Preparation of Fe(II)-montmorillonite by reduction of Fe(III)-montmorillonite with ascorbic acid

ContentslistsavailableatScienceDirect

AppliedClayScience

journalhomepage:www.elsevier.com/locate/clay

PreparationofFe(II)-montmorillonitebyreductionofFe(III)-montmorillonitewithascorbicacid

JayappaManjanna⁎

GraduateSchoolofEngineering,HokkaidoUniversity,Sapporo060-8628,Japan

articleinfoabstract

AnewandconvenientmethodisdescribedheretoprepareFe(II)-montmorillonite(mont)usingFe(III)-montasthestartingmaterial.Fe(III)-montwastreatedwithascorbicacid(vitaminC)solutionat70°Cfor~12hinAratmosphere.Alloftheadsorbediron(III)wasreducedtoformFe(II)-mont.ThepresenceofanyassociatedFe-oxidephaseinthestartingmaterial,Fe(III)-mont,wasratheradvantageoustoreachstoichiometricFe(II)-mont.TheFe2+/FetotalratiooftheironextractedfromfreshlypreparedFe(II)-montwasfoundtobeclosetounity.Thetotalamountofironextractedin1MNH4Clwasequivalenttothecationexchangecapacityoftheparentmontmorillonite(KunipiaFmontmorillonite)intermsofFe(II),providedtherewassufficientiron(III)inthestartingmaterial.Thebasalspacingatarelativehumidity(RH)of40%was14.7Å,typicalofmontmorillonitewithdivalentinterlayercations.FTIRspectrashowednosignificantchangesinthebasicclaymineralstructure.ThestructurewasfurtherprovedbyMössbauerspectraatroomtemperature(RT).ThusFe(II)-montpreparationasdescribedherewasfoundtobehighlyconvenientforroutinepreparationintheambientatmosphericcondition,alsoforlargequantities.Thefeasibilityofsolid-statepreparationofFe(II)-montwasalsodemonstrated.ThestabilityofFe(II)-montwasprovedindispersioninwaterunderpartiallydeoxygenatedcondition.Therewasabout30%and70%oxidationofinterlayerFe(II)ionsinabout47daysatRTand65°C.Thus,theinterlayerFe(II)ionsweregraduallyoxidized(i.e.,notrapidly,asanticipated).Forlong-timepreservationoftheFe(II)-montsample,itisessentialtokeepdeoxygenated,lowRH(b40%)conditionsandwithoutexposingtoelevatedtemperatures.

©2008ElsevierB.V.Allrightsreserved.

Articlehistory:

Received19September2007

Receivedinrevisedform30January2008Accepted11February2008Availableonline4March2008Keywords:

Fe(II)-montmorilloniteFe(III)-montmorilloniteAscorbicacid

1.Introduction

Environmentallybenignandredoxsensitivemetalions,Fe(II)andFe(III),intheinterlayerofmontmorillonitecanimpartmanytechnologicalapplications.MostoftheFe-montmorillonitesintheliteratureareconcernedwithFe(III)species(andnotferrousions)whichhavebeenstudiedforapplicationsindifferentfieldsthatincludepillaredclaysascatalysts(Zuritaetal.,1996;Palinkoetal.,1997;Kloprogge,1998;Kantametal.,1998;YamanakaandHattori,1988;Doradoetal.,2006)inorganicreactions,andadsorbentfortheremovaloftheenvironmentaltoxicants(Lenobleetal.,2002;Izumietal.,2005;Tabetetal.,2006;Masihetal.,2007).TheseFe-montmorillonitesarecommonlyobtainedbycationexchangeand/orintercalationofNa-montmorillonitewithhydrolyzedFe(III)species(HerreraandPeech,1970;Valverdeetal.,2005).IfFe(II)ionsaretheinterlayercationsofmontmorillonite,manyredoxreactionsareexpected.Sofar,onlythestructuralironofiron-richclaymineralshasbeenexplored(Stuckietal.,1984;Stucki,1988;Komadeletal.,

⁎PresentlyatNDEandScienceResearchCentre,FacultyofEngineering,IwateUniversity,Morioka020-8551,Japan.Tel./fax:+81196216350.

E-mailaddress:jmanjanna@rediffmail.com.0169-1317/$–seefrontmatter©2008ElsevierB.V.Allrightsreserved.doi:10.1016/j.clay.2008.02.005

2000;Komadel,2003;Leeetal.,2006),buttheapplicationoftheseclaymineralsseemstoberestrictedduetotherapidre-oxidationofreducedironand/ortheavailabilityofsuchredox-modifiedclaymineralsinbulkamountsunderambientatmosphericconditions.Thereareonlyafewstudies(Kameietal.,1999;Kozaietal.,2001,2007;Manjannaetal.,2007)dealingwithFe(II)-montpresumablyduetothelackofaconvenientmethodtoobtaintheFe(II)-mont.Conventionally,FeCl2couldbeusedfordirectcationexchangeunderstrictoxygencontrol(Kozaietal.,2001).However,theformationofionic-pairslikeFeCl+anditsstrongaffinitytotheclaymineralsurface(CharletandTournassat,2005)causedchlorideimpuritiesinthepreparedsampleofFe(II)-mont(Kozaietal.,2007).Also,whilewashingthemontmorillonitetoremoveexcessFeCl2,thehydrolysis/precipitationofironcompounds(duetosuccessiveincreaseofpH)mayleadtoexcessamountsofiron.ToovercomethedifficultiesassociatedwiththeuseofFe(II)salts,anewmethodwasproposedrecently(Manjannaetal.,2007).ThelessstableFe(II)-nitrilotriacetate(NTA)complex(logK=8.3)andnotthemorestableFe(III)-NTA(logK=15.9)facilitatedthecationexchangeofNa-montmorillonite.BulkpreparationwasnotpossiblebecauseofthecompetingprecipitationofFe(II)-NTAabove20mM(duetolowsolubility).TheexcessiveuseofNTAisharmfultotheenvironment(Fe-NTAiscarcinogenic).

