• Aucun résultat trouvé

Calcined Resin Microsphere Pelletization (CRMP) a novel process for sintered metallic oxide pellets synthesis.

N/A
N/A
Protected

Academic year: 2021

Partager "Calcined Resin Microsphere Pelletization (CRMP) a novel process for sintered metallic oxide pellets synthesis."

Copied!
12
0
0

Texte intégral

(1)

HAL Id: cea-02349240

https://hal-cea.archives-ouvertes.fr/cea-02349240

Submitted on 20 Nov 2019

HAL

is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire

HAL, est

destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

synthesis.

E. Remy, S. Picart, S. Grandjean, T. Delahaye, N. Herlet, P. Allegri, O.

Dugne, Renaud Podor, Nicolas Clavier, P. Blanchart, et al.

To cite this version:

E. Remy, S. Picart, S. Grandjean, T. Delahaye, N. Herlet, et al.. Calcined Resin Microsphere Pelleti-

zation (CRMP) a novel process for sintered metallic oxide pellets synthesis.. Journal of the European

Ceramic Society, Elsevier, 2012, 32 (12), pp.3199-3209. �10.1016/j.jeurceramsoc.2012.04.011�. �cea-

02349240�

(2)

JournaloftheEuropeanCeramicSocietyxxx(2012)xxx–xxx

Calcined resin microsphere pelletization (CRMP): A novel process for sintered metallic oxide pellets

E. Remy

a,∗

, S. Picart

a

, S. Grandjean

a

, T. Delahaye

b

, N. Herlet

b

, P. Allegri

c

, O. Dugne

c

, R. Podor

d

, N. Clavier

d

, P. Blanchart

e

, A. Ayral

f

aRadioChemistryandProcessesDepartment,SCPS,LC2A,CEA,NuclearEnergyDivision,F-30207Bagnols-sur-Cèze,France

bFuelCycleTechnologyDepartment,SDTC,LEMA,CEA,NuclearEnergyDivision,F-30207Bagnols-sur-Cèze,France

cFuelCycleTechnologyDepartment,SGCS,LMAC,CEA,NuclearEnergyDivision,F-30207Bagnols-sur-Cèze,France

dInstitutdeChimieSéparativedeMarcoule,UMR5257CEA-CNRS-UM2-ENSCM,F-30207Bagnols-sur-Cèze,France

eHeterogeneousMaterialsResearchGroup,CentreEuropéendelaCéramique,F-87068Limoges,France

fInstitutEuropéendesMembranes,UMR5635CNRS-ENSCM-UM2,CC047,UniversitéMontpellier2,F-34095MontpellierCedex5,France Received8February2012;receivedinrevisedform5April2012;accepted9April2012

Abstract

Thisstudydealswiththepreliminarydevelopmentofapowder-freeprocesscalledcalcinedresinmicrospherepelletization(CRMP)usedfor thefabricationofmetallic oxidepellets.Thisdustlessprocesscouldbeusedforthefabricationof mixedU1yAmyO2±x pelletsdedicatedto thetransmutationofAminfastneutronreactors.Inthisstudy,porousCeO2 microspheres,usedasasurrogateofAmO2,wereobtainedafter mineralizationofceriumloadedionexchangeresinbeads.Thesemillimetricoxidemicrospheresweredie-pressedintopelletswhichwerethen sinteredunderairtoformceramicpellets.Theirdensitiesapproached95%oftheoreticaldensityofCeO2andahomogeneousmicrostructurewas obtainedbyusingoptimizedmicrospheres.Theinfluenceofcalcinationparametersonthecharacteristicsofmicrospheresandonthepropertiesof sinteredpelletsisdiscussed.

©2012ElsevierLtd.Allrightsreserved.

Keywords:Microsphere;Calcination;Pressing;Sintering;Precursors-organic

1. Introduction

Oneof the most promising optionsunder study toreduce theradiotoxicityandtheheatloadofnuclearwastestorageis thetransmutationofthelong-livedradionuclidessuchasminor actinides (MA),americium (Am) for instance,in afast neu- tronreactor.1Inthisrecyclingprocess,MAaretransformedby nuclearreactionsintoshort-lived orstablenuclides.Different modesof transmutationare investigated: homogeneoustrans- mutation,consistingof dilutingMA inthewholecoreof the

Abbreviations: MA, minor actinides; MABB, minor actinide bearing blanket;SGMP,sol–gelmicrospherepelletization;CRMP,calcinedresinmicro- spherepelletization;FT,finaltemperature;HR,heatingrate;TD,theoretical density;PHWR,pressurizedheavywaterreactor;WAR,weakacidresin;IER, ionexchangeresin;LMFBR,liquid-metalcooledfast-breederreactors;Qweight, scientificweightcapacity;Qvol,technicalvolumecapacity;meq,milliequivalent.

Correspondingauthor.Tel.:+33466791827;fax:+33466796567.

