001ElsevierScienceB.V.Allrightsreserved.

Keywords:Thermoanalyticalbehavior;Anodiccoatings;Magnesiumalloys;Thermalemittance;Spaceapplications

1.Introduction

Magnesiumischaracterizedbylowdensityandhighreactivity.Magnesiumalloysarethelowestdensityengineeringalloysusedinawiderangeofapplications.Thesealloysarethepromisingcandi-datesinaerospaceandalliedfieldsprimarilytosavethefuelcost.Attemptshavebeenmadeforadditionofvariousalloyingelementstomagnesiumsoastoproducealloysofhighstrength,highcreepresistanceandlowdensity.

Magnesium–lithiumalloysarethelighteststruc-turalmetallicalloysknowntoday.Additionoflithiumwitharelativedensityof0.53,inmagnesiumreduces

Correspondingauthor.Tel.:þ91-80-508-3106;fax:þ91-80-508-3203.

E-mailaddress:aks@isac.ernet.in(A.K.Sharma).

*thedensityofalloysignificantly.Furthermore,addi-tionof$11%lithiumconvertshexagonalclosepackedstructureofpuremagnesiumtothebodycenteredcubiclattice,markedlyimprovingformabil-ityofthealloy[1].Thebinarymagnesium–lithiumalloys,thoughlightandductilearenotstrong.Addi-tionofaluminumresultsinconsiderableimprovementinstrength,primarilyduetotheformationofMgLi2Almetastablephase[2–4].Magnesium–lithiumalloy,MLA9,containing11–13%Li,1.25–1.75%Al,andbalancedMgwt.%,isdevelopedatDefenceMetal-lurgicalResearchLaboratory(DMRL),Hyderabad,India.Thesealloysare25%lighterthanconventionalmagnesiumalloys(density$1.35g/cm3)resultingintoaconsiderableweightsaving.

Chemicalconversioncoatings,viz,anodizing,chro-mating,etc.areoftenemployedasanidealmeansofimprovingoneormoresurfaceproperties,chemical,mechanical,electricaloroptical.Theseareusedto

0040-6031/01/$–seefrontmatter#2001ElsevierScienceB.V.Allrightsreserved.PII:S0040-6031(01)00534-2

68A.K.Sharmaetal./ThermochimicaActa376(2001)67–75

preventatmosphericcorrosion,toincreasemicrohard-nessandreducingfrictiononslidingsurfaces;toprovidebetteradhesionforpaints,lubricantsandadhesives;toimpartthermal/electricalresistance;asasolidfilmlubricanttopreventcoldweldinginspaceconditionsandtoprovideadequateopticalsurfaceforthermalcontrolapplications[5].

Theblackanodiccoatingsonmagnesiumalloyshavebeendevelopedforthermalcontrolofspacecraftandrelatedapplications.Thesecoatingsareusedontheelectronichousingofspacecrafttoimproveradia-tivecoupling.Inspacecraftsomeoftheelectronicpackages,whichareinoperation,generateenormousheatwhileothers,whicharenotinoperation,facethecolddeepspace.Thisgeneratesahighdegreeoftemperaturegradientacrosstheinternalpackagesofspacecraft.Theflatabsorberblackanodiccoatingwithhighabsorptanceandthermalemittancehelpsinminimizingtemperaturegradientacrossthepackagesbyimprovingtheirheatradiationcharacteristicsasperthefollowingequation[6].SAPa¼seAT4orT¼

󰀁

SAPa󰀂1=4

sAe

whereSisthesolarconstant(meanvalue1353W/m2);APtheprojectedsurfaceareaofthespacecraft(m2)perpendiculartothesunrays;athesolarabsorptanceoftheprojectedarea;stheStefan–Bolzmanconstant(5:67Â10À12W/cm2k4);Athetotalsurfaceareaofthespacecraft(m2);etheinfraredemittanceofthesurfaceofexposedarea;andTistheabsolutetem-peratureofthespacecraft.

AsS,AP,s,andAareconstantsinthisrelationship,itclearlyshowsthatthetemperatureofanygivenareaofspacecraftisdirectlycontrolledbythea/Eratio.Here,theterm‘absorptance’referstoallsolarradia-tion(X-ray,ultraviolet,visible,infrared,radiofre-quency,etc.),whereastheterm‘emittance’isrestrictedtoinfraredrangebecausethermalradiationsoccurmainlyintheinfraredregion.

Thecoatingsusedinspacetechnologyrequirehigherstandardsandbettercontrolthanusedforgroundapplicationssincespaceconditionsareverysevereandon-orbitspacecraftisnotapproachableforrepair.