Furthermore,inviewoftheclaymineralalterationduetoironcorrosionproductsinthegeologicaldisposalofnuclearwaste(JNC,

J.Manjanna/AppliedClayScience42(2008)32–3833

2000;Guillaumeetal.,2004;Wilsonetal.,2006;Carlsonetal.,2007),Fe(II)-montisenvisagedasoneofthesimplestalterationproductatreducingconditions.HencethepropertiesofFe(II)-mont(suchasswelling,hydraulicconductivity,diffusionofradionuclidesetc.)areveryimportant.Forthesestudies,theavailabilityofFe(II)-montinlargeramountsisneeded.

InanefforttodevelopasimpleandenvironmentallybenignmethodforthepreparationofFe(II)-montatambientconditions,Fe(III)-montwasconvertedtoFe(II)-montthroughthereductionofinterlayer/adsorbedFe(III)ionswithascorbicacid(vitaminC)solution.

2.Materialsandmethods

WeusedNa-montmorillonite(KunipiaF,Japan)withacationexchangecapacity(CEC)~113meq/100g(Kozaietal.,1993)andapproximatechemicalcomposition,(Na0.431K0.002Ca0.002)(Al1.56Mg0.305Fe0.099Ti0.007)oct(Si3.949Al0.051)tetO10(OH)2nH2O.

Allsolutionswerepreparedusingdistilledwater(conductivity0.06μScm−1),andthechemicalswereofanalyticalreagentgrade.Theexperimentswereperformedatroomtemperature,20±3°C(RT)unlessotherwisementioned.Ar-gaswaspurgedintheappropriatesolutions.

TheconcentrationofironinsolutionsandtheFe2+/Fetotalratioi.e.,[Fe(II)/(Fe(II)+Fe(III))]wasdeterminedbytheo-phenanthrolinemethod(Bassettetal.,1978)usingaUV–Visiblespectrophotometer(λmax510nm).Insomecases,Fetotalwasdeterminedbyinductivelycoupledplasmaatomicemissionspectroscopy(ICP-AES).Forthesemeasurements,thedispersedclayparticleswereremovedbycentrifugationfollowedbyultra-filtration(0.01μmfilter).

Forcomparison,Na-,Ca-andCr(III)-montwerealsopreparedfromKunipiaFmontmorillonitebytheconventionalcationexchangewith1MsolutionsofNaCl,CaCl2andCrCl3respectively(Kozaietal.,1996;Kozakietal.,1999;Juangetal.,2004).2.1.Startingmaterial,Fe(III)-montmorillonite

Fe(III)-montwasobtainedbycationexchangeofNa-montin0.4MFeCl3solution(pH~1.7).About10goftherawmontmorillonite(KunipiaF)wasaddedto1lof0.4MFeCl3solutionandstirredfor~24h.Thedispersionwasallowedtosettleovernightandafterdecantingthesupernatantsolution,themontmorillonitewasre-dispersedinfresh0.4MFeCl3solution,andthisprocesswasrepeatedtwice.Then,thedispersionwascentrifuged.InthecaseofFe-1,themontmorillonitewaswashedseveraltimeswithdistilledwater(~500mlofwaterper10gofclay)byfiltrationthrough0.2μmfilterpaper.FiltrationwasfoundtobemuchfasterandeasierthandialysisbecauseoftherestrictedswellingofFe(III)-mont,unlikeNa(I)-orCa(II)-mont.Fe-2waswashedwithdistilledwaterbycentrifugationtosqueezeoutmostoftheFeCl3solution,andthenwashedasinthecaseofFe-1.Duetothisdifferenceinthewashingstep,Fe-1containedhigheramountsofexternallyprecipitatedferricionsthaninFe-2,Table1.BoththesampleswerewasheduntilnoCl−wasdetectedbyAgNO3test.Theyweredriedinvacuumat25°Cforseveralhoursandthengroundtopowder.Thesampleswerestoredinvacuumdesiccatorforsubsequentuse.

ThetotalamountofadsorbedFe(III)inFe-1andFe-2wasdeterminedbyextractionwith50mMNa2EDTAsolutionat70°Cfor~24h(Table1).Thesample(≈0.5g)wasdispersedin100mlofthefreshreagentfor3timesandthecumulativeironreleasedwasconsideredasthetotaliron(Fetotal).

2.2.ConversionofFe(III)-toFe(II)-montmorillonite

ThesamplesofFe-1orFe-2werefirstdispersedinwaterfor~12hintheratioof1g/100ml.Ascorbicacidwasthenaddedandthedispersionwasstirredat~70°Cfor12hunderAr.Theascorbicacidconcentrationwasabout8and5mMrespectivelypereachgramofFe-1andFe-2basedonthecorrespondingamountsadsorbedFe(III).Themontmorillonitewasseparatedbycentrifugation,washedthoroughlywithdeoxyge-natedwater.ThecompleteremovalofascorbicacidwasensuredbytestingofsuccessivewashingsforFe(II)withferroinreagent(assumingferrous-ascorbatecomplexation).Theabsenceofascorbicacidwasalsotestedwithmixtureofiodineandstarchsolution.

Table1

Fecontentandspecificsurfacearea(S)ofFe(III)-andFe(II)-montsamples(MB—methyleneblue)Samples

Fein

MB

Sinm2/gmeq/100g

adsorbed(mol/g)

SMBSBETFe(III)-montFe-12266.4×10−450070Fe-2

1427.8×10−461047Fe(II)-montFe-1r1208.7×10−468045Fe-2r

97

9.2×10−4730

45

Fe(II)-montthusobtainedfromFe-1andFe-2wasdriedinvacuumat25°C,anddesignatedhereasFe-1randFe-2r(Table1).

