A new approach to the modeling of deactivation in the conversion of methanol on zeolite catalysts.docx
《A new approach to the modeling of deactivation in the conversion of methanol on zeolite catalysts.docx》由会员分享,可在线阅读,更多相关《A new approach to the modeling of deactivation in the conversion of methanol on zeolite catalysts.docx(25页珍藏版)》请在冰豆网上搜索。
![A new approach to the modeling of deactivation in the conversion of methanol on zeolite catalysts.docx](https://file1.bdocx.com/fileroot1/2023-2/28/79acb1bd-227f-420d-9352-334253ef6ec1/79acb1bd-227f-420d-9352-334253ef6ec11.gif)
Anewapproachtothemodelingofdeactivationintheconversionofmethanolonzeolitecatalysts
Copyright©2009ElsevierLtdAllrightsreserved.
Anewapproachtothemodelingofdeactivationintheconversionofmethanolonzeolitecatalysts
TonV.W.Janssens
a,
aHaldorTopsøeA/S,Nymøllevej55,DK-2800Lyngby,Denmark
Received8January2009;
revised18February2009;
accepted8March2009.
Availableonline1May2009.
Abstract
Thedeactivationofazeolitecatalystintheconversionofmethanoltohydrocarbonsisdescribedasareductionoftheeffectiveamountofcatalystwithtimeonstream.Withtheassumptionsthattheconversionofmethanolisafirst-orderreaction,andthatthelossofactivecatalystisproportionaltotheconversion,anexpressionfortheconversionwithtimeonstreamisobtained,whichdescribestheexperimentaldatawell.Thisexpressioncontainstherateconstant,thatcharacterizestheactivity,andadeactivationcoefficientthatdescribesthedeactivationbehaviorasparameters.Itisshownthatactivecatalystsshowamoresuddendecreaseinconversion,andthatthedeactivationratedeterminesthetimeatwhichthedecreaseinconversionisobserved.Iftheinitialconversioniscloseto100%,thelifetimeto50%conversiondoesnotdependontheactivity,andthedeactivationcoefficientisdirectlyderivedfromtheexperimentaldata,bydividingthemeasuredlifetimeto50%conversionbytheappliedcontacttime.Thelifetimetoallotherconversionlevelsisdependentonbothdeactivationandactivity,whichimpliesthatacatalystlifetimetobreakthroughofmethanoldoesnotscalewiththedeactivationrate.Likewise,itisshownthattheconversioncapacityisagoodcharacterizationofthedeactivation,andthiscanbereadilycalculatedastheproductofthespacevelocityofmethanol(WHSV)andthelifetimeto50%conversion.Theamountofconvertedmethanolatotherconversionlevelsdependsonthedeactivation,theactivity,andappliedcontacttime(spacevelocity),andisthereforelessappropriatetouseasacharacterizationofthedeactivationbehavior.
Graphicalabstract
Thelifetimeto50%conversionisentirelydeterminedbythedeactivationproperties;differencesintheslopeofthedecayreflectdifferentcatalyticactivity.
Full-sizeimage(17K)
High-qualityimage(120K)
Keywords:
Conversionofmethanoltohydrocarbons;Zeolite;Methanoltogasoline;ZSM-5;Deactivation;Kineticmodel;Catalystlifetime;Conversioncapacity
ArticleOutline
1.Introduction
2.Experimental
3.Thedeactivationmodel
4.Discussion
4.1.Catalystlifetime
4.2.Methanolconversioncapacity
4.3.Finalremarks
5.Conclusions
References
1.Introduction
Themethanoltogasoline(MTG)reactionistheconversionofmethanoloverazeolite-basedcatalysttolightolefins(C2–C4)andliquidproductsintheboilingpointrangeofgasoline,whichtypicallyoccursinthetemperaturerangeof300–400 °C.IntheTIGAS(TopsøeIntegratedGASoline)process,theMTGreactioniscombinedwiththemethanolsynthesis,andinthiswayanefficientprocessconvertingsynthesisgastogasolineisobtained[1]and[2].
TheMTGreactioncanberegardedasasequentialreaction,consistingofthefollowingsteps[3]:
Methanol
DME→lightolefins→gasolineproducts
andthisisknowntooccuronavarietyofzeolites[4].TheZSM-5zeolite,however,isthemostcommonlyappliedone,sinceitappearstobesuperiorforthisreaction.TheMTGreactionoveraZSM-5-basedcatalystis,likemanyotherhydrocarbonreactionsoverzeolitecatalysts,alwaysaccompaniedbycokeformation,whichleadstodeactivationofthecatalyst.Thedeactivationbycokeformationoccursthroughblockingtheaccesstotheactiveacidsites,eitherbydepositionofcarbonaceouscompoundsdirectlyontheacidsitesitselfandinthemicroporechannelsofthezeolite(internalcoke),orbyblockingtheentrancetothemicropores,therebypreventingthediffusionofmethanolmoleculesintothezeolitestructure(externalcoke)[5],[6],[7],[8],[9],[10]and[11].Lossofactivityduetothecokeformationis,inprinciple,reversible,andthecatalyticactivitycanberestoredbyaregenerationthatremovesthedepositedcoke.Acommonprocedureistoburnoffthecokewithoxygenat500–600 °C[7],[8],[12]and[13].Regeneratedcatalystsoftenshowasomewhatloweractivitythanfreshones,possiblyduetodealuminationofthezeolite,whichtypicallyoccursatelevatedtemperatures(>500 °C)inthepresenceofwater.