E-mailaddress:elodie.remy@cea.fr(E.Remy).

reactor andheterogeneoustransmutation,whereMAarecon- centratedinuraniumdioxidefuelsintroducedintheperiphery ofthecore.2Thesespecificfuelsareoxidematerialscalledminor actinide bearing blanket (MABB). Their fabrication requires special attention due tothe highradiotoxicity of MA.3 Cur- rently, the fabricationof these mixed oxide pellets follows a powder metallurgyroutewhichconsists innumerousstepsof grindingandmillingofUO2andAmO2powders,pressingand sintering.4,5Preparationofgranularmaterialbeforecompaction andespeciallygranulation,whichisrequiredtoproducefree- flowing feed material for automatic pellet presses, generates fineandhighlycontaminantparticles.Dustlessprocesseswhich avoidthepresenceofhighlycontaminatingpowderinthefab- ricationline,usingoxideintheformoffreeflowingspherules withmillimetricsize,arerecommended.6

The firstprocess developedinthe 70s, basedonthe com- paction ofmillimetric spherestoform pellets,wascalled the sphere-calprocess.7–9AuthorsstudiedthecompactionofUO2

andThO2beads,obtainedbysol–gelprocess,intopelletsand

0955-2219/$seefrontmatter©2012ElsevierLtd.Allrightsreserved.

http://dx.doi.org/10.1016/j.jeurceramsoc.2012.04.011

(3)

thesubsequentsinteringtoformsimpleormixedoxidefuelpel- letsforadvanced pressurizedheavywaterreactors (PHWRs).

Their work focused especially on the influence of the gela- tionandcalcinationprocessesonbeadcompactionintopellets andonthesubsequentsintering.Exhaustiveinvestigationswere carried out tooptimize heattreatment parameters inorderto obtainUO2microspheressuitableforpelletization.Morerecent studieshavebeen published on thisprocessrenamed sol–gel microsphere pelletization (SGMP) to produce ThO2–UO2

pellets.10–12 Pelletizationof soft porous microsphereswith a low crushing strengthleadsto thesynthesis of homogeneous and dense sintered pellets (≥94% TD). The SGMP process was then studied for the elaboration of UO2 pellets13,14 and mixeduranium,cerium orplutonium oxidepelletsfor liquid- metalcooledfast-breederreactors(LMFBR).Inthesestudies, cerium was used as a surrogate of plutonium.15,16 Pellets with high density (≥95% TD) anda tailored microstructure containing uniformlydistributed closed pores were produced by using porous microspheres, obtained by the addition of carbon black as a pore former or by adjusting gelation parameters.14,17 Theviabilityof the SGMProuteforfabrica- tion of (U,Pu)C and (U,Pu)N has also been established.18,19 Afterwards, uranium–plutonium nitrides as low-density pel- lets have been prepared and characterized for transmutation application.20

Aswiththesphere-calortheSGMPprogramme,ourstudy dealswiththedevelopmentofaspheruleroutewhichiscalled the calcinedresinmicrospherepelletization(CRMP) process.

It consists of the synthesis of mixed oxide precursors of sphericalformandmillimetric sizeobtainedbyanadaptation of the weak acid resin (WAR) process21–24 and their com- pactionintopelletsbeforesintering.Themicrospheresynthesis is based on the fixation of cerium cations into beads of ion exchangeresin(IER)andtheirmineralizationinairtoformthe oxide.

Thispresentstudyconstitutesafirststepintheelaborationof mixedoxidepelletsbytheCRMPprocesswiththeelaborationof cerium(asasurrogateofamericium)oxidepellets.Itdescribes theconversionmethod,theoptimizationofthecalcinationstep andtheresultsofcompactingandsinteringbyexaminationof thepellets’microstructureanddensity.

2. Materialsandmethods

2.1. Startingmaterials

TheacrylicresinwasobtainedfromDowChemicalsCom- pany (RohmandHaas, Chauny, France)and consistedof an IMACHP335geltypeacrylicexchangerintheformofbeads of3differentsizedistributions(batches1,2and3).Theirchar- acteristicsarelistedinTable1.

Concentratedammoniasolution(25%,Merck,ProAnalysis) and concentrated nitric acid solution (70%, Fisher Chem- ical, Certified ACS Plus) were used to prepare washing solutions. Hexahydrate cerium(III) nitrate (Ce(NO3)3·6H2O, 99.99%pure,Prolabo)wasusedtoprepare0.25MCe(III)stock solution.

Table1

Characteristicsofthe3batchesofIMACHP335ionicexchangeresin(nc,not communicated).

IMACHP335resin Commercial form:batch1

Finebeads:

batch2

Coarsebeads:

batch3

Structure Gel

Functionalgroup COOH

Physicalform Opaquebeads

Matrix Polyacrylic

Exchangecapacity(meq/mL RH)

3.85 4.19 4.02 Moistureholdingcapacity 52–58% 58.6 54.6

Density(H+form) 1.14–1.18 nc nc

Harmonicmeansize(m) 500–700 <300 >1000

2.2. Solutionanalysis

Theceriumcontentofthedifferentsampleswasdetermined by atomicemissionspectrometry (ICP-AES) usingan Activa HORIBALABequipment.Acalibrationwasperformedusing diluted Ce certifiedstandard (Merck, Certipur, 1000mg/L in 2–3%HNO3)in0.5M HNO3.Scandiumwasalsoused asan internalstandard(Merck,Certipur,1000mg/Lin2–3%HNO3) atafixedconcentrationof25mg/L.Twowavelengthswerecho- senforceriumanalyses(394.275and413.380nm)andonefor scandiummeasurement(358.963nm).

The ammoniumconcentration wasmeasured byacid–base titration carried out with aTitrando 809 Metrohmautomatic titratorinthepresenceofpotassiumoxalateusedtoshiftcerium hydrolysis. The evolution of pH was followed continuously usingaConsortC861analyserandaglasselectrode(Blueline 14pH,SchottInstruments).