Toevaluatethefeasibilityofsolid-statepreparation,1gofFe(III)-mont,(Fe-2)andascorbicacid(0.06g)weregroundinopenatmosphereatdifferentrelativehumidity(RH)conditions,20–80%.ThetimerequiredfortheconversionofallFe(III)toFe(II)couldbeseenfromthecolorchangei.e.,theinitialbrowncolorofthemixturebecamegrayishwhiteinafewminutesdependingonRH.Itwasfurtherestimatedbyperi-odicallydrawingaliquotofthemixtureandreactingwithasolidKSCNuntilnoredcolorwasseen,indicatingtheabsenceofFe(III).TheformationofFe(II)-montwasconfirmedbypowderX-raydiffractionandmagneticsusceptibilitymeasurementatRTusingasuperconductingquantuminterferencedevicemagnetometer.2.3.CharacterizationofFe(II)-montmorillonite

Thetotalamountofferrousiron(andferriciron,ifany)adsorbedonthemontmorillonitewasdeterminedbyextractionwithviz.,0.05MH2SO4,1MNH4Cl,and50mMNa2EDTA.Ineachcase,aknownamount(≈0.5g)ofthesamplewasdispersedin100mlofthereagentsfor~24hunderAr.ThisextractionprocedurewasrepeatedtwicebutnotwithH2SO4topreventthepossibledissolutionofthemontmorillonite.TheFe2+/Fetotalratiowasdeterminedfromtheironreleasedin0.05MH2SO4solution,whileotherreagentsyieldedFetotal.Fe(II)-montwascharacterizedbyX-rayfluorescence(XRF,Rhtarget),powderX-raydiffraction(XRD,Cu-Kαradiation),Fouriertransforminfrared(FTIR)(KBrpellets),and57FeMössbauerspectroscopyatRTusing57Co(Rh)source.

2.4.SpecificsurfaceareabymethyleneblueandN2-adsorption

Thespecificsurfaceareaofthesamples(Fe-1,Fe-2,Fe-1r,andFe-2r)wasdeterminedbymethyleneblue(MB)adsorption,andN2-adsorptionmethod(BET).

About0.5gofthesampleswasdispersedin100mlofwaterfor~24h.Then,MB(0.01M)wasaddeddrop-wise.Adsorptionwasmonitoredbytheso-calledspottest(Santamarinaetal.,2002).TheMBconcentrationattheendpointwasdeterminedbyaUV–Visiblespectrophotometeratλmaxof665nmafterequilibratingfor~24h.InthecaseofFe(II)-mont,theadsorptionwascarriedoutinAratmosphere.

FortheBETmeasurements,thesampleswerede-gassedatRTundervacuumformorethan24h.

2.5.StabilityofFe(II)-montmorillonite

About0.5gofthesample(Fe-1r)wasdispersedin200mlofwaterat(a)RTand(b)65°C,withArbubbling,butoccasionallyexposedtoairwhilesampling.Hencetheyaretreatedhereaspartiallydeoxygenatedconditions.About5mlofthisdispersionwasperiodicallycollectedandtreatedwith10mlof0.05MH2SO4for~24h,andthenFe2+/Fetotalratiowasdetermined.Forcomparison,Na-montwasalsotreatedsimilarlytoseeanystructuralironreleasedduetodissolution.

3.Resultsanddiscussion

3.1.Startingmaterials,Fe(III)-montmorillonite

BasedontheironextractedbyNa2EDTAsolution,Table1,totalamountofFe(III)inFe-1andFe-2wasabout2and1.25timestheCECoftheparentmontmorillonite(~113meq/100g).Theexcessironinthesesamplesispresumedtobeferric(hydr)oxidephase(Krishna-murtietal.,1998)duetotheprecipitationofFe(III)becauseofitslowsolubilityasthepHwasincreasedsuccessively(acidictoneutral)duringwashing.Also,hydrolysisandformationofoligomerse.g.,

FexOH(3yx−y)(KnightandSylva,1975)couldcontributetohigherFe(III)contents,butthisshouldbenegligiblysmallandthereforewasnottakenintoaccounthere.

Inthisstudy,theFe(III)-montmorillonitewasdriedatRTundervacuumbecauseairdryingwasfoundtobehighlytimeconsuming.Alternatively,itmaybeadvantageoustousefreezedrying.However,theeffectofdryingprocessontheexchangeableFe(III)ionsofthesampleneedstobeverified.DuringthepreparationofFe(III)-mont,thetotalinteractiontimeofthemontmorillonitewithFeCl3solutionwas~100h.Duringthistime,partialdissolutionofthemontmor-illonitecannotberuledout.Nevertheless,theSi/Alratio(asanalyzedforelementalcompositionbyXRF)oftheclaymineral,beforeandafterthecationexchangereaction,remainedthesame(≈3.3).ToavoidthedissolutionofclaymineralinacidicreagentslikeFeCl3,weproposetotestotherFe(III)reagents(ca.ferriccomplexeshavingintermediatestabilitysothatitcandissociatetoaidcationexchangewithNa-mont)infuturestudies.

34J.Manjanna/AppliedClayScience42(2008)32–38

Fig.1.WetanddriedNa-,Fe(III)-andFe(II)-mont.Upperrowphotos:freshlypreparedwetsamplesafterintensewashinginwater.Bottomrowphotos:samplesafterdryingatRT.Insets:sedimentsofdispersionusing2wt.%ofNa-andFe(III)-mont,and0.5wt.%ofFe(II)-mont.(Forcolorandclarity,thereaderisreferredtothewebversionofthisarticle.)