2.Experimental2.1.Anodizingprocess

Galvanicblackanodizingonmagnesiumalloy,AZ31B(Al3.0%,Zn2.0%,balanceMgwt.%)andmagnesium–lithiumalloy,MLA9(Li11.9%,Al1.5%,balanceMgwt.%)wascarriedoutbythemethodsdescribedinourpreviouspublications[7,8].Abriefdescriptionoftheseproceduresisgivenbelow.

2.2.A.Magnesiumalloy,AZ31B

1.Ultrasonicsolventdegreasinginisopropanolfor5–10min.

2.Alkalinecleaningfor5–10mininasolutioncontaining50g/dm3sodiumhydroxideand10g/dm3trisodiumorthophosphate,operatingat60Æ58C.Waterrinse.

3.Acidpicklingin180g/dm3chromiumtrioxide,40g/dm3ferricnitrateand3.75g/dm3potassiumfluoridefor2min.Waterrinse.

4.Blackanodizinginasolutionformulatedandoperatedasfollows:Potassiumdichromate,25g/dm3;ammoniumsulfate,25g/dm3;pH5.8;temperature,25Æ28C;cathode,stainlesssteelanodizingtank;anodetocathoderatio,1:10;anodizingtime,30min,forthestandardcoatingthicknessof5–8mm.Waterrinse.5.Heattreatmentat708Cfor2h.2.3.B.Magnesium–lithiumalloy,MLA9

1.Step1and2aresameasdescribedformagne-siumalloy,AZ31B.

2.Acidpicklingin500g/dm3chromiumtrioxide,1g/dm3ferricnitrateand0.5–1g/dm3potassiumfluoridefor3–5min.Waterrinse.

3.Fluorideactivationbydippingthespecimeninhydrofluoricacid(40%)50ml/dm3for10min,followedbywaterrinsing.

4.Blackanodizinginasolutioncomposedandoperatedasfollows:Potassiumdichromate,25g/dm3;ammoniumsulfate,25g/dm3;pH5.5;temperature,25Æ28C;cathode,stainlesssteelanodizingtank;anodetocathoderatio,1:10;anodizingtime,60min;forthestandardcoatingthicknessof14to18mm.Waterrinse.

A.K.Sharmaetal./ThermochimicaActa376(2001)67–7569

5.Heattreatmentat708Cfor2h.

IsolationofBlackAnodicCoatingsfromthesub-strateforchemicalandthermoanalyticalanalysiswascarriedoutbygentlyscrapingthefilmfromthesubstrateusingasharpscalpel.Whilescrapingthefilmcarewastakenthatnosubstratematerialshouldgetscrapedalongwiththecoating.2.4.Measurementtechniques

ThethicknessoftheanodiccoatingwasmeasuredusingIsoscopeMP2B-T33B,HelmutFischer(Ger-many),coatingthicknesstester.Thisinstrumentworksontheeddycurrentprincipleandisusedtomeasurethethicknessofnon-conductivecoatingsonacon-ductivesubstrate.

Thechemicalnatureofcoatingwasdeterminedbyusinginfraredspectroscopy.IRspectrawererecordedonFTIRSpectrometer(NicoletModelIMPACT400D)usingKBrÀpellets1inthewavenumberregionof4000–400cm.

ThethermalstabilityoftheanodiccoatingwasinvestigatedbyusingDupont951thermogravimetricanalyzertorecordthermogravimetric(TG)andderi-vativethermogravimetry(DTG)curvesandDupont910differentialscanningcalorimetry(DSC)systemstorecordDSCcurves.Thethermoanalyticalcurveswererecordedinargonatmospherewithagasflowrateof40cm3/sandasampleheatingrateof108C/min.TGandDTGcurveswererecordedfor19.4mgsamplessimultaneouslyupto10008C.DSCdatafor15.9mgsampleswerelimitedtoatemperatureof5008C,duetolimitedtemperaturecapabilityoftheinstrument.

3.Resultsanddiscussion3.1.Infraredspectra

TheassignmentofIRbandshasbeenmadeonthebasisofpublisheddata[9–11].Thespectraofgalvanicblackanodicfilmonmagnesiumalloys,AZ31Bandmagnesium–lithium,MLA9areshowninFigs.1and2.StrongbandsassociatedwithOHstretchingvibra-tionsofwaterandhydroxylgroupsoccurbetween3200and3700cmÀ1.Waterofhydrationiseasily

distinguishedfromhydroxylgroupsbythepresenceofH–O–Hbendingmotion,whichproducesamediumbandintheregion1600–1650cmÀ1[9].Theappear-anceofIRabsorptionbandsinFigs.1and2at$3435cmÀ1areassociatedwithOHstretchingvibra-tionsofwaterofhydration.Thebandsat1631cmÀ1inFig.1and16cmÀ1inFig.2showtheH–O–Hbendingmotion,whichconfirmsthepresenceofchromiumhydroxide.

