Experimental and theoretical study of methane adsorption on granular activated carbons

Carbons

YuguoWang,CemalErcan,AnwarKhawajah,andRashidOthman

Research&DevelopmentCenter,SaudiArabiaOilCo.,Dhahran,31311,SaudiArabia

DOI10.1002/aic.12611

PublishedonlineApril27,2011inWileyOnlineLibrary(wileyonlinelibrary.com).

Theexperimentalandtheoreticalstudyofmethaneadsorptionongranularactivatedcarbonsispresented.Theadsorptiondataaremodeledbyvariousisothermequations.Tothequationisfoundtohavethebestfit.Theisostericheatdecreaseswithloadingandincreasesweaklywithtemperature,whichisanindicationofheterogeneityofthemethaneandgranularactivatedcarbonsystem.UsingoptimizedparametersfromTothequation,anovelprocedureisdevelopedtocalculatetheintegralheatofadsorption,whichisthetotalamountofisostericheatofadsorptionatagiventemperatureand

C2011AmericanInstituteofChemicalEngineerspressureduringtheadsorptionprocess.V

AIChEJ,58:782–788,2012

Keywords:methaneadsorption,Tothequation,activatedcarbon,isostericheat,

integralheatofadsorption

Introduction

Dependingonthesourceandgeographicallocationofpro-duction,naturalgascancontainupto95mol%methane(CH4),theremainingbeingcarbondioxide(CO2),nitrogen,andsmallamountsofhighermolecularweighthydrocarbons,suchasethane(C2H6),propane(C3H8),andbutane(C4H10).Itiswell-knownthatnaturalgasburnscleanerthangasoline,dieselandtheotherfuels.Becauseofthisenvironmentallyfriendlybehavior,itiscommonlyusedforheatinganditsusecontinuestogrow.Itisalsousedinsomevehiclesbystoringitascompressednaturalgas(CNG)atpressuresupto270atm.Incurrentpractice,naturalgasismainlystoredasCNGforvehicleuseorliquefiednaturalgas(LNG)forshipment.However,CNGrequiresexpensivevesselsandmultistagecompressionandLNGalsorequiresexpensivecryogenicprocess.AnattractivealternativetoCNGandLNGmightbeadsorbednaturalgas(ANG),wherethegasisstoredonporousmaterialpackedintoavesselatmuch

CorrespondenceconcerningthisarticleshouldbeaddressedtoC.Ercanatcemal.ercan@aramco.com.

C2011AmericanInstituteofChemicalEngineersV

lowerpressure.ANGusesmicroporousadsorbentsinsidea

vessel,whichoffershigherenergydensityandhighervol-umetovolume(v/v)storagecapacitycomparedwithCNGattypicalnaturalgaspipelineconditions,pressurelessthan50barsandtemperaturelessthan55󰀁C.1–9So,atthesecon-ditions,ANGhashighpotentialforexploitationnotonlyintransportationbutalsoinlarge-scaleapplicationsuchasANGstorageclosetonaturalgasconsumers.

InANGoperation,themostimportantistheselectionordevelopmentofamicroporousmaterialwithhigh-storagecapacityandstabilityundercyclicoperation.Potentialmicroporousmaterials,whichcanmeettheserequirements,includeactivatedcarbon,metal-organicframeworks(MOF),andotherorganicorinorganicsolids.10–20AlthoughMOFsandtheotherorganicsolidsdisplayattractivesorptionprop-ertieswithav/vadsorptioncapacityashighas230ofabso-lutemethaneadsorptionat290Kand35bar,whichalsoexceedstheDOEtargetof180v/v,methaneisstronglyentrappedinMOF’sstructuresandheatingupto100󰀁Cisnecessaryfordesorption.13Theotherfactors,whichmayalsoaffecttheapplicabilityofMOFsasnaturalgasadsorb-entincludetheirstabilityandabilitytotolerateimpuritiessuchasH2S,blackpowder,mercaptans,etc.,whichare

AIChEJournal

782March2012Vol.58,No.3

Figure1.Schematicdiagramoftheexperimentalsetup.

commoninindustrialapplications.Inorganicsolidslikezeo-liteshavelowerperformanceformethanestoragethanactivatedcarbons.Moreover,zeolitespresentmuchmorehydrophilicsurfacesthancarbonsandtendtoadsorbwaterpreferentially.Therefore,activatedcarboncurrentlyremainstheonlycommerciallyviableadsorbentfornaturalgasstor-ageintermsofadsorptioncapacityandstability.