3.2.ConversionofFe(III)-toFe(II)-montmorillonite

ForconvertingFe(III)-toFe(II)-mont,ascorbicacid(vitaminC)wasusedasareducingagent.Thermodynamically,ascorbicacidisamildreducingagent(E0=+0.41Vvs.SHE)butcanbeakineticallystrongerreducingagentasrevealedinthereductivedissolutionofvariousFe-oxides(Manjannaetal.,2001;ManjannaandVenkateswaran,2002).Ascorbicacidishighlywatersolubleandanenvironmentallybenignreagent.DuringtheFe(III)-montreduction,slightlymorethanthestoichiometricamountofascorbicacidwasaddedtothedispersionofFe-1orFe-2underAratmosphere.Thereductionwascompletein~4hinboththesamples.Nevertheless,toachievethecompletedissolutionoftheexcessFe-oxidephase(especiallyinFe-1),thereactionwascontinuedfor~12h.ThisreactiontimecouldbeshorteneddependingontheFecontentofstartingmaterials.TheamorphousFe-oxidephaseformedatRTcanbeeasilydissolvedatthistemperature(Josephetal.,1998)unlikethesinteredoxides.Thusinthisstudythetemperaturewasraisedto~70°CtoaidthereductionofallFe(III)ions.TheconversionofFe(III)-toFe(II)-montwasvisiblyindicatedbythecolorchangefrombrowntograyishwhite,whichwassubstantiatedbymonitoringtheFe2+/Fetotalratio.TheconversionofFe(III)-toFe(II)-montmaybeshownas:2Fe(III)-mont+(H2A↔2H++2e−+A)→2Fe(II)-mont+A,whereH2AandAareascorbicacidanddehydroascorbicacidrespectively.ThereleasedprotonsmayreactwithhydroxylionsofhydrolyzedFe(III)and/orferric(hydr)oxide.Thus,pHofthedispersion,justafteraddingascorbicacidandattheendofthetreatmentwas~2.5and3.7,showingtheconsumptionofH+.ItwasnotexpectedthattheprotonsreplacethestronglyadsorbedFe(II)ionsontheclaymineralsurface.InthecaseofFe-2r(Table1)theferrousionsaccountonly~86%oftheCECoftheparentclayNa-mont(113meq/100g).Undersuchsituation,H+ionscouldcompensatethemissingamountofFe(II)ions,aschargebalancingcations.ThepresenceofBrønstedacidsitesinsuchsamples(Fe-2r)mayserveasadirectevidenceforthepresenceofH+ions.Attheendofthereduction,theresidualascorbicacidwaswashedoffwithexcesswaterbycentrifugation(pHattheendwas~5.5).AsrevealedfromthereactionofFe(II)-montwithCr(VI)solution(Manjannaetal.,2006),ascorbicacidseemedtobeaweakreductanttowardsstructuraliron.Thusthereductantusedinthisstudyhadseveraladvantages.

AsstatedaboveinSection3.1,thedissolutionoftheclaymineralduringthepreparationisanimportantissue.AscheckedbyXRF,theSi/AlratioofFe-1randFe-2rwasclosetothatofthestartingmaterial

(≈3.3)indicatingthatmontmorillonitedissolutionwasnegligiblysmall.Inagreementwiththisobservation,thedissolutionrateofsmectiteatpHof3wasreportedtobeverysmall,intheorderof10−13molm−2s−1(cf.Huertasetal.,2001).Accordingtothisrate,dissolutionduringtheinteractionwithascorbicacidhereinisestimatedtobeb10−3wt.%.

InFig.1,theappearancesofstartingmaterialsandtheendproducts(i.e.,Na(I)-→Fe(III)-→Fe(II)-mont),inwetanddryforms,arecom-pared.UnlikeNa(I)-mont,Fe(III)-andFe(II)-montdonotformagelinwetcondition(upperrowphotos)becauseofrestrictedswelling(insetphotosinbottomrow).ThecolorofFe(II)-montwasgrayishwhiteinwet/beforedryingandwhitewhendriedorre-dispersedinwater.ThedegreeofswellingdecreasesintheorderNa(I)-mont≫Fe(II)-montNFe(III)-mont.Suchadifferenceinswellingisanticipatedtoaffectthe

Fig.2.PowderXRDpatternsofNa-,Ca-,Cr(III)-,Fe(II)-andFe(III)-mont.

J.Manjanna/AppliedClayScience42(2008)32–3835

hydraulicconductivity,anddiffusion/adsorptionofradionuclidesetc.,whichareveryimportantcharacteristicsinviewoftheclaymineralalteration,Fe/clayinterface,inthegeologicaldisposalofnuclearwaste.

3.3.AdsorbedironofFe(II)-montmorillonite

TheFe2+/FetotalratioofFeextractedwith0.05MH2SO4fromfreshlypreparedsamples(Fe-1randFe-2r)wasclosetounity.Itmeansmostoftheinterlayerironisindivalentstate.Na-montreactedwith0.05MH2SO4releasedb1%ofthatinFe(II)-mont.Hencenegligiblysmallamountofclaymineraldissolutionoccurredduringthistreatment.Thus,theFe2+/FetotalratioascribedhereforinterlayerironofFe(II)-montwasnotinfluencedtoanyappreciableextentbytheclaymineraldissolution.However,long-timecontactoftheclaymineralisexpectedtoinfluencethisratio.

BasedonFetotalextractedwithNH4Cl,theCECofFe-1randFe-2rwereabout120and97meq/100g(Table1).AlmostthesameamountofFetotalwasobtainedinNa2EDTA.IftherewasanyundissolvedFe-oxidephase,FetotalinNa2EDTAwasexpectedtobehigherthaninNH4Cl.ThefactthatFetotalremainedthesameinboththereagentsindicatestheabsenceofFe-oxidephasesinthepreparedsamples.

ThedifferenceinFetotalofFe-1randFe-2rcanbeascribedtothedifferentamountsofFeinthestartingmaterials,Fe(III)-mont(Table1).TheexcessamountofadsorbedFepresentinthestartingmaterial,Fe-1,wasdissolvedcompletelyduringascorbicacidtreat-ment.Thus,additionalamountsofferrousions(probablyasferrousascorbateduetoreductivecomplexingmechanism)wereavailableinsolution.TheseFe(II)ionscanbeeasilyadsorbedontheclaymineralsurfacetocompensatethechargedeficitcreatedbythereduction

Fig.3.FTIRspectraofNa-,Fe(II)-,andFe(III)-mont.