Asaconsequenceofthedeactivation,thetotalamountofmethanolthatcanbeconvertedoveraZSM-5catalystislimited,anditisthereforeimportanttoknowthedeactivationbehaviorofthecatalyst.Tocharacterizecatalystdeactivation,itisrequiredtodescribehowthecatalyticactivitydecreasesduringthecatalystlifetime.Differentapproachesforsuchadescriptionhavebeendeveloped.Theapproachoftenencounteredintheliteratureisbasedonadescriptionofcarbonformationasafunctionofcontacttime,e.g.bytheempiricalVoorhiesequation[11],[14]and[15].Thecatalystactivityisthenrelatedtothecarboncontentinthecatalyst,using(semi-)empiricalrelations,resultinginthedescriptionoftheactivityasafunctionoftime[15],[16],[17],[18]and[19].Thepercolationmodeldescribesthecokeformationinthezeolitechannelsandtheaccessibilityoftheactivesitesinmoredetail,fromwhichthentheactivityisderived[20],[21],[22]and[23].Clearly,intheseapproaches,catalystdeactivationisdirectlyrelatedtotheformationofcoke,anddeactivationbydealuminationisnotincludedassuch.Theadvantageofthesemodelsisthatthecatalystdeactivationisrelatedtothechemistryofcarbonformationinthezeolitechannels,whichishelpfulinthedesignofzeolitestructureswithastableactivityforMTG.
Anotherapproachistodescribethechangeincatalyticactivitywithtime,andincorporatethisinakineticmodel.Thisyieldsaphenomenologicaldescription,andisindependentofthecauseofdeactivation.Usually,theactivationisdefinedbytheratiooftherateexpressionafteragiventimeonstreamandthatforthefreshcatalyst,whichresultsessentiallyinatime-dependentrateconstant[24].Thedeactivationoftendependsonparameters,suchasconversion,currentactivity,orcertainproductconcentrations,andistypicallydescribedbysimilarexpressionsasareactionrate,basedonrateconstantandanactivationenergyfordeactivation.Thedeactivationisthencharacterizedbyoneormorerateconstantsandactivationenergies[25],[26],[27],[28]and[29].
Inthisarticle,itwillbeshownthatcatalystdeactivationinMTGalsocanbedescribedasalossoftheeffectiveamountofactivecatalystinthereactorduringcatalystoperation,insteadofadecreaseintherateconstant,alsoresultinginaphenomenologicaldescriptionofthedeactivation.Therateatwhichtheeffectiveamountofactivecatalystdecreasesisthenadirectmeasureforthecatalystdeactivation.Thisapproachissimilartotheconceptofaneffectivespacevelocity,whichwasproposedpreviouslybyDahlandKolboe[30]and[31]andSapre[32],asareducedamountofactivecatalystresultsinaneffectivelylowercontacttimeorhigherspacevelocity.
Tomeasurethecatalystdeactivationexperimentally,theconversionwithtimeonstreamisfollowedunderconstantprocessconditions.Thedeactivationrateisthenderivedfromamodelthatdescribestheobserveddecreaseinconversionintermsofalossofactivecatalyst.Inthisarticle,experimentaldataforafewselectedZSM-5zeolites,showingdifferentdeactivationbehavior,arepresented.Thedeactivationratesdeterminedinthiswaycan,inprinciple,beusedforfurtherstudyoftheinfluenceofcatalystproperties,e.g.acidity,particlesize,orofprocessconditions,e.g.temperature,pressure,feedcomposition,onthecatalystdeactivation.Thescopeofthepresentarticle,however,isconfinedtothederivationofadeactivationmodel,andtheproceduresthatcanbeusedtoextractinformationonthedeactivationbehaviorfromtheexperimentaldata,withtheaimtodevelopamethodforthecharacterizationofcatalystdeactivation.
Themodelthatisderivedhereisbasedontheassumptionsthattheconversionofmethanolisafirst-orderreaction,andthatthedeactivationrateisproportionaltothetotalconversionofmethanolandDME.Therateconstantandthedeactivationrateappearastheparameterstobedetermined.Sincethedeactivationisregardedasareductionoftheamountofactivecatalyst,orofthenumberactivesites,ratherthanareductionoftheintrinsicactivityofanactivesite,therateconstantdoesnotchangewithtimeinthemodel.Althoughthisseemstobeaseveresimplification,itturnsoutthatitdescribestheobservedconversionofmethanolwithtimeonstreamwell,and,inaddition,asurprisinglyeasymethodtoextractthedeactivationratefromtheexperimentaldataisfound.Theexperimentallydetermineddeactivationratesandrateconstantsareusedtocalculatethecatalystlifetimeandtheconversioncapacity.Themodelgivesaclearinsightonhowthecatalyticactivity,thespacevelocity,andthedeactivationbehavioraffectthemeasuredconversions,thecatalystlifetime,andtheconversioncapacityofthecatalyst.
2.Experimental
Tomeasurethedeactivation,thetotalconversionofmethanoltohydrocarbons,i.e.thehydrocarbonyieldcalculatedonaC-atombasis,ismeasuredasafunctionoftime.ThoughitdoesnotincludetheconversionofmethanoltoDME,thisisreferredtoas“conversion”throughoutthispaper;methanolandDMEareassumedtobeinequilibriumandtreat