forthenanotexturalcharacterizationofthematerials,otherthantheusuallowt - PDF document

forthenanotexturalcharacterizationofthematerials,otherthantheusuallowtem- peraturenitrogenadsorption.Thisis,forinstance,thecaseoftheadsorptionof moleculesofvolatileorganiccompounds(VOCs),withvariousdimensionsand compounds · LandÞllgases 2.1Introduction Pillaredinterlayeredclays(PILCs)arefarmorestudiedconcerningtheirpreparation methodologiesandtheircatalyticproperties[1,2]thanwhatconcernstheirpossibil- lowerstructuralregularitythan,forinstance,themoreusualzeolites,butPILCs structuralregularityishigherthanotherimportantclassesofadsorbentmaterials, suchasactivatedcarbons.Therefore,thelackofapreciseregularstructurecannot beregardedasamajordrawbackforPILCapplicationsasadsorbents.Furthermore, PillaredClaysandRelatedCatalysts , DOI10.1007/978-1-4419-6670-4_2, C  SpringerScience+BusinessMedia,LLC2010 26J.PiresandM.L.Pinto presentsanisothermwhichisrelativelysimilartotheSiÐTi-PILC.Inthecaseof methanol(azirconiumoxide-pillaredsampleÐZr-PILCÐwasincludedinthiscase), probablybecausethisisthesmallestmoleculeoftheseries,theisothermsarenow moredifferentamongthevarioussolids. WhentheadsorptionofchlorinatedoroxygenatedVOCsisconsidered,the isothermscanbeverydifferentinthesamePILC[9,13,14].But,inthiscaseand becausethesetypesofmoleculeshavenormallyahighdipolemomentorarehighly polarizable,thosedifferencescanbeentirelyattributednotonlytosizeexclusion affectsbutalso,inanon-measurableway,tospeciÞcadsorbateÐadsorbentinterac- tions.Inthisway,theuseofchlorinatedoroxygenatedVOCsasprobemolecules forthecharacterizationofthemicroporositymaynotbeadequate.Theadsorption oforganicgasesandvaporsinPILCshasbeenmodeledaccordingtotheBET[7], Langmuir[7,14]virial[15],DubininÐRadushkevich[12],andDubininÐAsthakov [14]equations. 2.2.2ForSeparation/PuriÞcation ItiswellknownthatanumberofVOCs,liberatedinvariousindustrialactivities,are toxicandcontributetotheatmosphericpollution,withdirectorindirectactionon humantissues[16].TheabatementofVOCscanbemadebythermaloxidation,an efÞcientbuthighlyenergycostlyprocessthat,additionally,isnotentirelyadequate forchlorinatedVOCs[17].ThecatalyticoxidationofVOCsisanimportantproce- durebutadsorptioncan,inthelimit,makepossiblethere-useofaparticularVOC [18].Additionally,sinceVOCsareusuallypresentindilutedairstreams,adsorbents maybeusedtoconcentrateadilutedstreamandthereforeimprovingtheefÞciency ofthermalorcatalyticoxidation[18].Whileactivatedcarbons,duetotheirspread useandmoderate-to-lowcost,areusuallyconsideredasgoodcandidatesasadsor- bents,intheÞeldoftheremovalofVOCsthesematerialspresentsomedrawbacks sincetheirßammabilityposesdifÞcultiestotheirregeneration[19].Zeolitesare anotherfamilyofadsorbentsthatcanbeconsideredfortheVOCsabatement.Some limitationsofzeolites,atleastforthemorecommon,areduetotheirhydrophilic nature,sincehumidityisalsonormallypresentinthestreamswhereVOCsareto beremoved.Zeolitescanbepreparedlesshydrophilic[19],forinstance,bydea- luminationprocesses[19,20],butthisisanadditionalpreparativestepandthe limitingadsorbedamountstendtobelowerintherespectivedealuminatedsamples [20].PillaredclayshaveintermediatehydrophobicÐhydrophilicpropertiesbetween activatedcarbonsandthemorecommonzeolites[21]andcouldbeaninterest- ingalternativeasadsorbentsintheÞeldofVOCsremoval/abatement,atleastif anappropriatedmethodologyofmakingPILCseconomicallyinlargeamountsis developed. Sincetheremovalofnoxiousvolatileorganiccompoundsisusuallytobecon- sideredinthelowpressureregion,Table2.1includestheamountsofselectedVOCs adsorbedinseveralPILCsattherelativepressureof0.1. 