2.3. Synthesisofthesolidoxidemicrosphereprecursor 2.3.1. Preparationofresinbeads

ThewetresinwasfirstmechanicallyscreenedwithanAS200 BasicRetschapparatusunderdeionizedwaterthroughsucces- sivelyfinerstandardProlabosieves(100␮m,250␮m,400␮m, 630␮m,800␮m,1000␮mand1250␮m).Thenamanualsec- ondscreeningwasperformedoneachsizerangeinordertolimit cloggingofthesieves.

The selected resin waswashed inacolumn intwo cycles consisting of percolating successively a 1M aqueous nitric acidsolution,deionizedwater,1Maqueousammoniasolution and deionized water. Two cycles were necessary to remove contaminants such as sodium and synthesis impurities. The resin inits finalammonium formwas drainedand recovered semi-wet.

2.3.2. Resinmetalloading

Abatchmethodwasusedtoperformthecationexchangeon thewashedresininitsammoniumform.Thoseresinbeadswere fullyrehydratedonacolumn2hbeforeexchangeindeionized waterandtheirpackedbedvolumewasmeasured(15–45mL).

Different size distributions of resin were contacted with the 0.25M cerium nitratestocksolution (100–300mL)for 4hin

(4)

avesselatroomtemperature(22±2C)undergentleagitation withamagneticstirringbar.Theresinwasthentransferredtoa columnandwashedwithdeionizedwater.Afterdrainingunder vacuum,theloadedresinwasdriedat110Covernight.During exchange,thebulksolutionwassampledtofollowceriumand ammoniumconcentrations.

2.3.3. Calcination

Calcinationof theloadedresinwasperformedinatubular furnace(Nabertherm,HTRH100-300;aluminatube1200cmin lengthand10cmininternaldiameter)undersyntheticairflow (AirLiquide,20%O2–80%N2).Temperaturesintherangeof 500–1400Cweretested.Themicrosphereswereplacedas a singlelayerinanaluminacrucibleinordertoobtainahomoge- neouscalcinationasmentionedbyFerreiraetal.25

2.3.4. Pressingandsintering

Ceriumoxidemicrospheresobtainedaftercalcinationwere coldcompactedintogreenpelletsina10TpneumaticEnerpac pressat500MPa.Tungstencarbidedie-punchsetsof5.165mm diameterwereusedandwerelubricatedbyzincstearatepriorto eachcompaction.Thegreenpelletsweresinteredat1400Cfor 6hinthesamefurnaceandsameairflowaspreviouslydescribed.

Theheatingandcoolingrateswerefixedat10C/min.

2.4. Solidcharacterization 2.4.1. Loadedresin

An elemental analysis was performed on asample of the 630–800␮m loaded resin. A mass of 150mg of this sam- ple was placed in a column and contacted by percolating 25mLof1MHNO3during100min.Thepercolatewasrecov- eredin a50mL volumetric flask and completed by washing waterfromtheresin.ICP-AESanalyseswerethenconducted on this solution to determine the concentration of cerium released during dissolving corresponding to the resin metal content.

Themetal content wasalso estimated bythermogravimet- ricanalysis(TGA)anddifferentialscanningcalorimetry(DSC) performedwithaSTA449CNetzschequipment.Microspheres wereplacedinanaluminacruciblefittedwithacoverandwere calcinedupto1100Cinairflow(3NL/h)ataheatingrateof 5C/min withan isothermaltreatmentof 1hat1100C.The residuewaspureceriumoxidewhichwasconfirmedbyX-ray diffraction(XRD)andcarbonanalysis.

2.4.2. Oxidemicrospheres

The effective density26 of oxide microspheres was mea- sured by helium pycnometry (Accupyc 1330 Micromeretics pycnometer). The pycnometer was equipped with a 10cm3 volume cell and a sample module of 1cm3. It worked with helium (Alphagaz; Air Liquide, 99.995% pure). The 1cm3 volumemodulewascalibratedusinga0.718502cm3standard ball.Eachsamplewasdegassedbeforemeasurementovernight at140C under vacuumusing a VacPrep061 Micromeretics apparatus.

Apparentdensity26ofmicrosphereswasdeterminedbyvol- ume and mass measurements of asample of a few hundred beads.Numberofspheresandthesamplevolumewereestimated from image analysis of a layer of microspheres (considered as perfect spheres). Images were taken by an optical video microscope(FortSVM01)andtreatedbyEllixpatternrecog- nitionsoftware(Microvision).Unitaryweightequivalenttothe weightofoneoxidemicrospherewasalsodeterminedbythis method.

SpecificsurfaceareawasobtainedfromN2adsorptionand desorption at77Kusing the Brunauer–Emmett–Teller(BET) method.ATristar II3020 Micromereticsapparatus wasused formeasurements.Sampleswerepretreatedinavacuumoven toremovevolatiles.

Theresidualcarboncontentinoxidemicrosphereswasmea- suredwithaCS230LECOcarbon–sulphuranalysercalibrated withcarbonstandardsobtainedfromLECO(501-024,501-676 and502-809).The crystallographicstructure of cerium oxide microsphereswascharacterizedbypowderXRD(D8advanced BRUCKERdiffractometer,CuK␣radiation).Gold(NBSStan- dard)wasaddedtoallsamplesasaninternalstandardtocalibrate theangularpositionsoftheobservedXRDlines.XRDdiagrams weretreatedbyDiffracplusEvasoftware.Thelatticeparameters oftheoxidewererefinedbypatternmatchingusingFULLPROF Suite.27

2.4.3. Greenandsinteredpellets

Effectivedensityofsinteredpelletswasmeasuredbyhelium pycnometry. Bulkdensity of pellets was obtained byvolume measurementandgravimetricanalysis.Adilatometrictestwas carried outonagreenpellet withaSetsys evolutionSetaram dilatometerwhichallowedcontinuousmonitoringoftheshrink- agekineticsupto1400Cwithaheatingrateof10C/minunder air flow (syntheticair, Alphagaz, Air Liquide, 20% O2–80%

N2).