Inthisarticle,bothexperimentalandtheoreticalinvestiga-tionswerecarriedouttodeterminethemethaneadsorptioncapacityofvariousgranularactivatedcarbonsundertempera-turerangeupto56󰀁C.Temperature-dependentSips,TothandUnilanisothermequationswereusedtofittheadsorptiondata.D-Aequationwasusedtooptimizeparametersforadsorptioncharacteristiccurve.Anovelprocedureisalsodevelopedtopredictthetotalamountofisostericheatofadsorptionprocess.

characteristicswereused.Thehigh-puritymethaneandultra-high-purityheliumwereusedwithoutfurtherpurification.

Characterizationofactivatedcarbons

MicromeriticsASAPsurfaceareaandporosityanalyzerwereusedforN2adsorption/desorptionat77K.Thephysi-calpropertiesmeasuredareinTable1.Here,thetotalporevolumefromnitrogenporosimeterisfortheporesthathaveadiameterrangeof0–20nm,andthemicroporesaredefinedastheporesthathavediametersrangeof0–2nm.Forthemercuryporosimeter,mesoporevolumeisforporeswithawidthrangeof2–50nm,andmacroporesareporeswithawidthrangeof50–10,000nm.

TheamountN2adsorbedvs.relativepressureat77KongranularactivatedcarbonsareshowninFigure2.TheshapesofthesefourisothermsareType-IIisotherm,whichindicatesthatthesegranularactivatedcarbonsareessentiallymicropo-rous.Aftertherelativepressureof0.8,theincreaseofadsorbedamountismostprobablyduetothepresenceofmesopores,wherethecondensationofN2occursasthepres-suregoesup.ItisalsonotedthattheisothermplateauisreachedatP/P0¼0.5.Theslopeoftheplateauisrelatedtothemultilayermechanismofadsorptionontheexternalsur-faceofthematerials.Itisalsoobviousthatthehighersur-faceagranularactivatedcarbonhas,thehighermasstomass(m/m)adsorptioncapacityithasfornitrogen.

Table1.PhysicalPropertiesoftheGranularActivated

CarbonSamples

SampleASTMmeshsize

AC18Â16

AC230Â70

AC32Â60

AC412Â400.542.0599990.5000.45620.640.3140.1370.176

Experimental

Experimentalsetupandprocedure

Inthisstudy,thevolumetricmethodwasusedtomeasurethemethaneadsorptionongranularactivatedcarbons.ThemajorcomponentsoftheexperimentalsetupareshowninFigure1.Avacuumpump,notshowninFigure1,isalsoconnectedtothesystemfordegassing.Theadsorberhasathermaljacketconnectedtoaheater/chillersothatthetem-peratureoftheadsorbercanbesettothedesiredtemperatureformeasurement.

Todeterminetheadsorptionisotherm,theadsorberwasfirstfilledupwith50gofgranularactivatedcarbonandwasdegassedatapressureof2.5Â10À4Torrat120󰀁Cfor4h.Then,theadsorberwascooleddowntoandmaintainedatdesiredtemperaturesof10,21,38or56󰀁C.Thereferencevesselhasavolumeoftwoliters,andtheadsorberhasavol-umeof120cm3.Thereferencevesselwasfirstchargedwithmethanetoacertainpressure,then,valve3wasopenedandthepressurebetweentwovesselswaslefttoequalize.Read-ingsofthetwopressureindicatorsweretakenafter20minofreachingequilibrium.