2PillaredInterlayeredClaysasAdsorbentsofGasesandVapors27 Table2.1 AmountsofselectedVOCsadsorbedinPILCsattherelativepressureof0.1.When indicatedintheoriginalwork,thestructuralformulaoftheclays,whichwereallmontmorillonites, (perO 10 (OH 2 ))isgiven VOCTypeofpillarsStartingclay Amount adsorbed(cm 3 /g) andreference BenzeneSiO 2 ÐTiO 2 (Si 3.89 Al 0.11 )(Al 1.60 Mg 0.32 Fe 0.08 )0.1[7] BenzeneChromiumoxide0.13[12] BenzeneTitaniumoxide0.14[12] TolueneAluminumoxide0.12[10] CyclohexaneSiO 2 ÐTiO 2 (Si 3.89 Al 0.11 )(Al 1.60 Mg 0.32 Fe 0.08 )0.065[7] n- HexaneSiO 2 ÐTiO 2 (Si 3.89 Al 0.11 )(Al 1.60 Mg 0.32 Fe 0.08 )0.09[7] CCl 4 IronoxideÐzirconium oxide (Si 4 )(Al 1.67 Mg 0.34 Fe 0.08 )0.31[9] CHCl 3 IronoxideÐzirconium oxide (Si 4 )(Al 1.67 Mg 0.34 Fe 0.08 )0.28[9] TrichloroethaneAluminumoxide0.12[10] TrichloroethaneZirconiumoxide(Si 3.70 Al 0.30 )(Al 1.16 Fe 0.51 Mg 0.26 )0.07[13] MethanolAluminumoxide(Si 3.70 Al 0.30 )(Al 1.16 Fe 0.51 Mg 0.26 )0.12[13] MethanolZirconiumoxide(Si 3.70 Al 0.30 )(Al 1.16 Fe 0.51 Mg 0.26 )0.08[13] SomestudiesaddressedtheseparationoforganicmoleculesusingPILCs asselectiveadsorbents[13,22].Indynamicexperiments[22]separationfactors foranaluminumoxide-pillaredclaywithvaluesof2.08for n- hexane/benzene (at523K);1.90for n- hexane/ n- heptane(at594K),and2.68forcyclohex- ane/benzene(at624K)werereported.Themodelingofselectiveadsorption ofVOCs,byamethodologybasedontheDubininÐRadushkevichtheorywas alsoattemptedinmontmorillonitespillaredwithaluminumorzirconiumoxide pillars[13].VariousVOCsweretestedandseparationfactorsnearthree(at acoverageof0.5)forthemethanol/propanonesystemandthemethylethylke- tone/propanonesystemwereobtained.Theeffectsofthetypeofpillarsinthe separationfactorswerenotentirelyclear,althoughthetotalamountsadsorbed werehighestforthealuminumoxide-pillaredmaterials,inlinewiththehigh- estspeciÞcsurfaceareaofthesesamples[13].Chlorinatedsamplessuchasthe 1,1,1-trichloroethane/trichloroethylenewerealsotestedbuttheresultswereless promisinginthiscase[13]. AlimitednumberofstudiesweremadeconsideringtheadsorptionofH 2 S, whichisalsoanoxiousgaseouspollutant.Theseconsideredeitherthestudyof theeffectofthepillaringwithdifferentoxidepillars,mainlyofaluminumor iron[23]ortheparticularcaseofH 2 Sremovalrelatedtotheconservationof worksofart[24].Inthelattercase,amontmorillonitefromWyomingpillared withaluminumoxidepillarsresultedtobemoreefÞcientintheremovalofH 2 S atverylowrelativepressures,inspiteofislowesttotalmicroporousvolume whenconsideredwithothermorecommonadsorbents,suchassodiumXorY zeolites. 28J.PiresandM.L.Pinto 2.3AdsorptionofWater Wateradsorptionstudiesinpillaredclayshavebeenmadebyvariousauthors,essen- tiallywithatwofoldobjective:thepotentialuseoftheseadsorbentsinmoisture removalandalsothecharacterizationofthesurfaceofPILCs.Inthissection,we consideredtheanalysisofpublishedwateradsorptionresultsintermsoftheirpoten- tialitiesasdesiccants,usingsomeindicationsintheliteraturethatmaterialsshould have,inparticularlyiftheenvisageuseisasdesiccantsingas-Þredcoolingsystems anddehumidiÞcation[25].Forthis,oneimportantcriterion,althoughofcoursenot