2.4.4. Microstructureanalysis

Scanningelectron microscopy (SEM)micrographsof pel- letsandmicrosphereswerecarriedoutwithaSupra55ZEISS instrument.Microsphereswerecoldmountedwithepoxyresin withinaringmouldandthengrindedwithSiCsandpaper(1200, 2400and4000gritsizes)toaccessinternalmicrostructure.A diamondsuspension(10␮m)wasusedtopolishsamplesanda finalpolishingwithcolloidalSiO2suspensionwasperformed.

Internal microstructure of microsphereswas observed on the polishedsurface.

Sintered pelletswere cut andsections were coldmounted andpolishedfollowingthesameprocess.Insituhightempera- tureenvironmentalscanningelectronmicroscopy(HT-ESEM) experimentswereperformed usingafieldemissiongun envi- ronmentalscanningelectronmicroscope(modelFEIQUANTA 200 ESEM FEG) equippedwitha 1500C hot stage.28 Dur- ingthiswork,observationswereperformedunderwatervapour atanoperatingpressureof233Paandaspecificdetectorwas usedforinsitugaseoussecondaryelectronimagingathightem- perature.Anindividualbeadofceriumloadedresinwasfixed

(5)

Table2

SizedistributionsoftheIMACHP335resin.

Batch 1 2 3

Sieveaperture(m) Noncumulativetotal(betweensieves)(%)

100–250 100

250–400 5.4

400–630 47.2

630–800 40.9

800–1000 6.4 19.7

1000–1250 79.0

1250–1400 1.3

onrefractorycementandwasheatedto1000C.Micrographs wererecordedatregularintervalsduringtheheattreatmentof themicrosphere.

3. Resultsanddiscussion

3.1. Exchangecapacity

Theresinwassortedfirsttoworkonhomogeneousbatches forexchangekinetics.Thesizedistributionsofthethreebatches ofIMACHP335resinwereobtainedbysieving(seeTable2).

Foursizerangeswerequalifiedinthisstudy:100–250,400–630, 630–800and1000–1250␮m.

Exchangecapacityoftheresinwasdeterminedbymeasuring theamountofammoniumandprotonexchangedduringthelast washingcycle.Averagevaluesofscientificweightcapacityof twosize distributions(Qweight)inmilliequivalentpergramof driedresininprotonform(meq/gdryH+ form)andtechnical volumecapacity29ofeachsizedistribution(Qvol)inmilliequiv- alent per millilitre of packed bed (meq/mL packed bed) are shown inTable 3. The swelling factorwhich corresponds to volumeexpansionofthepackedbedafterH+/NH4+conversion isalsomentioned.Swellingisimportantbecausetheactionof ammoniaccausescompleteiondissociationintheresinbutalso introducesalargesolvatedammoniumcounterion.29,30

The scientific weight capacity does not vary significantly withsizedistributionandisequalto12.6meq/gdryH+ form, which is not far from the theoretical value of 13.3meq/g dry H+ form (knowing the weight of 1mol of carboxylic moiety to be 75g, which represents 1 equivalent, the theo- retical weight capacity corresponds to 1/75=13.3meq/g dry H+).

3.2. Kineticsoffixation

EvolutionsofpH,ammoniumandceriumconcentrationsdur- ing conversion step of the 630–800␮m size distribution are plottedinFig.1.

The amount of cerium in solution rapidly decreased dur- ing the first hour of conversioncorresponding to the loading ofceriumintheresinandthesubsequentreleaseofammonium.

Accordingly,thequantityofNH4+insolutionincreased.After 1h,thepHstabilizedatavaluebetween5and6corresponding toachievementofequilibrium.

Fig.1.EvolutionofCe3+andNH4+concentrationsinsolutionduringthecon- versionstep.

3.3. Ceriumcontentintheloadedresin 3.3.1. Balancefromsolutionanalyses

Experimentalconditions andresults of conversionof each sizedistribution,suchasfinalmassoftheloadedresin,quantity ofceriumfixedintheresin,quantityofammoniumreleasedin solution, effectiveexchange volumecapacity andweightper- centageofceriuminthedriedloadedresinarelistedinTable4.

Thequantityofceriumfixedintheresiniscalculatedonthe basis of the differencebetween the initialquantity of cerium andtheamountofCefoundinthesolutionatequilibrium,in samplingandwashing.TheweightpercentagesofCe(wt%Ce) for everysizedistributionsarequitehomogeneousvaluesand theirmeanvalueequals35±3%.

Thequantityofammoniumreleasediscalculatedinthesame waybyaddingtheamountofammoniuminsolutionatequilib- rium,samplingandwashing.