Fig.4.57FeMössbauerspectraofNa-,Fe(III)-,andFe(II)-montatRT(Fe-2andFe-2rareshownastypicalsamples).

ofinterlayerFe(III)toFe(II)ions.InthecaseofFe-2r,theCECintermsofFe(II)ionsis97meq/100gwhichis~68%ofstartingmaterial(Fe-2;CEC142meq/100g)andcorrelateswiththechargereduction,Fe3+→Fe2+.Basedonthiscorrelation,Fe-2rprobablycontainssomeprotonsascompensatingcations.Thus,aslightlyexcessamountofFe(III)inthestartingmaterial(maybeasamorphousFe-oxide)isadvantageous.

Fig.5.Magneticsusceptibility(χ)ofthesolidmixtureofFe(III)-mont(Fe-2)andascorbicacid(RH≈80%)atdifferenttimeintervals.InsetisthemagnetizationoftheFe-2sampleasafunctionofappliedfield.

36J.Manjanna/AppliedClayScience42(2008)32–38

Fig.6.StabilityofFe(II)-montunderdifferentconditions.

3.4.XRD,FTIRandMössbauerspectra

Thebasalspacing,d001,indicatingthedistancebetweentheadjacentlayersisknowntobeaffectedbythecharge,sizeandhydrationbehaviorofinterlayercations(WatanabeandSato,1998).InFig.2,thepowderXRDpatternsofFe(II)-montiscomparedwiththeNa(I)-,Ca(II)-,Fe(III)-andCr(III)-montatarelativehumidity(RH)of40%.Clearly,XRDpatternsdisplaybothbasal/symmetricreflections(00l,whosepositionvarywithseparationbetweenlayers)andgeneralhkreflections(twodimensional),asaconsequenceoftheturbostraticstackingoflayers.Basalspacing,d001,ofFe(II)-montdeterminedhere(14.7Å)isinagreementwiththereportedvalue(Manjannaetal.,2007),andalsowithatypicaldivalentclay,Ca(II)-mont,showninFig.2.Further,d001ofFe(II)andFe(III)-montwerefoundtobesimilarintheentireRHrange(notshownhere).Thisisprobablyduetothecompensationofionicchargevis-à-visionicradius(Fe2+is0.77ÅandFe3+is0.60Å)onthehydrationbehavior.ThesharpbasalreflectionsintheXRDpatternalsoconfirmthecrystallinityofthesamples.

FTIRspectraofFe(II)-andFe(III)-montarecomparedwithNa(I)-montasshowninFig.3.AmongtheFe-montsamples,theplotsshownhereareforFe-1andFe-1r(firstset),however,itwasseenthatthereisnodifferencebetweenthecorrespondingsamplesoftheotherseti.e.,Fe-2andFe-2r.Thus,allthesamplesstudiedhereshowthechar-acteristicbandsofmontmorilloniteclay,forinstance,theSi–O–Albendingvibration~520cm−1,hydroxylbendingvibrationsofAl–Mg–

OH(845cm−1)andAl2OH(918cm−1).However,thestretchingvi-brationsofSi–Ogroup~1038cm−1isslightlybroadenedinthecaseofFe-montwhencomparedtoNa(I)-mont.Thisisprobablyduetotheinteractionofinterlayer/adsorbedFewiththetetrahedralsilicatesheet.Theplausibleclaydissolutionand/orsitedistortioncannotberuledoutastheclaywasinitiallyinteractedwithacidicsolutionofFeCl3andlaterinascorbicacid,andhencethedetailedstudiesarerequired.TheFTIRspectrainthehigherenergyregion(Fig.3)areshownhereforFe(II)-,andFe(III)-mont.Thevibrationbandscorrespondingtoadsorbedwater(1630cm−1),interlayerwater(3426cm−1)andstruc-turalOHgroups(3625cm−1)haveremainedmoreorlessthesameasthatobservedfortheparentclay,Na(I)-mont.

MössbauerspectroscopyishighlysuitedforthestudyofclayswhoseindustrialusesdependontheFecontent.Nevertheless,itshouldbeviewedincombinationwithdatafromothertechniques.Fig.4showstheMössbauertransmissionspectraatRTforNa(I)-,Fe(III)-andFe(II)-mont.AmongtheFe-montsamples,Fe-2andFe-1rareshownhereasatypicalcaseofferric-andferrous-mont.MössbauerfeatureinFig.4isduetobothinterlayerandstructuraliron.Thestructuraliron(Fe2+andFe3+)ispresentintheoctahedralsheet.TherearetwositesforFewithrespecttohydroxylsintheoctahedral,cis(M2)andtrans(M1).InMössbauerspectraofsmectiteclays(Stucki,2006;Murad,1998),thecentralquadrupoledoubletatabout0.3mm/sisascribedtothestructuralFe3+andtheothersetofdoublet/scomprisingtheadditionalfeatureatabout2.2mm/sisforFe2+.InfactthereisnodetailedinformationintheliteratureabouttheinterlayerFe(andthusthedatapointsinFig.4arenotfittedhere).ItispresumedthattheinterlayerFepresentinthecaseofFe(III)-andFe(II)-montalsogivesadditionalset/sofdoublets.AscanbeseeninFig.4,theinterlayerFe(III)wasseentoenhancethecentraldoublet/swhiletheinterlayerFe(II)wasfoundtoenhancetheadditionalfeatureatabout2.2mm/s.InthecaseofFe(III)-mont,althoughtheassociatedFe-oxidephasewaspresent,itcouldnotbedetectedatRTbecauseofitspoorly-crystallineandmagneticallynotorderedphase.Nevertheless,suchoxidephaseisexpectedtogive6-linepatternatcryogenictemperatures(Stucki,2006).

3.5.Specificsurfaceareabymethyleneblue,andBETmethod

UV–VisiblespectraoftheMBsolution(0.01M)usedhereshowedanabsorptionmaximumat665nm,whichischaracteristicofMBmonomers.MB–claymineralinteractionisextremelysensitiveforprobingthesurfacepropertiesofsmectitesinaqueousdispersions.InthecaseofNa-mont,theadsorptioncapacitywas~10.3mol/g.Basedonanareaof130Å2permoleculeofMB(Santamarinaetal.,2002),thetotalspecificsurfacearea(internal+external)oftheNa-montwas~810m2/g.TheamountofMBadsorbed(inmolespergramsofclay)

Fig.7.ColorchangesduetoairoxidationofpowderedFe(II)-mont(Fe-1r)atRTandRHb40%.A—(whitecolor):freshlypreparedsample,B—(lightbrown):50%oxidizedsample,andC—(darkbrown):completelyoxidizedsample.(Forinterpretationofthereferencestocolorinthisfigurelegend,thereaderisreferredtothewebversionofthisarticle.)