theonlyone,isthatthewateradsorptionisothermmustbeofaÒmoderateÓtypeI. Roughly,theamountadsorbedatarelativepressureof0.1shouldbeabout50%the totalamountadsorbed[25,26]. Figure2.3collectsthemaximumwateradsorbedamounts,estimatedfromthe publisheddataforrelativepressuresbetween0.85and0.95andtherespectiveper- centageofamountadsorbedattherelativepressureof0.1.InthisÞgure,thedata forsamples1Ð4arefrom[26],forsamples5Ð9from[27],forsamples10Ð13from [28],forsamples14Ð16from[29],andforsamples17Ð22from[12].Samples23 and24[8]arenotPILCs,theyare,respectively,YandAzeolites.Thelattersamples wereincludeinFig.2.3toillustratetheeffectofasteeptypeIisotherm,whichis usuallyobtainedforwateradsorptioninthesehydrophiliczeoliticmaterials.Infact, forthesezeolites,theamountadsorbedatarelativepressureof0.1ismorethan 80%(almost90%forAzeolite)ofthetotaladsorbedvolume. AllmaterialsinFig.2.3havealumina-basedpillarsexceptsamples17Ð19[12], whichhavechromiumoxidepillars,andsamples20Ð22[12]thathavetitanium 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 sample amount adsorbed (cm 3 g –1 ) 0 10 20 30 40 50 60 70 80 90 100 % at p /p 0 = 0.1 amount adsorbed % at p/p 0 = 0.1 34567891112 14151619212324 22 20 18 17 13 10 2 1 Fig.2.3 Adsorptioncapacitiesforwaterandrespectiveadsorptionat p / p o = 0.1forvarious pillaredclaysandzeolites(seetextforreferences) 2PillaredInterlayeredClaysasAdsorbentsofGasesandVapors29 oxidepillars.AscanbeseeninFig.2.3,thehighestwateradsorptionamountsseem toberegisterednotforPILCsthathavealuminumoxidepillarsbutforthosethat havechromiumortitaniumoxidepillars. Intheseriesofsamples5Ð9[27],samples5and6arealsoamongthosewith thehighestwateradsorbedamounts.Inthatworktheauthorsstudytheeffectof theprogressiveincreaseofthecalcinationtemperature,instepsof100  Cbetween 200  C,forsample5,and600  C,forsample9,andobservedthatthemaximum adsorbedamountofwaterdecreasesasthecalcinationstemperatureincreases.Inthe caseofthelowestcalcinationtemperature(200  C),thehighestadsorbedamount observedmayalsobeduetothefactthattheoligomericcationicspeciesthatarethe precursorsofthepillarscanbenotentirelytransformedintotherespectiveoxide andtheresidualOHgroupsmaythenincreasetheadsorptionviaspeciÞcinteraction withthewatermolecules.Forallsamplesofthisseries(samples5Ð9)thepercentage adsorbedat p / p o = 0.1isalwayswellbelow50%,puttinginevidencetherelatively lowhydrophilicityofthestudiedsamples. Inthecaseofthesamples1Ð4,theauthors[26]triedtoimprovethepropertiesof thePILCsforwateradsorptionbyincreasingthepercentageoftheamountadsorbed at p / p o = 0.1byvaryingtheagingconditionsofthepillaringsolution.Aslightly betterresultwasobtainedinthesamestudyinaPILC,whereCa 2+ cationswere introducedafterthepillaring/calcinationprocess,wherethatpercentageincreases to32%.Otherauthorshavealsostudiedtheeffectonwateradsorptionproperties ofPILCsbyintroducingcationsafterthepreparationofthePILCs[28],bydirect cationexchangeathighpH,namely,Ca 2+ ,Na 2+ ,Mg 2+ ,andLi + (samples10,11, 12,and13,respectively).AscanbeseeninFig.2.3,themostfavorableresultswere obtainedforCa 2+ ,whileLi + andMg 2+ resultedonlyinamodestimprovement.The authors[28]interpretedthelatterfactasduetothemigrationofthesmallestLi + and Mg 2+ cationsintotheoctahedrallayersoftheclaysheets,accordingtotheknown HoffmannÐKlemeneffect. Waterisapolarmolecule,andaswellstatedintheliteratureforwateradsorption, thecontributionofthedispersionforcestoadsorptionisusuallysmallwhencom- paredwiththecontributionfromtheinteractionswiththedipole[30].Therefore, wateradsorptionhasahighdegreeofspeciÞcinteractionswiththesurface.Inthe