The stoichiometry of the exchange reaction (ratio of the quantity of ammonium releasedin solutionover the quantity of ceriumfixedintheresin)iscloseto3whichconfirms that the lanthanidecationtakesup3 exchangingsitesintheresin and thereforereleases 3 ammoniumcations insolution.This ratioof3hasalreadybeenreportedintheliterature31forLa3+, Ce3+,Pr3+inthecaseofaNa-formresinatpH6–7.Thecation exchangereactioncanbewrittenasEq.(1):

3RNH4+Ce3+→ R3Ce +3NH4+ (1) whereRrepresentsanexchangingsiteoftheresin.

3.3.2. Balancefromsolidcharacterization

Elementalanalysisalsoindicatesthattheweightpercentage of cerium equals 33±2% for the630–800␮mloadedbeads.

Metal contentinthe fully loadedresin canalsobe measured after calcinationunderairat800Cfor eachsizedistribution consideringthattheresidueispureceriumoxidefreeofcarbon (confirmedbycarbonanalysespresentedlaterinTable6),the residualmassisrepresentativeofthemetaloxidematerialsleft aftercalcinationandallowsassessmentoftheinitialmetalcon- tentintheresin.TheweightpercentageofCeissimilarforeach sizedistributionandthemeanvalueis34.4±0.2%.

(6)

Table3

TechnicalvolumeandscientificweightcapacitiesofdifferentsizedistributionsofIMACHP335.

Sizedistributions(m) 100–250 250–400 400–630 630–800 800–1000 1000–1250 Meanvalues

QvolH+form(±0.2)(meq/mL) 4.1 4.1 3.8 3.7 3.6 3.9 3.9±0.1

QvolNH4+form(±0.1)(meq/mL) 2.66 2.56 2.52 2.67 2.63 2.53 2.60±0.04

QweightH+form(±0.1)(meq/g) 12.5 12.6 12.6±0.1

Swellingfactor(±0.1) 1.4 1.5 1.4 1.4 1.4 1.4 1.4

Table4

ResultsofCeconversionfordifferentsizedistributions.

Sizedistribution(m) 100–250 400–630 630–800 1000–1250

VresinNH4+(mL) 15.0±0.3 15.0±0.3 42.0±0.6 15.0±0.3

VCe(NO3)3(mL) 100.0±0.6 100.0±0.6 300.0±0.6 100.0±0.6

FinalpH 4.38±0.05 4.68±0.05 5.28±0.05 5.22±0.05

n(Ce3+)loaded(mmol) 14.0±0.8 12.6±0.8 38.9±2.0 13.0±0.8

n(NH4+)released(mmol) 43±2 40±2 121±1 39±2

Effect.Capa.(meq/mL) 2.8±0.1 2.5±0.1 2.8±0.2 2.6±0.1

EquivalentNH4+/Ce3+ 3.1±0.3 3.2±0.3 3.1±0.4 3.0±0.3

mresinCe(g) 5.88±0.01 5.43±0.01 14.25±0.01 4.92±0.01

wt%Celoadedresin 33±2 33±2 38±5 37±2

Thesizedistributionhasnoinfluenceontheweightpercent- ageofCeinthedriedresinwhichisabout34%accordingtoall characterizations.Thoseresults areinagreement withresults concerning the fixation of lanthanide(III) cations studied by Mokhtari.32 Thetheoreticalvalueof weightpercentageof Ce inthedriedresin corresponding toafullexchange is38.7%, calculatedfromthefollowingequation:

wt% Ceth= MW(Ce) MW(Ce)+MW(R)

= 140.12

140.12+3×74 =38.7 (2)

with MW being the molecular weight and R the carboxylic inorganicgroupinitsdeprotonatedform.

Theresinexchangecapacityisnotfullyused owingtothe establishmentofequilibriumgeneratedbytheaccumulationof NH4+ insolution duringthe batch exchange.33 The effective capacityreaches90%oftheoreticalcapacity.

3.4. Solidanalyses 3.4.1. TGA

Thethermalconversionofasampleof630–800␮mloaded resinwasinvestigatedusingTGAataheatingrateof5C/min fromroomtemperatureto1100Cinair(Fig.2).Thethermal behaviourofoneindividualmicrospherewasalsocharacterized byHT-ESEMwhichenablestofollowtheevolutionofthemicro- spherediameterwithtemperature(seeFig.2andAppendixB).

SeriesofimagesisgiveninFig.3.Thefirstsignificantweight loss(5%)betweenroomtemperatureand290Ccanbeassigned todehydrationandlossofweaklyadsorbedwatercoupledwith reduction of3% inmicrospherediameter. Theelimination of carbonmatterintheloadedresinbeadsoccurredinthetemper- aturerangeof290–690Cwithalossof52%oforiginalmass.

Thisprocesscanbedividedintothreeexothermicmechanisms, withafirstdegradationstepbetween290Cand420Ccaus- ingadiameterdropof12%followedbyasecondstepbetween 420Cand490Cconcomitantwitha3%diameterdecrease andafinaldegradationuntil690Ccorrespondingtoadiame- terreductionof4%.Themicrospherecontinuestoshrinkabove 690C.Thisphenomenonisattributedtothebeginningofsin- tering.Upto800C,adecreaseof24%inmicrospherediameter isobserved.

3.4.2. Microspherecharacteristics

Thediametralandvolumetricshrinkagesoftheloadedresin (R3Ce)forcalcinationupto800Cweredeterminedbyimage analysisforeachsizedistribution.Meanvaluesare22±3%and 53±8%respectively.Theformervalueofdiametralshrinkage isinagreementwiththedataof24%obtainedfromHT-ESEM images(Fig.2).