J.Manjanna/AppliedClayScience42(2008)32–3837

bytheFe-montsamples,andthecorrespondingspecificsurfacearea(SMB)aregiveninTable1.TheadsorptioncapacityofFe-mont,ingeneral,wassmallerthanthatobservedforNa-mont,anddecreasedwithincreasingFecontentofthesamples.TheadsorptionwashigherforFe(II)-mont(Fe-1randFe-2r)whencomparedtothestartingmaterials(Fe-1andFe-2).TheseresultsshowthatMBmoleculecannotreplacealloftheinterlayerironspecies.ThisispresumablyduetohighaffinityofFe(III)ionsfortheclaysurface.Also,thereleasedFe(III)ionsareexpectedtoprecipitateduetotheirlowsolubilityandre-adsorbontheclay,thereby,MBadsorptionisaffected.Thus,inthecaseofFe-mont,theadsorptioncapacityofMBcouldnotbedirectlyrelatedtothelayerchargeofclayminerals,unlikeinNa-mont.

ThespecificsurfaceareameasuredbyN2-adsorption(SBET)forFe-2,Fe-1randFe-2rwas~45m2/g,and~70m2/gforFe-1(Table1).Gasadsorptiongivesonlytheexternalsurfaceareaoftheselayeredmaterials.Inthisstudy,thesamplesweredriedatRTi.e.,notexposedtoelevatedtemperaturesinviewoftheirstability(toavoidtheoxidationinthecaseofFe(II)-montorthepossibleformationofseparateFe-oxidephaseinthecaseofFe(III)-mont).Thus,thelowspecificsurfaceareaobtainedhereisattributedtothepoorlydriedsamples.TherelativelyhigherSBETofFe-1maybeduetothepresenceofassociatedFe-oxidephasesastheFecontentis~2timestheCECoftheparentmontmorillonite.

3.6.Solid-statereactionofFe(III)-intoFe(II)-montmorillonite

Duringthesolid-statereaction,therewasagradualdisappearanceofinitialbrowncolorofFe(III)-mont.ThereactionwaseffectiveonlyathigherRHcondition(N70%),andcompletereductionwasachievedin~6h.AtlowerRH,thereactionwasincomplete,forinstance,~30%reductionwasobtainedwhenRH=60%,andreductionwasnegligiblysmallatRHb40%.Thus,RHof70%andhigherwasrequiredtocom-pletethereductionreactioninreasonabletime.TheFe(III)ionsin-tercalatedinthestartingmaterial(Fe-2)weresufficientlyhydrated(athigherRH)tokeeptheinterlayerspacesexpanded(d001=15.5ÅatRH≈70%)whichpromotethediffusionofascorbicacidbetweenthelayers.TheformationofFe(II)-montwasindicatedbypowderXRD(d001=14.7Å)andmagneticsusceptibilitymeasurementatRT(Fig.5).Therewasabouta5-foldincreaseoftheparamagneticsusceptibilityduringthereduction.TheabsenceofFe(III)ionsintheendproductswasalsocheckedbythechemicaltestwithKSCN.Thesolid-stateion-exchangeandintercalationreactionsarewellknownformontmor-illonite(Crockeretal.,1992;Ogawaetal.,1995).3.7.StabilityofFe(II)-montmorillonite

TheFe2+/FetotalratioofFereleasedfromFe(II)-montunderdifferentconditionswasmeasuredduringsevenweeks(Fig.6).ThedecreaseofFe2+/FetotalindicatedtheoxidationofinterlayerFe(II)ions.ThestabilityofFe(II)-mont,includingourpreviousdata,(a)and(d),ofthesamplepreparedbytheFe-NTAmethod(Manjannaetal.,2007),decreasedasdrypowderNinwater/Ar/RTNinwater/air/RT≫inwater/Ar/65°C.Thereactionsduringoxidationmaybewrittenas:Fe2+→Fe3++e−;H2O+0.5O2+2e−→2OH−and/or2H2O+2e−→H2+2OH−.Atelevatedtemperature,thesereactionsarekineticallyfavored,thereforehigheroxidationwasobservedat65°C.Detailedstudiesunderstrictoxygencontrolareneededtounderstandthemechanism.

Clearly,Fe(II)-montisgraduallyoxidizedi.e.,notrapidlyasanticipated.ThereasonablestabilityofinterlayerFe(II)ionsmaybeattributedtotheirprimaryroleaschargebalancingions.Therefore,theoxidationprocessseemstobekineticallyhindered,whereasinstrongoxidizingmediumsuchasaqueousCr(VI),theoxidationwasrapid(Manjannaetal.,2006)atRT.WhenthedrypowderofFe(II)-montwasexposedto~100°Cinair,theoxidationoccurredinlessthan10min.Forlongerstorage,itisessentialtomaintainadeoxygenated,lowRH(b40%)conditionatRTorbelow.

Thecolorchangeoftheoxidizedsamples(white→lightbrown→darkbrown)wasvisibleafterabout30%oxidation(Fig.7).FreshlypreparedFe-1r(Fe2+/Fetotal≈1)waswhite(A)whichturnedtolightbrowncolor(B)duetoairoxidation(Fe2+/Fetotal≈0.5,inabout50daysatRTwithRHb40%).Itwasdarkbrown(C)whencompletelyoxidized.Aftergradualoxidationto~50%,furtheroxidationwasfaster.Asimilarcolorationwasseenduringoxidationinwater.Thus,theextentofoxidationcouldbefollowedbyvisualinspection.