presentreviewitisdifÞculttodisclosefromthepublishedresultssomeaspects relatednamelywithdifferencesintheusedstartingclaysthatcouldalsopartially explainthevariationofresultsinFig.2.3.Forthepreparationofthealuminum oxide-pillaredclayscorrespondingtothesamples14Ð16[29]theauthorsusedthree montmorillonitesfromthreedifferentsoildeposits.Fromthesematerials,sam- ple14presentedthebestresults,particularlyconcerningthevalueoftheamount adsorbedat p / p 0 = 0.1,whichapproached50%ofthetotalamount.Thissam- plewaspreparedwithamontmorillonitethathadahighdegreeofsubstitutions inthetetrahedralsubstitutions(ofsiliconbyaluminum),whiletheglobalsilicon toaluminumratioofthethreesampleswasmuchsimilar.Thetetrahedralsheets are,ofcourse,moreexposedtotheinteractionswiththeadsorbedmoleculesand sospeciÞcinteractionscanbedevelopedwiththeadsorbedwatermolecules.This 30J.PiresandM.L.Pinto resultseemstoindicatethatstartingclayswithhighdegreeoftetrahedralsubstitu- tionsmaybebeneÞcialwhenpreparingPILCsformoistureadsorption,particularly forimprovingtheadsorptionpropertiesintheinitialportionoftheadsorption isotherm. 2.4AdsorptionofNaturalandBiogasComponents Oneimportantapplicationofadsorbentmaterialsisingasseparationprocesses, whereamixtureoftwoormoregasesisseparatedtoobtainoneormorepuri- Þedcomponents[31].PILCscanbepotentiallyusedinsuchprocesses,provided thattheypresenttheadequatepropertiesforaneffectiveseparationofagiven mixture.However,comparingwithothertypesofadsorbentmaterials,likeacti- vatedcarbonsandzeolites,relativelyfewstudiesandexperimentaldatacanbe foundontheliteratureabouttheadsorptionofpermanentgasesonPILCs.Works involvingadsorptionatpressuresabovetheatmosphericpressure(high-pressure adsorption)areevenscarcer.Nevertheless,someimportantandillustrativeresults wereobtained,namely,inthestudyofPILCsforthepuriÞcationofnaturalandbio- gascomponents.Thissectionpresentsthepublishedadsorptionresultsonthistopic alongwithadiscussionandanalysisoftheresultsintheviewofpossibleindustrial applications. PILCshavebeenstudiedforthemethane/ethaneseparation[32,33].Thissep- arationhasimportanceinthecontextofthenaturalgasextraction,sincetheseare thetwomainhydrocarboncomponentsofnaturalgas[34].Ethanemaybesepa- ratedforuseasafeedstockforsteamcrackingfortheproductionofethylene.Other highermolecularweighthydrocarbonspresentinnaturalgasesareimportantfuels aswellaschemicalfeedstocksandarenormallyrecoveredasnaturalgasliquids. AnexampleofthemethaneandethaneadsorptionresultsispresentedinFig.2.4for theadsorptiononazirconiumoxidePILC(Zr-PILC).Asexpected,ethaneadsorbs considerablymorethanmethane,sinceitistheheaviercomponent,andatnear atmosphericpressuretheadsorbedamountsofethanearemorethanthreetimes higherthanthoseofmethane.Thevacancysolutiontheory(VST)[35Ð37]wasused topredicttheadsorptionisothermsofmethane/ethanebinarymixturesandsuccess- fullycomparedthemwithsomeexperimentalpoints[32].InFig.2.4,thebinary adsorptionisothermspredictedbyVSTandbytheidealadsorbedsolutiontheory (IAST)[38Ð41]arealsopresentedandasmaybeobservedagreefairlywellforthe gasphasecompositionscompared,althoughVSTtendstopredicthigheradsorption valuesatthesamecomposition.Itisinterestingtonotethat,forequalcomposition