Materials

Fourtypesofcommercialgranularactivatedcarbons,la-beledasAC1,AC2,AC3andAC4,withdifferentphysical

AIChEJournal

March2012Vol.58,No.3

Bulkdensity(g/cm3)0.470.390.49

32.2992.3632.402Skeletaldensity(g/cm)

NitrogenPorosimetry(77K)

123515891426BETsurfacearea(m2/g)

TotalPoreVolume(cm3/g)0.6290.7470.599

0.6000.7060.560MicroporeVolume(cm3/g)˚BJHaverageporewidth(A)18.0018.7017.47MercuryPorosimetry

0.3880.4570.360Totalporevolume(cm3/g)

30.1820.2050.164MesoporeVolume(cm/g)

0.2050.2520.194MacroporeVolume(cm3/g)

PublishedonbehalfoftheAIChEDOI10.1002/aic783

Figure2.N2adsorptionisothermsat77Kongranular

activatedcarbons.

Determinationofadsorptionofmethane

Aftertheadsorberwasfilledwithgranularactivatedcar-bon,itwastappedgentlyuntilthelevelofthegranularacti-vatedcarbondidnotgodownanymore.Atthispoint,thegranularactivatedcarbonwaspackedtoitsbulkdensity.Thetotalvolumeoftheadsorbercell(Vabsorber)consistsofthefollowingvolumes:adsorbentskeletal(Vsk),micropore(Vmic),mesopore(Vmes)andmacropore(Vmac),andthespaceamongtheadsorbentparticles(Vinterp).Forprecisedeterminationofthetotalvolumeofmicropore,mesopore,macroporeoftheadsorbentandthespaceamongadsorbentparticles,heliumwasusedsinceitisessentiallyinertforadsorption.Todeterminetheadsorptionamount,pressureP1wasrecordedforthereferencecellwithavolumeofVref,andequilibriumpressureP2wasalsorecordedafterthegatevalve3wasopen.Then,thetotalvolumebecomes(VT¼VrefþVabsorber).Thenonskeletalvolumeisdefinedas

Vnsk¼VTÀVsk

(1)

granularactivatedcarbons.AllofthemhaveType-Iiso-therm,whichindicatesthatthematerialsaremicroporous.Basedonthemolesofmethaneadsorbedpergramofgranu-laractivatedcarboninFigure3b,thecapacityformethaneadsorptionforfourgranularactivatedcarbonsincreaseintheorderofAC4\\AC1\\AC3\\AC2.ThisorderisthesameastheorderofBETsurfaceareaofAC4(999m2/g)\\AC1(1235m2/g)\\AC3(1426m2/g)¼AC2(1589m2/g).ThisisconsistentwiththeorderoftheamountofN2adsorbedat77KonthesesamplesasinFigure2.Thehigher-surfaceareaagranularactivatedcarbonhasthemoremolesofadsorbateitadsorbs.Basedonthevolumeofmeth-aneadsorbedperbulkvolumeofgranularactivatedcarboninFigure3a,thecapacityformethaneadsorptionincreasesintheorderofAC4\\AC2\\AC1\\AC3.Thiscapacityformethaneadsorptionchangecanbeattributedtothefactthatbulkdensityinvolumetovolumecomparisonbecomesamoreimportantfactor.Thebulkdensityforthefourgranu-laractivatedcarbonsareAC1(0.47g/cm3),AC2(0.39g/cm3),AC3(0.49g/cm3)andAC4(0.54g/cm3).AlthoughAC2hasthelargestBETsurfacearea,ithasthelowestbulkdensity.Then,forthesamevolume,AC2willhavethelow-estmasspacked.AC3hasthesecondhighestBETsurfaceareaandbulkdensity,andAC4hasthehighestbulkdensity,butlowestBETsurfacearea.Therefore,itisreasonabletosaythatthesynergeticeffectbetweenBETsurfaceareaandbulkdensitymakesAC3havethehighestvolumetovolumeadsorptioncapacity.