Absorptionbandsat$1100cmÀ1accompaniedÀ1byaconsiderablyweakerbandat$670cmareasso-ciatedwithchromiumsulphate[10].ThedoubletoccurredatÀ1917and802inFig.1andthesinglepeakat870cmappearedinFig.2revealthepresenceofmagnesiumchromate[11].

Thebandsatlowerfrequencies$530,$470cmÀ1areassignedtothestretchingfrequenciesofM–Obonds,indicatingthepresenceofchromiumandmagnesiumoxides.

3.2.Hemisphericalemittance

Measurementsoftotalhemisphericalemittancehavebeenmadeusingcalorimetricmethod[12].Thisconsistsofmeasurementofthepower-inputpara-metersandtemperaturesofsampleandtheshroud.Usingthesedata,emittanceoftheblackanodizedsurfaceundersteadystateconditionswerecalculatedusingfollowingenergybalanceequatione¼

p

sAðTS4ÀTO4Þ

whereeisthetotalhemisphericalemittance,Pthe-power,Heatinputtothesample(watts),s2theStefan–Boltzmannconstant(5:67Â10À12W/cmK4),Atheareaofthesample(cm2),TSthetemperatureofsample(K)andT0isthetemperatureofshroud(K).

Themeasurementsweremadeinastainlesssteeldoublewalledtestchamber.Theinnerwall(shroud)isisolatedfromtheouterwallwiththeTeflonspacers.Theinsidewallofthechamberthatreceivestheradiationfromtestspecimeniscoatedwithblackpaintformaximizingthethermalcoupling.Thischamberisconnectedtoavacuumsystemtomaintainthepressureinsidethechamberatthelevelof76Â10À7mbar.Themeasurementswerecarriedoutonaspeciallydesignedcuptypesamplemadeoftwocirculardiscsof75mmdiameterwiththe

70A.K.Sharmaetal./ThermochimicaActa376(2001)67–75

Fig.1.Infraredspectraofanodiccoatingofmagnesiumalloy,AZ31B.

guidinggrooves.Boththeoutsidecircularfacesandedgeofthesamplewereblackanodized.Thestandardcircularthermofoilheaterwassandwichedbetweenthesetwodiscs.Fourcopper-constantanthermocou-plesweremountedonthesampletomonitorthesampletemperature.Thehemisphericalemittanceofblackanodizedsamplesofmagnesiumandmagne-sium–lithiumalloyswithtemperatureaftererrorcor-rectionwithreferencetoinfraredEmissometerreadingsisshowninFig.3.Thehemisphericalemit-tanceofboththecoatingsincreaseswithtemperature.3.3.Thermogravimetricanalysis

Figs.4and5showtheTG,DTGandDSCdataofblackanodicoxidecoatingsonmagnesiumand

magnesium–lithiumalloys,respectively.Thesether-moanalyticalcurvesrevealthreemajorchangesthatoccurwhenblackanodicoxidecoatingsareheated:(1)dehydration,(2)decompositionofchromiumhydroxideandsulphate,and(3)decompositionofmagnesiumchromate.

3.3.1.Dehydration

Dehydrationofboththeanodiccoatingsonmag-nesiumaswellasmagnesium–lithiumalloysstartsat$508Candcontinuesuntil$2668C.Thecorrespond-ingDTGpeakofdehydrationisobtainedat978Cformagnesiumalloy,AZ31B.Whileforthecoatingonmagnesium–lithiumalloy,MLA9,thischangeisrepresentedbyadualpeakat87and2238C.ThedehydrationstepintheDSCcurvesisrepresentedas

A.K.Sharmaetal./ThermochimicaActa376(2001)67–7571

Fig.2.Infraredspectraofanodiccoatingofmagnesium–lithiumalloy,MLA9.

anendothermicpeak[13]at1308CforcoatingsonboththealloysAZ31BandonMLA9.