Fig.2.TGAandDSCcurvesof630–800mloadedresinbeadsandevolution byHT-ESEMofthediameterofoneloadedresinbeadduringheattreatment.

(7)

Fig.3.EvolutionbyHT-ESEMofoneloadedresinbeadduringcalcination.

Fig.4.LatticeparametersatdifferenttemperaturesrefinedbyFULLPROFand X-raydiffractionpatternoftheCeO2 microspherescalcinedat800C(Au standard).

3.4.3. Crystallineevolution

X-ray diffraction patterns were obtained on oxide micro- spheres calcined at 600C, 800C and 1000C. For all calcinationtemperatures,thediffractionpatternsarecharacter- isticoftheface-centredcubicCeO2.Thepatterncorresponding

Table5

LatticeparametersatdifferenttemperaturesrefinedbyFULLPROF.

Temperatureof calcination(C)

Latticeparameter (±0.001)( ˚A)

Averagecrystallite size(±3)(nm)

600 5.407 19

800 5.407 32

1000 5.408 48

tothetemperatureof800CisreportedinFig.4.Therefined latticeparameter(Table5),a=5.408 ˚Aisnotsignificantlydiffer- entfromthevalueof5.411 ˚Apublishedintheliterature.34The oxideisalreadyformedat600C.Thevariationofthecrystal- lizationstatereportedinFig.4isestimatedthroughtheaverage fullwidthathalfmaximum(fwhm)fromthemoreintenseXRD lines. The crystallite size increases withtemperature and the evolutionisingoodagreementwithpublisheddata.35

3.4.4. Microstructuralevolution

ESEMmicrographsrealizedonthecalcinationofaunique loaded resin microsphere enabled identification in situ the microstructuralchangesofthemicrosphereduringtheheating process(Fig.5).

Afterdehydrationandatthebeginningofmassloss,cracks appearatthesurfaceofthebeadscorrespondingtothebeginning

Fig.5.MicrostructuralevolutionofaceriumloadedresinmicrosphereduringthermaltreatmentobtainedbyHT-ESEM.

(8)

oftheremovalofcarbonintheformofcarbondioxide.Open porosityiscreatedatthismomentandseemstodisappearwhen temperatureincreasesandthediameterofthespheredecreases.

SEM micrographs performed on as-polished sections of CeO2beadsrevealthehomogeneityoftheirinternalmicrostruc- tureandmicrometre-sizedgrains(Fig.6).Theporosityevolution versustemperaturewasinvestigatedandresultswillbepresented hereafter.

3.5. Calcinationoptimization

Differentthermalcycleswereappliedto630–800␮mbeads tocomparetheinfluenceofeachparameterofthecyclesuchas:

heatingrate(HR)andfinaltemperature(FT).Adwelltimeof4h wasappliedatfinaltemperaturetoallbatchesofmicrospheres exceptformicrospherescalcinedat500C.

3.5.1. Heatingrate(HR)

Differentheatingrates wereappliedduringthe calcination oftheloadedresinbeads:0.1,0.2,0.5,1,2.5,5C/min(witha FT=800C).

Heatingratehasanimportantinfluenceonthemicrosphere fractureasshowninFig.7.

An excessive heating rate of 5C/min causes violentout- gassingofCO2,leadingtoimportantmechanicalstressonthe beadsandtheirdisintegration.Thepercentageofbrokenbeads calcinedat5C/minisabout20%.Moreover,anincreaseofheat- ingrateresultsingreaterresidualcarboncontentintotheoxide bead(2000ppmat5C/mincomparedto570ppmat1C/min).

Removalofcarbonisnotcompleteforexcessiverate.

Aslowheatingrate(≤1C/min)wasthenselectedduringthe mineralizationtoformhomogeneousoxidemicrosphereswith enoughmechanicalstrengthtoavoidemissionofsmallparticles duringthehandlingstep.

3.5.2. Finaltemperature(FT)

Finaltemperaturesof500C(withnodwelltime)and600C, 800C,1400Cweretestedforoptimizingthecalcinationof loadedresin.Characteristicsofcalcinedbeadsaresummarized inTable6.

Afirstobservationisthattheeffectivedensityofoxidebeads increases with temperature which is linked mostly with the decreaseofclosedporosityfrom4%at500Ctoavalueinfe- riorat1%at1400C.Openporosityalsodecreasesfrom800C butstillremainselevatedforbeadscalcinedathightemperature (68% at1400C) as we can observe inFigs. 8 and 9 which explainsthatbulkdensityofbeadsislimitedto31%ofTDin thoseconditions.

Thoseresultsareinagreementwiththoseof TGAandthe evolutionofmeandiameterduringmineralization(Fig.2).They revealthedensificationofmicrosphereabove690Cinrelation withagraingrowthobservableinFig.9andadecreaseofthe specificsurfaceofoxidebeadswithtemperature(from45m2/g at500Cto0.7m2/gat1400C).

A second observation concerns the evolution of carbon contentintotheoxidewhichisabout0.15%below600Cinde-

pendentlyofdwelltimeandwhichisnotsignificantabovethis Table6 Characteristicsofoxidemicrospheresobtainedunderdifferenttemperaturesofcalcination. ThermalcycleofmicrosphereContentofresidualBulkdensityBulkdensityEffectivedensityOpenporosityClosedporositySpecificsurfaceMeandiameterUnitaryweight 3aab2calcination(FT;HR)carbon(ppm)(±0.15)(g/cm)(±1%)(%TD)(±1%)(%TD)(±1%)(%V)(±1%)(%V)(±0.1)(m/g)(±20)(m)(±6)(g/u)bulkbulk (500,1)1350±1501.27188278444.948979 (600,1)1360±1001.34199380144664 (800,1)570±751.1816968318.249175 (1400,1)2.2431986810.739272 a343TD,theoreticaldensityofCeO(7.216g/cm).2 b%V:percentageofthebulkvolume.bulk

(9)

Fig.6.SEMmicrographsperformedonas-polishedsectionofCeO2beadsheattreatedat1400C.