inthegasphase(0.5methanemolarfraction),thepredictedisothermsareconsider- ablymoreclosetothepureethaneisothermthantothepuremethane.ThisisaÞrst indicationthatthematerialselectivelyadsorbsethaneovermethane. Forabetterunderstandingoftheadsorptionbehaviorofmixturesinanadsor- bentmaterial,atagivenÞxedpressure,itisusefultocomparethecompositionof thegasphase( y i )withthecompositionoftheadsorbedphase( x i ),byconstructing 2PillaredInterlayeredClaysasAdsorbentsofGasesandVapors31 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 pure ethane 0.25 0.5 0.75 pure methane methane molar fraction n ads /mmol g –1 100 80 60 040 20 p /kPa Fig.2.4 ComparisonbetweenIAST( black )andVST( gray )predictionsofbinarymixture isothermsofmethane/ethaneonazirconiumoxide-PILC,at289K.Experimentalvaluestaken from[32] theso-called x–y phasediagram[31].Forthesamezirconiumoxide-pillaredclay presentedinFigure4.1,itispossibletoobtainthediagraminFig.2.5usingthe IASTmodeladjustedtotheexperimentalresultsobtainedattwotemperatures.In thisÞgure,thecompositionofmethane(molarfractions)inthegasphase( y 1 )and adsorbedphase( x 1 )arerepresentedintheaxis.Inthisway,thecompositionofthe adsorbedphasemaybeknownforagivengascomposition,whichisaveryuseful informationinthecontextofgasseparationsinceitallowstoestimatetheexpected purityoftheseparatedgases.Thetwocurves(oneateachtemperature)showthat thegasphaseisalwaysricherinmethanethantheadsorbedphase,i.e.,theethane tendstoadsorbonthematerialandthemethanetendstostayinthegasphase.The sharperthekneeofthecurve,themoreeffectiveistheseparation.Infact,ifone 0 0.2 0.4 0.6 0.8 1 x 1 y 1 215 K 293 K 1 0.8 0.6 0.4 0.2 0 Fig.2.5 x–y phasediagram ofthemethane/ethane adsorption(compositionof theadsorbedandgasphases) onaZr-PILC,at215and 293Kand100kPa 32J.PiresandM.L.Pinto considersthegasmixturewith0.8inmethane( y 1 )itcanbeseenthatthecompo- sitionoftheadsorbedphaseat215Kisabout0.3inmethane( x 1 )(i.e.,about0.7 inethane),butincreasesto0.55inmethaneat293K.Therefore,Fig.2.5clearly showsthattheseparationismoreeffectiveat215Kthanat293K.Nevertheless, theresultsindicatethatthisPILCcouldbeusedforthemethane/ethane separation. The x–y diagramsareusefulfortheevaluationofamaterialforaspeciÞcsepa- ration,asshownbytheprecedingexample,buttheydonotgiveanyinformationon theadsorbedamountsinagivenmassofadsorbent.Thisinformationisalsoimpor- tantwhendesigningtheseparationequipmentforaparticularapplication,namely, forcalculatingthemaximumadsorbedamountsofaseparationcolumn.Forthis purpose,therepresentationoftheadsorbedamountsasafunctionofthegasphase compositionismoresuited,byconstructingagasphasediagramoftheadsorbed phase.WhenusingsuchrepresentationforthesameZr-PILC(showninFig.2.6)it becomesevidentthattheadsorptioncapacitiesat215and293KaresigniÞcantly different.Theinterceptionofthelinesfortheethaneandmethaneamountsindi- catesthecompositionofthegasphaseatwhichthecompositionoftheadsorbed phaseisequalinthetwocomponents.Thispointmovestolower y 1 valueswhen thetemperaturesincrease,indicatingthattheseparationbecomeslesseffective.The resultsat293KaremoresigniÞcantforpossibleapplications,sincetheyarecloseto ambienttemperaturesandseparationsarenormallypreformednearthesetempera- tures.Intheseconditions,Fig.2.6showsthattheamountsofethanevaryfromabout 0.45mmol/gofthepureethaneadsorptiontoabout0.16mmol/g,at0.79methane molarfaction,andforhighermethaneconcentrationonthegasphasetheamountsof 215 K 0 0.4 