Assumingthatthemethaneadsorptiononlytakesplaceinmicropores,methanemoleculesinsidethemesoporeandmacroporesbehavelikeingasstate.Gas-phasemethanevol-umeis,therefore(VnskÀVmic).Tocalculatetheamountofmethaneadsorbedpergramofgranularactivatedcarbon,nonidealbehaviorisconsideredandEq.2isused.Thisvolu-metricmeasurementofadsorptionissimilartothemethodsappliedbyGummarandTalu21andSalehietal22h

nadsorbed¼

P1VTZRTÀ

P2ÂðVnskÀVmicÞ

ZRTi

(2)

Mgac

whereZismethanecompressibilityfactor,Tistemperature,andRistheidealgasMgacisthemassofthegranularactivatedcarbonused.Zisdeterminedfromthewell-knownSoave-Redlich-Kwong(SRK)equationofstate.

ResultsandDiscussionExperimentalresults

Figure3aand3bshowtheplotsoftheamountofadsorbedmethanevs.pressureat21󰀁Cforthefourdifferent

784

DOI10.1002/aic

PublishedonbehalfoftheAIChE

March2012Vol.58,No.3

AIChEJournal

˚amongAC3hasthelowestaverageporewidthof17.47A

thefourgranularactivatedcarbonsascalculatedbyBarret-Joyner-Halenda(BJH)method(Table1).Theaveragepore

˚)\\AC1(18.00widthincreasesintheorderofAC3(17.47A

˚)\\AC2(18.70A˚)\\AC4(20.64A˚).EarliertheoreticalA

23–25studiesfoundthatformethaneadsorptiononslit-pore

˚toactivatedcarbon,theoptimumporewidthis11.2–11.4A

createthemaximumdensityfortheadsorbedphase.SincetheporewidthofAC3isclosetotheoptimumporewidth,ithasthehigheradsorptiondensity,whichcontributestoAC3’shighestvolumetovolumecapacityformethaneadsorption.

SinceAC3performsbetterthantheothersonvolumetovolumebasedcomparisonformethaneadsorption,intheremainingpartofthearticle,thedataforAC3willbeusedinthediscussionandmodeling.Figure4showsthetempera-tureeffectonmethaneadsorptiononAC3.Itisclearthatthehighertheadsorptiontemperature,theloweramountofmethaneadsorbed.Thisisduetothefactthatadsorptionisexothermic.AccordingtotheLeChatelier’sprinciple,theendothermicdesorptionwillbefavoredwhentemperatureincreases.Therefore,lessamountofmethaneisadsorbedathighertemperatures.

Tocalculatethestoredamountofmethaneinsidetheadsorbervessel,theamountofthemethanecompressedinsidethemesopores,macroporesandthespaceamonggran-ularactivatedcarbonparticlesisaddedtotheliquid-likeadsorbedphaseinsidemicropores.Figure5comparesthemethanestoredforANGandCNGatvariouspressuresat21󰀁C.Atabout50bars,theadsorberpackedwithAC3canadsorbandstore90and120timesthevolumeoftheadsorbercell,respectively.ThesameadsorberpackedwithAC3onlyrequiresapressureofabout10barstostorethesameamountofmethanebycompressionat50bars.So,theadvantageofadsorbedmethanestorageisobvious.

itdoesnotgiveproperHenrylawbehavioratlow-pres-sure,27anddoesnothaveafinitelimitwhenthepressureissufficientlyhigh.Itisgenerallyvalidinthenarrowrangeoftheadsorptiondata.AlthoughtheSipsequationaddressestheproblemofFreundlichequationofcontinuingincreaseoftheadsorbedamountwithanincreaseinpressure,itdoesnothaveaproperHenrylawbehavioratlow-pressure.How-ever,Tothisothermequationsatisfiesboththelow-andhigh-endrequirementanddescribesmanysystemswithsub-monolayercoverageverywell.Unilanisothermequation(uniformdistributionandLangmuirlocalisotherm)assumesthattheenergydistributionisuniformandlocalLangmuirisothermisapplied.

Here,methaneadsorptiononAC3wasmodeledwithtem-perature-dependentformofSips,TothandUnilanequations.TheequationsandtheiroptimizedparametersarelistedinTable2.Nonlinearleast-squaremethodisusedtofittheex-perimentaldataandin-housedevelopedMATLABprogramisusedtodrivetheMATLABoptimizationtoolboxsolverlsqcurvefit.TheaveragerelativeerrorforeachmethodisalsolistedinTable2.TheaveragerelativeerrorisdefinedinEq.4

󰀁P󰀁󰀁YiexpÀYimodeled󰀁

󰀁Yexp󰀁i

AveragerelativeerrorðAREÞ%¼

N

(4)

whereNisthenumberofexperimentaldatapoints,super-scriptsexpandmodeledstandfortheexperimentalandmodeledvalues,respectively,Yrepresentstheamountofmethaneadsorbed.

InTable2,then0,tandsintheSips,TothandUnilanequationscharacterizetheheterogeneityofthemethane-granularactivatedcarbonsystem;Cl,Cls,Cls0areadsorbedamount,saturationadsorbedamountandsaturationadsorp-tionamountatthereferencetemperature;b,b0,b,areadsorptionaffinityconstant,adsorptionaffinityconstantatreferencetemperatureandadsorptionaffinityconstantataverageadsorptionenergy,Risgasconstant,Pisadsorptionpressurex,a,Pareparameters,andQisheatofadsorption.QforSipsandTothequationsarequitedifferentfromeachother.ThisshouldnotcauseanyproblembecauseQintheSipsequationistheisostericheatofadsorptionata

EmpiricalModeling

Manysemiempiricalapproacheshavebeenproposedandtheyarequitesuccessfulindescribingequilibriumdata.26TheseisothermequationsincludeFreundlich,Sips,TothandUnilanequations.TheFreundlichisothermequationisnotvalidatthelow-andhigh-endofthepressurerangebecause

Table2.IsothermEquationsUsedtoModeltheExperimentalDataandtheirOptimizedParameters

EquationNameSips

EquationExpressionCl¼Cls

1nðbPÞ=

1=n1þðhbPÞÀÁiQT0b¼b0expRTÀ10

ÀTT0Á11

ÀTn¼n0þah1󰀆󰀇i

T

Cls¼Cls;0expz1ÀT0

OptimizedParametersusing294.15Kasreferencetemperature

Cls,0¼10.794mmole/g

v¼0

b0¼0.023753barÀ1Q¼10.516kJ/mole

n0¼1.6512a¼0.64163

Cls,0¼17.934mmole/g

v¼0

b0¼0.12929barÀ1Q¼20.259kJ/mole

t0¼0.42718a¼0.2573

Cls,0¼28.115mmole/g

v¼0

b0¼0.001025barÀ1󰀁¼15:803kJ=moleE

s¼6.1961

Emax¼30.956kJ/moleEmin¼0.6505kJ/mole

ARE(%)1.7563

Toth

bPCl¼Cls1s

bPÞt󰀄=

h½1þðÀÁiQT0

À1b¼b0expRTT0

ÀÁ0

t¼tDþah1󰀆ÀTT󰀇i

TCls¼Cls;0expv1ÀT0

1.3047

Unilan

󰀆󰀇

󰀁sPþbe

Cl¼Clsln11þ󰀁ÀsPbehÀÁi󰀁T0E󰀁b¼b0expÀRT01ÀT󰀁¼EmaxþEminE

s¼

Cls¼Cls;0expv1À

2EmaxÀEmin

2hRT󰀆T

T0

2.9886

󰀇i

fractionalloadingof0.5,whileQintheTothequationistheisostericheatofadsorptionatzerofractionalloading.Thevaluesofn0,tandsinTable2allindicatetheheterogeneityofthemethane-granularactivatedcarbonsystem.26Asindi-catedinTable2,Tothequationgivesthehighestaccuracyinfittingtheexperimentaladsorptiondata.Figure6showsthattheoptimizedtemperature-dependentTothequationfitsex-perimentaldatawellinallthepressurerangeatdifferenttemperatures.