Fig.7.CeO2microspheresobtainedbycalcinationinairwithdifferentTR:(a)0.2C/min;(b)1C/minand(c)5C/min.

Fig.8.Evolutionofeffectivedensity,openandclosedporosityofCeO2beads calcinedatdifferenttemperatures.

temperature inconformity withtheendof masslossinTGA curve.

Unitary weight corresponding to the weight of one oxide microsphere can be also calculated and is about 75␮g/microsphereequivalenttoapproximately61␮gofCeper microsphere.Thosedata haveagreatinterest inthedetermi- nationofcapacityrequirementsasmentionedbyHaasetal.36 Forinstance,thevolumeofinitialresininitsammoniumform, necessary for the productionof 1g of CeO2 (mCeO2) which representsapproximately1/75×106=13,300beadsofoxide (nspheres),isabout10mLofresinbed.Detailedcalculationsare giveninAppendixA.

Finaltemperatureisacriticalparameterandenablessynthesis ofmicrosphereswithdifferentfinalamountsofcarbon,specific surface,densityandporosity.Theseparameterswillhaveimpor- tantconsequencesforthemicrostructureofsinteredpellets,as willbediscussedinthefollowingsection.

Fig.9.SEMmicrographsofthesurfaceofCeO2beadscalcinedat(a)(600,1)and(b)(1400,1).

(10)

Fig.10.Dilatometricanalysisperformedonagreenpelletobtainedbycom- pactionof630–800mbeadscalcinedat800Cat1C/minfor4h.

3.6. Compactingandsintering

Microspherescalcinedabove800Cwereuniaxiallypressed at500MPaina5.165mmmatrix.Themicrospheresobtainedat lowertemperatureswerediscardedbecausetheircarboncontent wastoohigh.Thegreenpelletsweresubsequentlysinteredat 1400C for 6h in air. A dilatometric analysis performed on agreenpelletfollowingthisthermalcycleisshowninFig.10.

Sinteringbeginsat800Cwithamaximumlinearshrinkagerate at1090C.Thesinteringisendedat1400C.Thefinallinear shrinkageafter coolingisabout 14%.The shrinkagecurveis classicalandinaccordancewiththesinteringbehaviourofgreen

pelletsobtainedfromthecompactionofceriumoxidepowder preparedfromoxalateprecursorsevenifthefinalshrinkageis slightlyinferior: ameanvalue of18% isobtained withthese pellets.37

Thegreenpelletdimensionsweremeasuredbeforesintering todeterminebulkdensity.Effectiveandbulkdensitiesofceramic pelletsareshowninTable7.

The bulk density of green pellets is found to be between 62 and65% of TD which is greater than the value of 52%

obtainedforCeO2pelletsfabricatedbypowdercompaction.37 Aslightinfluenceofcalcinationparametersonthegreendensity isobserved,whichincreaseswithheatingrate.

Greenpelletsobtainedfromthecompactionofmicrospheres calcinedat1400Cweretoobrittleandnotsuitableforhandling.

Thiscanbeexplainedbythehardnessofbeadswhichistoohigh andnotappropriateforpressing.Thispropertyhasbeenwidely describedintheliteratureasaconsequenceoftheuseof“non- porous”andhardmicrosphereswhichdonotdisintegrateduring pelletization.9,25

Sinteringofgreenpelletsat1400Cfor6hresultedineffec- tivedensitybetween85and94%ofTDdependingontheheating rate. In particular, aheating rate of 5C/min duringcalcina- tionleadstosinteredpelletswithanimportantclosedporosity approaching15%probablyduetoaheterogeneouspopulationof beadsbeforecompactionasmentionedpreviously.SEMmicro- graphsofas-polisheddiametersectionsof sinteredpelletsare giveninFig.11.

Pellets obtained from compaction of beads calcined at a low rate of 1C/min up to 800C are dense and present a

Fig.11.SEMmicrographsofas-polishedsectionsofsinteredpelletsobtainedfrommicrospherescalcinedat(a)800Cat1C/minfor4hand(b)800Cat5C/min for4h.

(11)

Table7

Densitiesofgreenandsinteredpellets.

Thermalcycleofmicrospheres calcination(FT;HR)

Bulkdensityofgreen pellets(±0.2)(g/cm3)

Sinteredpellets Bulkdensity (±0.2)(g/cm3)

Effectivedensity (±0.02)(g/cm3)

Effectivedensity (±0.2)(%TD)

Open porosity(%)

Closedporosity (±0.1)(%)

(800,0.2) 4.5 6.8 6.66 92.3 0 7.7

(800,0.5) 4.5 6.8 6.69 92.7 0 7.3

(800,1) 4.5 6.8 6.71 93.0 0 7.0

(800,2.5) 4.6 6.8 6.78 94.0 0 6.0

(800,5) 4.7 6.3 6.15 85.2 0 14.8

(1400,1) Brokenaftercompaction

homogeneous microstructurewith6% of closedporosity and noopen porosity.Onthe contrary,ahighheatingrate causes aheterogeneousandporousmicrostructure.Afinalcalcination temperatureof800Cproducesbeadswithspecificsurfaceof 8.2±0.1m2/gandbulkdensityof1.18±0.15g/cm3andseems tobeanoptimizedtreatmenttoobtaindenseandhomogeneous pellets.Actually,thosetwocharacteristicshavebeendescribed intheliterature8,25asfundamentalpropertiesinthecaseofthe optimizationofthepelletizationprocessofmicrosphereprecur- sorsobtainedfromsol–geltechnology.Theauthors’conclusion istouselowdensifiedkernel(∼2g/cm3)withlowspecificsur- face(∼5m2/g)obtainedat850C.