0.8 1.2 1.6 total amount Ethane Methane 293 K 0 0.1 0.2 0.3 0.4 0.5 1 1 0.8 0.6 0.4 0.2 0 1 0.8 0.6 0.4 0.2 0 n ads / mmol g –1 n ads / mmol g –1 y 1 Fig.2.6 Phasediagrams (amountsadsorbed)attwo temperaturespredictedby IASTat100kPa,asa functionofthegasphase composition,foraZr-PILC 2PillaredInterlayeredClaysasAdsorbentsofGasesandVapors33 methaneadsorbedbecamehigherthanthoseofethane.Thisindicatestheamounts ofethanethatcouldberecoveredfrommixturesundertheseconditions,usingthis Zr-PILCastheadsorbentmaterial. ThepredictionsshowninFig.2.6maybecomparedwithexperimentalvalues obtainedfortheadsorptionofmethane/ethanemixtures.Whenthepredictionsmade usingtheVSTarecomparedwiththetotaladsorbedamountandgasphasecompo- sitiondeterminedexperimentally,deviationsofabout2Ð3%intheadsorbedamount werefound[32]whichisagoodresultconsideringthattheexperimentalvalueshave aboutthesameuncertainty.Thepredictionspresentedin2.6weremadebytheIAST mayalsobecomparedwiththeexperimentaldata,andinFig.2.7theexperimen- talpointiscomparedwiththepredictions.Ascanbeseen,theexperimentalpointis closetothesolidlineofthetotaladsorbedamount,withadeviationof5%.Thissup- portsthepredictionsmadebyIASTandconÞrmsthattheyareclosetothosemade byVST(alreadyseeninFig.2.4).Alsoveryimportantistheindicationthatthese theoriesaresuitableformodelingtheadsorptionofbinarymixturesinPILC,using thepurecomponentsadsorptionisotherms.Infact,themeasurementsofisotherms forgasmixturesaremuchmorecomplicatedthanforpuregases,duetoequilibrium problemsandtotheanalysisofthegasphasecomposition.Thus,IASTorVSTare importanttoolstotheanalysisoftheadsorptiondata,especiallywhencomparing differentPILCmaterialsforanenvisagedapplication.Morerecently,anadsorption modelbasedonstatisticalmechanicsforthepredictionofmixtureisothermson PILCswasproposed[42]withhigheraccuracyatveryhighpressures,butforthe rangeofpressuresnormallyusedonadsorptionexperimentsitgivessimilarresults toIAST. 298 K 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 Total amount Ethane Methane total amount (Exp.) 1 0.8 0.6 0.4 0.2 0 n ads /mmol g –1 y 1 Fig.2.7 Comparisonofthepredictedadsorbedphasediagramat73.84kPa,withanexperimental point,foraZr-PILC Mosttimes,thereisaninteresttostudytheseparationofagasmixturewith aspeciÞccompositionandthetemperatureorpressureeffectsontheseparation performanceoftheadsorbentmaterials.Onewayistousetheselectivityofthe separation,deÞnedas 34J.PiresandM.L.Pinto S 2,1 = x 2  y 2 x 1  y 1 (2.1) andevaluatethisparameterasafunctionofpressureortemperature.Usuallya separationwithselectivityhigherthanthreeisconsideredappropriateforindustrial applications[31]. Inthemethane/ethaneseparationappliedtonaturalgas,theabundanceofethane inthedistributiongasisonlyabout8%(i.e., y 1 = 0.92),andthisvaluecanbe usedasreferencetotheevaluationoftheselectivityofPILC.Figure2.8shows theselectivityfortheseparationofethanefrommethane,atthisreferencecom- position,usingzirconiumoxide-pillaredclayssynthesizedfromtwodifferentclays [32].Oneclay,labeledasPTS,isfromPortoSanto(Madeiraarchipelago,Portugal), andtheother,labeledasBEN,isfromBenavila(Alentejo,Portugal).Itisevident fromFig.2.8thatselectivitytowardethanedecreases,aspressuresandtemperatures increase,althoughatthehighertemperature(293K)theselectivityisalmostinsen- sitivetothepressure.ItisinterestingtonoticethattheselectivityvaluesforBEN