One-step aldol condensation reaction of dimethoxymethane and methyl acetate over supported Cs/ZSM-35 zeolite catalysts

Article abstract of DOI:10.1016/S1872-2067(18)63069-4

This study was performed for the development of a green and promising approach for the synthesis of methyl acrylate and acrylic acid by a one-step aldol condensation reaction of dimethoxymethane and methyl acetate over cesium oxide-supported on ZSM-35 zeo

Full text of DOI:10.1016/S1872-2067(18)63069-4

ChineseꢀJournalꢀofꢀCatalysisꢀ39ꢀ(2018)ꢀ1129–1137
ꢀ
ꢀ ꢀ ꢀ
ꢀ ꢀ ꢀ ꢀ
a v a i l a b l e a t w w w . s c i e n c e d i r e c t . c o m
ꢀ
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / c h n j c
ꢀ
ꢀ
Article
One‐stepꢀaldolꢀcondensationꢀreactionꢀofꢀdimethoxymethaneꢀandꢀ
methylꢀacetateꢀoverꢀsupportedꢀCs/ZSM‐35ꢀzeoliteꢀcatalystsꢀ
ZhanlingꢀMaꢀa,b,c,ꢀXiangangꢀMaꢀ ,ꢀHongchaoꢀLiuꢀ ,ꢀWenliangꢀZhuꢀ ,ꢀXinwenꢀGuoꢀ ,ꢀZhongminLiuꢀ *
b
b
b,#
a
ꢀ
b,
ꢀ
aꢀStateꢀKeyꢀLaboratoryꢀofꢀFineꢀChemicals,ꢀPSU‐DUTꢀJointꢀCenterꢀforꢀEnergyꢀResearch,ꢀSchoolꢀofꢀChemicalꢀEngineering,ꢀDalianꢀUniversityꢀofꢀTechnology,ꢀ
Dalianꢀ116024,ꢀLiaoning,ꢀChinaꢀ
bꢀNationalꢀEngineeringꢀLaboratoryꢀforꢀMethanolꢀtoꢀOlefins,ꢀDalianꢀInstituteꢀofꢀChemicalꢀPhysics,ꢀChineseꢀAcademyꢀofꢀSciences,ꢀDalianꢀ116023,ꢀLiaoning,ꢀ
Chinaꢀ
cꢀUniversityