AnXRDstudyoftheeffectoftheSiO2Na2Oratioonthealkali

AnXRDstudyoftheeffectoftheSiO2/Na2OratioonthealkaliactivationofflyashM.Criadoa,A.Fernández-Jiméneza,⁎,A.G.delaTorreb,M.A.G.Arandab,A.PalomoaaEduardoTorrojaInstitute(CSIC),c/SerranoGalvache,no.4,28033Madrid,SpainbDepartmentofInorganicChemistry,UniversityofMálaga,CampusTeationss/n,29071Málaga,SpainReceived13November2006;accepted26January2007AbstractSolublesilicahasaverysignificanteffectonthemicrostructuralandmechanicaldevelopmentofthecementitiousmaterialsproducedasaresultofthealkaliactivationofflyash.Inthisstudy,fourdifferentalkalinesolutionswithdifferentsolublesilicacontentswereusedtoactivateflyash.Theprimaryreactionproductwasasodiumaluminosilicategel,whiledifferenttypesofzeolitesappearedasminorityphases.Thepercentageandcompositionofthesereactionproductswerefoundtodependonboththesolublesilicacontentpresentintheactivatingsolutionsandthethermalcuringtime.Inaddition,theamountofgelwasobservedtohaveadecisiveeffectonthemechanicalstrengthdevelopinginthematerial.©2007ElsevierLtd.Allrightsreserved.Keywords:Flyash;Rietveldmethod;Silicate–sodiumoxideratio1.IntroductionIn1959,V.D.Glukhovshy[1]suggestedthattheconstruc-tionindustrycouldbenefitfromtheuseofwhathecalled“soilcements”,newcementitiousproductsresultingfromattackingcertain(naturalorindustrialwaste)aluminosiliceousmaterialswithalkalinesaltsolutions.Theflyashproducedinsteampowerplantsisonesuchaluminosilicatesuitableforalkaliactivation[2].Althoughthemechanismsthatregulatethereactionsbetweenalkalineactivatorandflyasharenotfullyunderstood,themodelsthataregraduallybeingdevelopedaffordafairlypreciseviewofwhatactuallytakesplace[3,4].Theactivationrateandchemicalcompositionofthereactionproductsdependonfactorssuchasashparticlesizeandchemicalcomposition,typeandconcentrationoftheactivator,etc.Nonetheless,themechanismsthatcontrolthegeneralactivationprocessareindependentofthevaluesassumedbythesevariablesatanygiventime.Theabovefindingsinformedthedecisiontoconductthepresentstudyontheeffectofthesolublesilicacontentintheactivatingsolutiononthereactionrateofthe“ash-activator”systemandthenatureofthereactionproducts.Priorresearch[5–7]hadshownthattheH2O/SiO2andOH−/SiO2ratioshaveaconsiderableimpactonthe“molecular”or“polymeric”speciespresentinthereactionmixandtherateofuptakeofthesespeciesintothethree-dimensionalstructureofthetectosilicates.Themainreactionproductofthealkaliactivationofflyashisasodiumaluminosilicategel.Thisamorphousmaterialhashithertobeenverydifficulttocharacterize,buttheunderstand-ingofitsstructureandcompositionisgrowingsteadilymoreprecisewiththeincreasinglypowerfulanalyticaltechniquesnowdeployed.AccordingtoNMRfindings,itdoes,however,exhibitshort-rangeorder,withathree-dimensionalstructureinwhichsiliconisfoundinavarietyofenvironments,Q4(nAl):onthisbasis,thegelcanberegardedtobea“zeoliteprecursor”[4].Rietveldmethodologyconsistsonthecomparisonbetweenthemeasuredandcalculatedpowderdiffractionpatterns.Theanalysisofthewholepatternminimisestheinaccuraciesderivedofsystematicerrorsinsidetherawdatasuchaspeaksoverlap,preferredorientation(s),peakbroadeningandlackofapurestandard(s).TheRietveldmethodisapowerfultooltosuccessfullycarryoutquantitativephaseanalysis(QPA)ofCementandConcreteResearch37(2007)671–679⁎Correspondingauthor.E-mailaddress:[e-mail](A.Fernández-Jiménez).0008-8846/$-seefrontmatter©2007ElsevierLtd.Allrightsreserved.doi:10.1016/j.cemconres.2007.01.013crystallinecomplexsystems[8–13]althoughthecrystalstructuresofallcrystallinephasesmustbeknown.Thismethodologyisalsosuitedtofollowstructuralchanges[14,15]duetodifferentelementalcomposition.Moreover,theRietveldmethodologyhasalsobeenextendedtoindirectlydeterminetheamorphouscontentinagivencrystallinesamplebyaddingasuitablestandard[16].Theaimofthepresentstudywastoexploretheeffectofthesolublesilicacontentinaseriesofactivatingsolutionsonthemicrostructuraldevelopmentof“ash-activator”systemreactionproducts,andtoidentifyandquantifysuchproductsandtheirimpactonmechanicaldevelopmentinthematerial.2.Experimental2.1.CharacterizationofinitialmaterialsAtypeF(asdefinedinASTMstandardC6128–03)flyashfromtheCompostillasteampowerplantinSpain,consistingprimarilyofSiO2andAl2O3wasusedinthepresentstudy.Thechemicalcompositionoftheash(determinedasdescribedinSpanishstandardUNE80–230–99)andpercentageofreactivesilica(establishedfollowingtheproceduresetoutinSpanishstandardUNE80–225–93)areshowninTables1and2,respectively.Ninetytwopercentoftheparticlesinthisashwereb45μm[17].Theashwasactivatedwithaseriesofalkalinesolutions,allwithapracticallyconstantsodiumoxidecontent(≈8%),butwithvaryingproportionsofsolublesilica.Theproductsusedtopreparethesolutionswerelaboratoryreagents:ACS-ISO98%pureNaOHpelletssuppliedbyPanreacS.A.andsodiumsilicatewithadensityof1.38g/ccwiththefollowingcomposition:8.2%Na2O;27%SiO2and64.8%H2O.ThechemicalcompositionandcertainotherpropertiesofeachsolutionaregiveninTable3,whichalsoshowsthedifferencesinboththesilicaandthewatercontentachievedbymixingsodiumsilicateandsodiumhydroxideindifferentproportions.2.2.Methodology2.2.1.Alkaliactivationo