andPTS-pillaredclaysarenotverydifferent.Nevertheless,inthemajorityofcondi- tions,selectivityishigherforthePTSsample,afactthatisrelated,mostprobably, tothehigherheatofadsorptionofethaneonthispillaredclay[32].Theselectiv- ityvaluescalculatedusingIASTandVSTpresentsomediscrepanciesat215Kfor thesamematerials,beingtheIASTvalueshigherthanthosecalculatedbyVST. However,theyqualitativelyagreeintheorderoftheexpectedselectivityforthe separation,andat293Kthevaluesareidentical.Fromthepracticalpointofview, thepotentialapplicationofpillaredclaysintheseparationofthemajorhydrocarbon componentsofnaturalgas(methaneandethane)isconÞrmed,sincetheselectivity coefÞcients,betweensixandeightat293K,comparewellwiththoseobtainedwith otheradsorbents[43]. 0 40 80 20406080100 p / kPa S PTS 215 (IAST) PTS 293 (IAST) BEN 215 (IAST) BEN 293 (IAST) PTS 215 (VST) PTS 293 (VST) BEN 215 (VST) BEN 293 (VST) Fig.2.8 SelectivityofZr-PILC(PTSandBEN)predictedbyIAST( black )andVST( gray ),at twotemperatures,forthemethane/ethaneseparation,at0.92methanemolarfraction 2PillaredInterlayeredClaysasAdsorbentsofGasesandVapors35 Anotherimportantgasseparation,whichhasalsobeenstudiedusingPILCsas adsorbents,isthecarbondioxide/methaneseparation.Thisseparationcanbeapplied forthepuriÞcationofnaturalgas,biogas,andlandÞllgas,whichareformedwitha considerableamountofcarbondioxidealongwiththecombustiblepart[34,44Ð46]. InFig.2.9theadsorptionresultsontwoPILCsusingthesamePortoSantoclay (PTS),butdifferentpillars,arepresented[47,48].TheZr-PTSpresenthigher adsorptioncapacityformethaneandcarbondioxideintherangeofcomparableval- ues(0Ð100kPa),althoughthetwomaterialshaveaboutthesamesurfacearea.This indicatesthatthenatureofthepillarsstronglyinßuencestheadsorptionbehavior ofthesegases,eitherbydifferencesonthechemicalnatureofthesurface(surface charges,acidity,etc.),orbydifferencesinthepillarÕsdensityandsize(i.e.,the sizeofthemicropores).Inarecentwork,aluminumandzirconiumoxide-PILCs preparedusingWyoming(WYO)andBenavila(BEN)clayshavebeenstudiedfor carbondioxide/methaneseparation,upto1,000kPa[33].Theresultsshowedthat thezirconiumPILCspresentedthehighestcarbondioxideadsorbedamountsper unitsurfaceareaofthestudiedmaterials,duetothemoreacidicnatureofZroxide- PILCsandtheirinteractionwiththequadrupoleofthecarbondioxide.However, whenthecomparisonismadebytheadsorbedamountintermsofthemassof materialonly,thecomparisonisnotbesoevident.Infact,somealuminum-PILC presentedhigheradsorbedamountsthananotherzirconium-PILC(Zr-BEN),due totheirhighersurfaceareasandmicroporousvolumes.Therefore,theworkcon- cludesthatthemorefavorableresultsobtainedforZr-WYOPILCisanoutcome ofthechemistryofZrO 2 pillarsintheWyomingclayhost.Thispointsoutthe difÞcultiesofdrawingsomegeneralconclusionsaboutthepillarsthataremore suitableforagivenseparation,sincetheywillalsodependontheparentclay usedinthepreparation.Thissubjectwillbefurtherdiscussedwithmoredetail hereafter. 0 0.25 0.5 0.75 1 1.25 p / kPa n ads /mmol g –1 Carbon Dioxide-Zr Methane-Zr Carbon Dioxide-Al Methane-Al 0100200300400500 Fig.2.9 Carbondioxideandmethaneadsorptionisotherms,onPILCsobtainedwithAlandZr oxidepillarsat298K 36J.PiresandM.L.Pinto Consideringapplicationsusingpressureswingadsorption[31],whichisthemost usedtypeofcyclicgasadsorptionseparation,thematerialsusedinthisprocess shouldbeeasilyregeneratedintheendoftheadsorptionstep,preferablybya