3.3.2.Decompositionofchromiumhydroxideandsulphate

Thesecondstagedecompositionrepresentingthedecompositionofchromiumhydroxideandsulphatealongwiththelossofremainingwatercontentsisobservedbetween266–38Cformagnesiumalloy,AZ31Band310–4468Cformagnesium–lithiumalloy,MLA9.ThecorrespondingDTGpeaksareobtainedat309and3478CforcoatingsonAZ31BandMLA9,respectively.ThischangeisrepresentedbyDSCasanendothermicpeakataround1308C(forAZ31B)and3418C(forMLA9).Thetotalweightlossinfirstandsecondstagedecompositionisaround30and23.8%fornon-heattreatedandheattreatedcoatings,respectively

incaseofmagnesiumalloy,AZ31B.InthecaseMLA9itisaround25.8and25.2%fornon-heattreatedandheat-treatedcoatings,respectively.

3.3.3.Decompositionofmagnesiumchromate

Thethirdstepdecompositionisthedecompositionofmagnesiumchromatetooxide.Thedecompositionstartsaround4438Candcontinuesupto6378CforcoatingsonAZ31B.Thisdecompositionisrepre-sentedbetween446and6948CforcoatingsonMLA9.TheCorrespondingDTGpeaksareobtainedat560and6268CforcoatingsonAZ31BandMLA9,respectively.Atotalmasslossofabout35.5and26.9%isobserveduptothistemperaturefornon-heattreatedandheat-treatedcoatings,respectively,forcoatingsonAZ31Band34.0and25.8%forcoatingsonMLA9.

72A.K.Sharmaetal./ThermochimicaActa376(2001)67–75

Fig.3.Infraredemittanceofanodiccoatingvs.temperature.

Table1

ThermoanalyticaldataofanodicoxidefilmsDecompositionsteps

Magnesiumalloy,AZ31BWithoutheattreatment

1.Dehydration

Decompositionrange(8C)Weightloss(%)

DTG,peaktemp.(8C)DSC,peaktemp.(8C)Heatofreaction(J/g)

50–26623971309

Heattreated50–26619121142263266–39.83313531466–6613.156026.9

Magnesium–lithiumalloy,MLA9Withoutheattreatment50–31017.687130692310–4468.2

223,347341770446–6948.262634.0

Heattreated50–31016.481127639310–4468.8

223,360341602446–6840.661425.8

2.DecompositionofchromiumhydroxideandsulphateDecompositionrange(8C)266–3Weightloss(%)7.0DTG,peaktemp.(8C)309DSC,peaktemp.(8C)347Heatofreaction(J/g)2653.DecompositionofmagnesiumchromatesDecompositionrange(8C)443–637Weightloss(%)5.5DTG,peaktemp.(8C)560Totalweightloss(%)35.5

A.K.Sharmaetal./ThermochimicaActa376(2001)67–7573

Fig.4.Thermoanalyticalcurvesofanodiccoatingofmagnesiumalloy,AZ31B;thedottedcurvesareforheattreatedcoatings(708Cfor2h).

Beyond6378C(forAZ31B)andbeyond6948C(forMLA9)andupto10008C,thereisnootherthermaleventandthecoatingsarethermallystablewithoutanyweightloss.

TheheatsofreactionfordifferentdecompositionstagesofanodicfilmwerecalculatedfromtheDSCcurvesusingtheexpressionH¼KA/m,whereHistheheatofreaction,Kisthecalibrationconstant,Aistheareaunderpeakandmisthemassofthesample.Atthepeaktemperatureof1308C,theheatsofreactionforthedehydrationofoxidecoatingonAZ31Balloywere9J/gfornonheattreatedcoatingand263J/g,forheattreatedcoating.Whileatthepeaktemperatureof3478C,theheatsofreactionrepresent-ingthedecompositionofchromiumhydroxideandsulphatealongwiththelossofremainingwaterwere265J/gfornon-heattreatedand1J/gforheattreatedcoatings,respectively.

Foranodiccoatingonmagnesium–lithiumalloy,MLA9,theheatsofreactionatthepeaktemperatureof1308C(representingdehydration)were692J/gfornon-heattreatedcoatingand639J/gforheat-treatedcoating.Whileatthepeaktemperatureof3418C,(representingdecompositionofchromiumhydroxideandsulphate),theheatsofreactionwere770J/gfornon-heattreatedand602J/gforheat-treatedcoatings,respectively.

Thethermo-analyticaldataofblackanodicoxidecoatingsonmagnesiumalloy,AZ31Bandmagne-sium–lithiumalloy,MLA9ispresentedinTable1.

74A.K.Sharmaetal./ThermochimicaActa376(2001)67–75

Fig.5.Thermoanalyticalcurvesofanodiccoatingofmagnesium–lithiumalloy,MLA9;thedottedcurvesareforheattreatedcoatings(708Cfor2h).