4. Conclusion

Thisworkconstitutesafirstapproachconcerning theelab- oration of dense cerium oxide pellets by the calcined resin microspherespelletization(CRMP)processbasedonthecom- paction of oxide microspheres into green pellets and their subsequent sintering to obtain dense ceramic pellets. Those homogeneousandporousoxidebeadsof150–750␮mindiam- eterwereobtainedfrommineralizationofresinbeadsloadedin ceriumfollowingdifferentthermaltreatments.Differentcalci- nationtemperaturesandheatingrateshavebeeninvestigatedto produceoxidebeadssuitable forpressing.Inparticular, com- pactionofporousoxidebeadscalcinedatslowmineralization rate(≤1C/min)andatemperatureof800Cresultsindense andhomogeneousmetaloxidepellets(95%ofTD).

In contrast, an overly high temperature of calcination (1400C)resultedinhardbeadsnotsuitableforcompaction.

Themechanicalbehaviourundercompressionofthoseoxide beadsproducedindifferentcalcinationconditionsisunderstudy andshouldimprovethecontroloftheceramicmicrostructure.

EventuallytheoptimizedCRMPprocessshouldbeappliedtothe elaborationofmixed(U,Ce)O2pelletsand(U,Am)O2ceramics.

Acknowledgements

The authors would like to thank the MATINEX French ResearchGroupforitsfinancialsupportandD.Lacoumefrom RohmandHaas,France(subsidiariesof DowChemicals)for the supplyof the IMAC HP335 ionicexchange resin.Agnès Grandjean andCyrielleReyare thankfully acknowledgedfor

thecarbonanalyses.TheauthorsaregratefultoBénédicteArab- ChapeletforherhelpwithXRD.

AppendixA. Determinationofthevolumeofinitial resininitsammoniumform,necessarytotheproduction of1gofCeO2(mCeO2)

mRCe= mCeO2

1−WL = 1

1−0.57 =2.32 (A.1)

wheremRCerepresentsthemassoftheceriumloadedresinand WL isthe weightloss inpercentageduringcalcinationtaken fromTGAequalto57%.

mCe=wt% Ce×mRCe= 0.344×2.32=0.8 (A.2) withwt%Cerepresentingtheweightpercentageof Ceinthe loadedresinwhichisequalto34.4%.

So

VRH(mL)= mCe

MWCe×3×1000 QvolH+

= 0.8

140.12×3×1000

2.6 =6.6 (A.3)

whereMWCeistheCemolecularweightandQvolH+isthetech- nicalvolumecapacityoftheprotonresinexpressedinmeq/mL (Table3).

The estimation of the resin bed volumein its ammonium form (VRNH4)isobtainedby takingintoaccount theswelling factor described in Table 3: in this case, VRNH4 is equal to 6.6×1.4–10mL.

AppendixB. Supplementarydata

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/

j.jeurceramsoc.2012.04.011.

References

1.BoullisB,WarinD.Recentprogressinadvancedactiniderecyclingpro- cesses. In: OECD NuclearEnergy Agency 11th information exchange meetingonactinideandfissionproductpartitioningandtransmutation.

2010.

2. KoningsRJM, HaasD.Fuels andtargetsfor transmutation.CR Phys 2002;3:1013–22.

Références

Documents relatifs

ةصالخاو ةيمومعلا رارضلأا ىلع ينمأتلا تاكرش جاتنإ في ينمأتلا ءاطسو ةمهاسم ينب ةنراقلما : م ةيلاعف ىدم ةفرعم لجأ نم علا رارضلأا ىلع ينمأتلا تاكرش في ينمأتلا

Ahmed Fakhry, Mathieu Pinault, Émeline Charon, François Tran Van, Fouad Ghamouss.. Electrode- position of MnO2 on VACNT and the use of these composites in

The previous sections have revealed the necessity of consid- ering all kinetic trapping-detrapping parameters in diffusion- trapping systems, previously referred to as DTS, along

The largest dif- ferences between the observed and expected limits are found for scenarios of top squark pair production with moderate mass splittings and result from observed

disparition avec acrimonie dans son journal du 19 avril 1815 : « on a enterré aujourd’hui M. François Jourdain, professeur à l’école de dessin. C’était un honnête homme mais

Ce décalage peut s'interpréter comme le signe d'une transition par l'emploi inachevée, à cause en particulier de politiques du logement favorables au secteur social, qui iraient

The pra ti al pseudo-random generators proposed for Prophet and the possibility to a hieve ollision-free dynami address allo ation are dis ussed in Se tion 5.. Finally, we on

Par exemple, si vous assignez une valeur à la propriété Text d’un label, comme nous l’avons fait, le contrôle conservera cette valeur durant toutes les requêtes qui seront