simpledecompressionofthesystem.Thefactthattheadsorptionisothermsofcar- bondioxideandmethaneonPILCsaremosttimesslowrisingcurves,likethose presentedinFig.2.9,isfavorablefromthisviewpoint.Thisisanadvantageover manyzeolites,sincetheseusuallyhaverectangularshapedisothermsandbecome almostsaturatedwithcarbondioxidebelowatmosphericpressure,hamperingthe effectiveregenerationbyasimpledecompressionstep.Takingintoaccountthe exampleforthealuminumoxide-PILCinFig.2.9,onecanseethattheadsorp- tionamountrisesfromabout0.5mmol/gat100kPa(atmosphericpressure)tomore than1mmol/gat500kPa,givingaworkingcapacityof0.5mmol/gforthatpres- surecycle.Similarworkingcapacityforpressureswingadsorptionandvacuum swingadsorptionresultswerealsorepostedforotherPILCs,supportingthepossible applicationofPILCsforcarbondioxide/methaneseparation[33]. ForabetteranalysisoftheadsorptionresultspresentedinFig.2.9, x–y phasedia- gramsandselectivitycoefÞcientswerecalculatedusingtheIASTandarepresented inFigs.2.10and2.11.BothÞguresshowthatthePILCwithaluminumoxidepillars ismoreeffectivefortheseparationunderstudy,sinceitallowsobtainingmethane 0 0.2 0.4 0.6 0.8 1 00.20.40.60.81 x 1 y 1 Zr Al Fig.2.10 Zr-PTSand Al-PTS x–y phasediagrams ofcarbondioxide/methane mixture,at100kPaand 298K(component1Ð methane) 1 6 11 16 050100 p / kPa S Al Zr Fig.2.11 Selectivity coefÞcientsasfunctionof pressure,at y 1 = 0.5 (methane),forthetwoPILCs preparedfromPTSclay,at 298K 2PillaredInterlayeredClaysasAdsorbentsofGasesandVapors37 withaslightlyhigherpurity.Itisinterestingtonotethatthesimplecomparisonof theisothermswouldclearlyleadtothisconclusion.Infact,Zr-PTSpresentedthe highestadsorptionamountsforbothmethaneandcarbondioxide.However,itisthe relationbetweenthesetwoamountsthatinßuencestheselectivityoftheseparation. TheresultspresentedinFigs.2.10and2.11areinlinewiththoseobtainedwiththe sametypeofpillarsinotherclays[33]. Asalreadymentioned,mostgasseparationsarecarriedatpressuresaboveatmo- spheric,andthereismuchinteresttostudytheadsorptionofgasesinthispressure range.However,thereisscarcedataintheliteratureonhigh-pressureadsorptionof gasesonPILCs.Recently,theÞrstcomparisonworkonthehigh-pressureadsorp- tionofmethane,ethane,carbondioxide,andnitrogenondifferentaluminumand zirconiumPILCs[33]waspresented.Thisworkwasacontributiontoevaluatethe realapplicabilityofPILCsfortheseparationofbinarymixturesofthestudiedgases, whichisimportantinthecontextofthenatural,bio,andlandÞllgasupgrade.For thestudiedbinarymixtures,thepresentedresultsindicatethepossibleapplication intheseparationofmethane/ethane,methane/carbondioxide,nitrogen/ethane,and nitrogen/carbondioxide.Thelatterseparationhasalsointerestforapplicationsin carbondioxideemissionsabatement. Inadditiontotheapplicabilityevaluation,theresultsobtainedinabroadpressure rangealsogivetheopportunitytostudytheevolutionofthephasediagrams,similar tothoseinFigs.2.6and2.7,butnowasfunctionofpressure.Thisapproachleads totri-dimensionalrepresentationofthephasediagramsasinFig.2.12.ThisÞgure showsthephasediagramforazirconiumoxide-pillaredclay,whichpresentedthe bestresultsofthestudiedPILCsforthemethane/carbondioxideseparation.The Fig.2.12 Phasediagram(compositionoftheadsorbedphase)predictedbyIAST,asafunctionof thegasphasecompositionandpressureforazirconiumoxide-PILC.The black surfacerepresents thetotalamountadsorbed,thelower gray meshrepresentstheadsorbedmethaneamountsandthe upper gray meshrepresentstheadsorbedcarbondioxideamount