equationisusedtocalculatetheisostericheatofadsorption.UsingClausius-Clapeyronequation(Eq.5),theequationforisostericheatofadsorptionisderivedfromtheTothequationasshowninEq.6

ÀDH¼RT2ð@lnP=@TÞh(5)8679

67=<7Â1bPtÃ65ÀDH¼QÀðaRTOÞlnðbPÞÀ1þðbPÞln4ÀÁ1=:;t1þðbPÞtt

#(\")

1hlnh¼QÀðaRTOÞlnÀt1=tð1ÀhtÞtð1þhÞ82󰀆󰀈󰀇39>>>>Cl<=ln67C1Clsl67¼QÀðaRT0Þln4󰀆󰀇1=tÀ󰀆󰀆󰀈󰀇t󰀇5>>t>>1ÀClCls:;CtlsÀCtl

(6)

Theisostericheatofadsorption(Eq.6)isafunctionof

loadingofadsorbatesorpressure.Now,themeaningofQinEq.6isclear;itistheisostericheatwhenthefractionalloadingiszero.Figure7showsthevariationoftheisostericheatofadsorptionforAC3withloadingatfourdifferenttemperatures.Thevalueofisostericheatofadsorptiondeter-minedbyEq.6forAC3isclosetothosereportedvaluesof20KJ/molintheliterature.30Thedecreaseoftheisostericheatwithloadingphysicallymeansthatmethanemoleculesprefertoadsorbontothesitesofhigh-energy.Then,asadsorptionprogressesmethanemoleculesadsorbontothesitesoflow-energy,whichresultsinaslowincreaseintheamountofadsorbedvs.pressure.ThisfindingisalsoinagreementwiththeslopeofadsorptionisothermasindicatedinFigure6.

March2012Vol.58,No.3

AIChEJournal

IsostericHeatofAdsorption

Theknowledgeoftheadsorptionequilibriumandisostericheatofadsorptionisessentialforproperdesignandopera-tionofanygas-phaseadsorptionprocess.Theisostericheatofadsorptionisusuallyestimatedfromthetemperaturede-pendenceoftheadsorptionisotherm.19,28,29AsdiscussedintheExperimentalresultssection,theoptimizedTothequationgivesthebestresultinfittingexperimentaldata.Here,this

Figure6.ModelingandexperimentaldataforAC3at

differenttemperatures.786

DOI10.1002/aic

PublishedonbehalfoftheAIChE

highertemperatureislessthanthatatlowertemperature.Thisisbecauseofthelessamountofmethaneadsorbedathighertemperature.

Conclusions

Thedetailedanalysisofhigh-pressuremethaneadsorptionandstorageresultsatthetemperatureandpressurerangeof10–56󰀁C,and0–50barshasallowedustodrawthefollow-ingconclusions.

PhysicalcharacteristicsofgranularactivatedcarbonsuchasBETsurfacearea,microporevolume,packingdensityandpore-sizedistributionallplayimportantroleindeterminingtheamountofmethaneadsorbed.

Isostericheatofadsorptionofmethaneongranularacti-vatedcarbonsincreaseswithtemperatureanddecreaseswithloading,whichindicatestheheterogeneityofthemethane-granularactivatedcarbonsystem.

Aprocedureisdevelopedtocalculatetheintegralheatofadsorptionintheadsorptionprocess.Theintegralheatofadsorptionincreaseswithdecreaseofadsorptiontemperature.Thehigheramountofadsorptioncanaccountforthiseventhoughaslightlowerisostericheatofadsorptionatloweradsorptiontemperature.Thisprocedurecanbeusedtopre-dicttheintegralheatofadsorptionreleasedintheisothermaladsorptionprocessanditisimportantinformationforthedesignandoperationofanindustrialadsorber.