Duetooxidation,about33%oftheFeisexpectedtochangeitsstatusfromchargecompensatingcationstoneutralspecies.Asindicatedbythecolorchange,suchFeseemstoprecipitateontheclaymineralsurfaceasamorphousiron(hydr)oxide.Basedonthestabilitydata,itisclearthattheFe(II)ionsintheinterlayerspaceofmontmorillonitearereasonablystableunlikeitwasanticipated.HenceitwaspossibletoprepareandhandletheFe(II)-montunderambientatmosphericconditions.4.Conclusions

PreparationofFe(II)-montmorillonitefromFe(III)-montthroughthereductionofinterlayer/adsorbedFe(III)byascorbicacidwasfoundtobehighlyconvenient.Thismethodmaybeadoptedforpreparationinlargeamounts.ThepresenceofexcessFe-(hydr)oxidephasesinthestartingmaterialwasratheradvantageoustoobtainstoichiometricFe(II)-mont.Thefeasibilityofsolidstatereductionwasalsodemonstrated.ThestabilityofFe(II)-montunderdifferentconditionsdecreasedas:drypowderNinwater/Ar/RTNinwater/air/RT≫inwater/Ar/65°C.Forlong-timepreservation,Fe(II)-montmustbestoredindeoxygenatedand/orlowrelativehumidity(b40%)conditionatRTorbelow.Acknowledgements

FinancialsupportforthisworkwasprovidedbytheJapanSocietyforthePromotionofScience.ThanksareduetoProf.T.KozakiandProf.S.Sato,HokkaidoUniversity,Japan.References

Basset,J.,Denny,R.C.,Jeffery,G.H.,Mendham,J.,1978.Vogel'sTextbookofQuantitative

InorganicAnalysis,fourthed.ELBS/Longman,London.

Carlson,L.,Karnland,O.,Oversby,V.M.,Rance,A.P.,Smart,N.R.,Snellman,M.,Vahanen,

M.,Werme,L.O.,2007.Experimentalstudiesoftheinteractionsbetweenanaerobicallycorrodingironandbentonite.Phys.Chem.Earth32,334–345.

Charlet,L.,Tournassat,C.,2005.Fe(II)–Na(I)–Ca(II)cationexchangeonmontmorillonite

inchloridemedium:evidenceforpreferentialclayadsorptionofchloride–metalionpairsinseawater.Aquat.Geochem.11,115–137.

Crocker,M.,Herold,R.H.M.,Emeis,C.A.,Krijger,M.,1992.Preparationofacidicformsof

montmorilloniteclayviasolid-stateion-exchangereactions.Catal.Letters15,339–345.

Dorado,F.,deLucas,A.,García,P.B.,Romero,A.,Valverde,J.L.,2006.Copperion-exchangedandimpregnatedFe-pillaredclays:studyoftheinfluenceofthesynthesisconditionsontheactivityfortheselectivecatalyticreductionofNOwithC3H6.Appl.Catal.,A.305,1–196.

Guillaume,D.,Neaman,A.,Cathelineau,M.,Mosser-Ruck,R.,Peiffert,C.,Abdelmoula,M.,

Dubessy,J.,Villieras,F.,Michau,N.,2004.Experimentalstudyofthetransformationofsmectiteat80and300°CinthepresenceofFeoxides.ClayMiner.39,17–34.Herrera,R.,Peech,M.,1970.Reactionofmontmorillonitewithiron(III).Proc.SoilSci.

Soc.Am.34,740–742.

Huertas,F.J.,Caballero,E.,JimenezdeCisneros,C.,Huertas,F.,Linares,J.,2001.Kineticsof

montmorillonitedissolutioningraniticsolutions.Appl.Geochem.16,397–407.Izumi,Y.,Masih,D.,Aika,K.I.,Seida,Y.,2005.Characterizationofintercalatediron(III)

nanoparticlesandoxidativeadsorptionofarseniteonthemmonitoredbyX-rayabsorptionfinestructurecombinedwithfluorescencespectrometry.J.Phys.Chem.,B.109,3227–3232.

JapanNuclearCycleDevelopmentInstitution(JNC),2000.H12:SafetyAssessmentof

theGeologicalDisposalSystem,JAEA,Orai,Japan.

Joseph,S.,Venkateswaran,G.,Moorthy,P.N.,1998.Influenceofthermalhistoryofiron

oxidesontheirdissolutionbehaviourincitricacid–EDTA–ascorbicacidmixture.IndianJ.Chem.Technol.5,222–226.

Juang,R.S.,Lin,S.H.,Huang,F.C.,Cheng,C.H.,2004.StructuralstudiesofNa-montmorilloniteexchangedwithFe2+,Cr3+,andTi4+byN2adsorptionandEXAFS.J.ColloidInterfaceSci.274,337–340.

Kamei,G.,Oda,C.,Mitsui,S.,Shibata,M.,Shinozaki,T.,1999.Fe(II)–Naionexchangeat

interlayersofsmectite:adsorption–desorptionexperimentsandanaturalanalo-gue.Eng.Geol.,15–20.

38J.Manjanna/AppliedClayScience42(2008)32–38

Kantam,M.L.,Kavita,B.,Figueras,F.,1998.Oxidativecouplingof2-naphtholscatalysed

byclays-supportedironcatalysts.Catal.Letters51,113–115.

Kloprogge,J.T.,1998.Synthesisofsmectitesandporouspillaredclaycatalysts:areview.

J.PorousMater.5,5–41.

Knight,R.J.,Sylva,R.N.,1975.Spectrophotometricinvestigationofiron(III)hydrolysisin

lightandheavywaterat25°C.J.Inorg.Nucl.Chem.37,779–783.

Komadel,P.,2003.Chemicallymodifiedsmectites.ClayMiner.38,127–138.

Komadel,P.,Madejova,J.,Laird,D.A.,Xia,Y.,Stucki,J.W.,2000.ReductionofFe(III)in

griffithite.ClayMiner.35,625–634.

Kozai,N.,Ohnuki,T.,Muraoka,S.,1993.Sorptioncharacteristicsofneptuniumby

sodium-smectite.J.Nucl.Sci.Technol.30,1153–1159.