Theseanodicoxidecoatingsonmagnesiumareverysoft(geltype)whenformed.Afterdryingandheattreatmenttheyloosesomewaterofhydrationandattainsufficienthardness.Becauseofdifferenceinthepercentageofwaterofhydrationinasanodized(nonheattreated)andheattreatedcoatingsthevaluesoftheirheatsofreactionaredifferent.

Theanodicoxidecoatingsonmagnesiumalloysdescribedhereinareconversioncoatings.Thechemi-calcompositionofcoatingchangesasthecoatingthicknessincreases.Initiallythecoatinghassmoothfinegrainmorphologywhileathigherthicknessitissomewhatroughandpowdery.Thepowderycoatinghaslesspercentageofwaterofhydration.

Onmagnesiumalloy,AZ31Bwhichiscomparativelylessproneforatmosphericcorrosion,ananodiccoatingthicknessof5–8mmissufficient.Whileforahighlyreactivemagnesium–lithiumalloy,MLA9acoatingthicknessof14–18mmisrequiredforadequatecorro-sionprotection.ThisexplainsthemarkeddifferenceintheheatsofreactionofasanodizedandheattreatedanodicoxidefilmonAZ31Balloy(lesscoatingthick-ness,moredifferenceinthepercentageofwaterofhydrationofasanodizedandheattreatedanodicfilm).

A.K.Sharmaetal./ThermochimicaActa376(2001)67–7575

4.Conclusions

Thermalbehaviorofblackanodicoxidecoatingsonmagnesiumalloy,AZ31Bandmagnesium–lithiumalloy,MLA9hasbeeninvestigated.Infraredspectralstudiesrevealthatthesecoatingsconsistofchromiumhydroxide,chromiumsulphate,magnesiumchromateandwaterofhydration.

Theblackanodiccoatingsprovidehighemittance(>0.80),indicatingtheirextremesuitabilityforther-malcontrolapplications.Studiesonmeasurementofhemisphericalemittanceofcoatingsversestempera-turerevealthattheemittanceofcoatingincreaseswithtemperature.

Thermoanalyticalstudies(TG,DTGandDSC)showthatthedecompositionofcoatingproceedsthroughthreesteps,thefirstisdehydration,thesecondstepsinvolvesthedecompositionofchromiumhydro-xideandsulphateandinthelaststepmagnesiumchromateisdecomposedtooxide.

Acknowledgements

TheauthorsaregratefultoDr.P.S.Goel,Director,ISROSatelliteCentre(ISAC),Bangalore,Mr.A.V.Patki,DeputyDirector,MSA,Mr.H.Narayanamurthy,

GroupDirector,TSG,andMr.H.Bhojraj,Head,TFD,ISACfortheirkeeninterestandencouragement.TheauthorsdulyacknowledgethesupportreceivedfromMr.A.Ramasamy,V.Ramakrishnan,andN.K.Sundar-esan,TTD.References

[1]C.R.Chakravorty,Bull.Mater.Sci.17(6)(1994)733.[2]J.McDonald,Trans.ASM61(1961)505.[3]J.McDonald,J.Inst.Metals97(1969)353.[4]J.McDonald,J.Inst.Metals99(1971)24.[5]A.K.Sharma,Trans.SAEST30(1)(1995)1.

[6]B.N.Agarwal,DesignofGeosynchronousSpacecrafts,Printice-Hall,EnglewoodCliffs,NJ,1986,p.281.[7]A.K.Sharma,MetalFinishing91(6)(1993)57.

[8]A.K.Sharma,R.UmaRani,H.Bhojaraj,H.Narayanamurthy,J.Appl.Electrochem.23(1993)500–507.

[9]

A.RichardNyquist,R.O.Kagel,InfraredSpectraOfInorganicCompounds,Vol.4,AcademicPress,NewYork,1971,p.3.

[10]F.A.Miller,C.H.Wilkins,Anal.Chem.24(8)(1952)1259.[11]K.Nakamoto,IRSpectraofInorganicandCo-ordinationCompounds,Wiley,NewYork,1963,p.107.

[12]

A.Ramasamy,N.K.Sundaresan,Padmanabhan,P.P.Gupta,H.Narayanamurthy,SAETechnicalpaperseries921329,in:Proceedingsofthe22ndInternationalConferenceonEnvironmentalSystems,Washington,July1992.

[13]

A.K.Sharma,H.Bhojaraj,PlatingSurf.Finishing76(2)(19)59–61.