IntegralHeatofAdsorption

Equation6canbenumericallyintegratedtogettheinte-gralheatofadsorption,31whichisthetotalamountofheatofadsorptionreleasedduringtheadsorptionprocess.There-fore,thefollowingprocedurecanbedevelopedtoestimatetheintegralheatofadsorption,whichneedstoberemovedfromanadsorberbedtocontrolthebedtemperatureconstantduringadsorptionprocess:

(a)Experimentalmeasurementofadsorptionisothermsatdifferenttemperatures.

(b)Useexperimentaldatatooptimizetheparametersofthetemperature-dependentTothequationparameters.

(c)UseTothequationtopredicttheamountofadsorptionatacertaintemperatureandpressure.

(d)Integrateisostericheatofadsorptioncurvetogettheintegralheatofadsorption—thetotalamountofisostericheatofadsorptionuptoacertainpressureatacertaintem-perature.

Usingtheaforementionedprocedureandassumingacon-stantbedtemperature,theintegralheatofadsorptionheatreleasediscalculatedforanadsorberpackedwith1kgofAC3atatemperaturerangeof10–60󰀁C,andapressurerangeof0–50bars.AlthoughFigure7showsthatisostericheatisslightlyhigherathigheradsorptiontemperature,Figure8showstheintegralheatofadsorptionreleasedat

LiteratureCited

1.CookTL,KomodromosC,QuinnDF,RaganS.Adsorbentstoragefornaturalgasvehicles.In:BurchellTD.CarbonMaterialsforAdvancedTechnologies.Oxford:ElsevierScience,Ltd;1999:269–302.

2.LuX-C,LiF-C,Ted,WA.AdsorptionmeasurementinDevonianshales.Fuel.1995;74:599–603.

3.ChangKJ,TaluO.Behaviorandperformanceofadsorptivenaturalgasstoragecylindersduringdischarge.ApplThermEng.1996;16:359–374.

4.InomataK,KanatawaK,UrabeY,HosonoH,ArakiT.Naturalgasstorageinactivatedcarbonpelletswithoutabinder.Carbon.2002;40:87–93.

5.WegrzynJ,GurevichM.Adsorbentstorageofnaturalgas.ApplEnergy.1996;55:71–83.

´S.Efficientdynamicchargeanddischargeofan6.GoetzV,Biloe

adsorbednaturalgasstoragesystem.ChemEngComm.2005;192:876–896.

7.MenonVC,KomarneniS,Porousadsorbentsforvehicularnaturalgasstorage:areview.JPorousMater.1998;5:43–58.

´S,GoetzV,MauranS.Dynamicdischargeandperformance8.Biloe

ofanewadsorbentfornaturalgasstorage.AIChEJ.2001;47:2819–2830.

9.BasumataryR,DuttaP,PrasadM,SrinivasanK.Thermalmodelingofactivatedcarbonbasedadsorptivenaturalgasstoragesystem.Carbon.2005;43:541–549.

10.CelzardA,FierroV.Preparingasuitablematerialdesignedfor

methanestorage:acomprehensivereport.EnergyFuels.2005;19:573–583.

11.PfeiferP,Ehrburger-DolleF,RiekerTP,GonzalezMT,Hoffman

WP,Molina-SabioM,Rodriguez-ReinosoF,SchumidtPW,VossDJ.Nearlyspace-fillingfractalnetworksofcarbonnanopores.PhysRevLett.2002;88:115502–1–115502–4.

12.ShaoX,WangW,ZhangX.Experimentalmeasurementsandcom-putersimulationofmethaneadsorptiononactivatedcarbonfibers.Carbon.2007;45:188–195.