Kozai,N.,Ohnuki,T.,Matsumoto,J.,Banba,T.,Ito,Y.,1996.Astudyofspecificsorptionof

neptunium(V)onsmectiteinlowpHsolution.Radiochim.Acta75,149–158.

Kozai,N.,Adachi,Y.,Kawamura,S.,Inada,K.,Kozaki,T.,Sato,S.,Ohashi,H.,Ohnuki,T.,

Banba,T.,2001.CharacterizationofFe-montmorillonite:asimulantofbuffermaterialsaccommodatingoverpackcorrosionproduct.J.Nucl.Sci.Technol.38,1141–1143.

Kozai,N.,Inada,K.,Adachi,Y.,Kawamura,S.,Kozaki,T.,Sato,S.,Ohnuki,T.,Sakai,T.,Sato,

T.,Oikawa,M.,2007.CharacterizationofhomoionicFe2+-typemontmorillonite:potentialchemicalspeciesofironcontaminant.J.SolidStateChem.180,2279–22.Kozaki,T.,Sato,H.,Sato,S.,Ohashi,H.,1999.Diffusionmechanismofcesiumionsin

compactedmontmorillonite.Eng.Geol.,223–230.

Krishnamurti,G.S.R.,Violante,A.,Huang,P.M.,1998.Influenceofmontmorilloniteon

Fe(II)oxidationproducts.ClayMiner.33,205–212.

Lee,K.,Kostka,J.E.,Stucki,J.W.,2006.ComparisonsofstructuralFereductionin

smectitesbybacteriaanddithionite:aninfraredspectroscopicstudy.ClaysClayMiner.,195–208.

Lenoble,V.,Bouras,O.,Deluchat,V.,Serpaud,B.,Bollinger,J.-C.,2002.Arsenicadsorption

ontopillaredclaysandironoxides.J.ColloidInterfaceSci.255,52–58.

Manjanna,J.,Venkateswaran,G.,2002.Preparationandkineticconsiderationsforthe

dissolutionofCr-substitutedironoxidesinreductive-complexingformulations.Can.J.Chem.Eng.80,882–6.

Manjanna,J.,Venkateswaran,G.,Sherigara,B.S.,Nayak,P.V.,2001.Dissolutionstudiesof

chromiumsubstitutedironoxidesinreductive-complexingagentmixtures.Hydrometallurgy60,155–165.

Manjanna,J.,Kozaki,T.,Kozai,N.,Sato,S.,2006.UseofFe(II)-montmorilloniteforthe

reductionofaqueousCr(VI).19thProc.Intern.MineralogicalAssociation,Kobe,Japan,p.270.

Manjanna,J.,Kozaki,T.,Kozai,N.,Sato,S.,2007.AnewmethodforFe(II)-montmorillonitepreparationusingFe(II)-nitrilotriacetatecomplex.J.Nucl.Sci.Technol.44,929–932.

Masih,D.,Izumi,Y.,Aika,K.,Seida,Y.,2007.Optimizationofanironintercalated

montmorillonitepreparationfortheremovalofarsenicatlowconcentrations.Eng.LifeSci.7,52–60.

Murad,E.,1998.Claysandclayminerals:whatcanMössbauerspectroscopydotohelp

understandthem?.Hyp.Interact.117,39–70.

Ogawa,M.,Hagiwara,A.,Handa,T.,Kato,C.,Kuroda,K.,1995.Solid-stateionexchange

reactionsbetweenhomoionic-montmorillonitesandorganoammoniumsalts.J.PorousMater.1,85–.

Palinko,I.,Lazar,K.,Kiricsi,I.,1997.Cationicmixedlayerclays:infraredandMössbauer

characteristicsofthepillaringagentsandthepillaredstructuresinFe,AlandCr,Alpillaredmontmorillonites.J.Mol.Struct.410–411,7–550.

Santamarina,J.C.,Klein,K.A.,Wang,Y.H.,Prencke,E.,2002.Specificsurface:deter-minationandrelevance.Can.Geotech.J.39,233–241.

Stucki,J.W.,1988.In:Stucki,J.W.,Goodman,B.A.,Schwertmann,U.D.(Eds.),IroninSoils

andClayMinerals.Reidel,Dordrecht,TheNetherlands.

Stucki,J.W.,2006.HandbookofClayScience,Developments.In:Bergaya,F.,Theng,B.K.

G.,Lagaly,G.(Eds.),ClayScience,vol.1.Elsevier,Amsterdam.

Stucki,J.W.,Golden,D.C.,Roth,C.B.,1984.Preparationandhandlingofdithionite

reducedsmectitesuspensions.ClaysClayMiner.32,191–197.

Tabet,D.,Saidi,M.,Houari,M.,Pichat,P.,Khalaf,H.,2006.Fe-pillaredclayasaFenton-typeheterogeneouscatalystforcinnamicaciddegradation.J.Environ.Manag.80,342–346.

Valverde,J.L.,Romero,A.,Romero,R.,Garia,P.B.,Luz,S.M.,Asencio,I.,2005.Preparation

andcharacterizationofFe-PILCs:influenceofthesynthesisparameters.ClaysClayMiner.53,613–621.

Watanabe,T.,Sato,T.,1998.Expansioncharacteristicsofmontmorilloniteandsaponite

undervariousrelativehumidityconditions.ClaySci.7,129–138.

Wilson,J.,Cressey,G.,Cressey,B.,Cuadros,J.,Ragnarsdottir,K.V.,2006.Theeffectofiron

onmontmorillonitestability:(II)experimentalinvestigation.Geochim.Cosmo-chim.Acta70,323–336.

Yamanaka,S.,Hattori,M.,1988.Ironoxidepillaredclay.Catal.Today2,261–270.

Zurita,M.J.P.,Vitale,G.,DeGoldwasser,M.R.,Rojas,D.,Garcia,J.J.,1996.Fe-pillaredclays:

acombinationofzeoliteshapeselectivityandironactivityintheCOhydrogenationreaction.J.Mol.Catal.,AChem.107,175–184.