13.SekiK.Designofanadsorbentwithanidealporestructurefor

methaneadsorptionusingmetalcomplexes.ChemCommun.2001;16:1496–1497.

Figure8.Integralheatofadsorptionformethane

adsorptionprocess.

14.EddaoudiM,KimJ,RosiN,VodakD,WachterJ,O’KeeffeM,

YaghiOM.Systematicdesignofporesizeandfunctionalityinisore-ticularmofsandtheirapplicationinmethanestorage.Science.2002;295:469–472.15.Du†renT,SarkisovL,YaghiOM,SnurrRQ.Designofnewmateri-alsformethanestorage.Langmuir.2004;20:2683–2689.

16.MaS,SunD,SimmonsJM,CollierCD,YuanD,ZhouH-C.Metal-Organicframeworkfromananthracenederivativecontainingnano-scopiccagesexhibitinghighmethaneuptake.JAmChemSoc.2008;130:1012–1016.

17.AtwoodJL,BarbourLJ,JergaA.StorageofmethaneandFreonby

interstitialvanderWaalsconfinement.Science.2003;296:2367–2369.

18.HarlickPJE,TezelFH.Adsorptionofcarbondioxide,methaneand

nitrogen:PureandbinarymixtureadsorptionbyZSM-5withSiO2/Al2O3ratioof30.SepSciTechnol.2002;37:33–60.

19.LiP,TezelFH.AdsorptionSeparationofN2,O2,CO2,andCH4Gasesbyb-Zeolite.MicroporousMesoporousMater.2007;98:94–101.

20.LiP,TezelFH.Pureandbinaryadsorptionequilibriaofcarbon

dioxideandnitrogenonsilicalite.JChemEngData.2008;53:2479–2487.

21.GummaS,TaluO.Gibbsdividingsurfaceandheliumadsorption.

Adsorption.2003;9:17–28.

22.SalehiE,TaghikhaniV,GhotbiC,LayEN,ShojaeiA.Theoretical

andexperimentalstudyontheadsorptionanddesorptionofmethane

23.24.25.

26.27.28.29.30.31.

bygranularactivatedcarbonat25󰀁C.JNaturalGasChem.2007;16:415–422.

TanZ,GubbinsKE.Adsorptionincarbonmicroporesatsupercriti-caltemperatures.JPhyChem.1990;94:6061–6069.

MatrangaKR,MyersAL,GlandtED.Storageofnaturalgasbyadsorptiononactivatedcarbon.ChemEngSci.1992;47:1569–1579.ChenXS,McEnaneyB,MaysTJ,Alcaniz-MongeJ,Cazorla-AmorosD,Linares-SolanoA.Theoreticalandexperimentalstudiesofmethaneadsorptiononmicroporouscarbons.Carbon.1997;35:1251–1258.

DoDD.AdsorptionAnalysis:EquilibriaandKinetics,SeriesinChemicalEngineering.London,UK:ImperialCollegePress;1998.TaluO,MyersAL.Rigorousthermodynamictreatmentofgasadsorption.AIChEJ.1988;34:1887–1893.

HacskayloJJ,LeVanDM.CorrelationofadsorptionequilibriumdatausingamodifiedAntoineequation:anewapproachforpore-fillingmodels.Langmuir.1985;1:97–100.

SundaramN,YangRT.Isostericheatsofadsorptionfromgasmix-tures.JColloidsInterfaceSci.1998;198:378–388.

EstevesIAAC,Lopes,MSS,NunesPMC,MotaJPB.Adsorptionofnaturalgasandbiogascomponentsonactivatedcarbon.SepPurifTechnol.2008;62:281–296.

MyersAL.Thermodynamicsofadsorptioninporousmaterials.AIChEJ.2002;48:145–160.

ManuscriptreceivedOct.24,2010,andrevisionreceivedFeb.21,2011.

788DOI10.1002/aicPublishedonbehalfoftheAIChEMarch2012Vol.58,No.3AIChEJournal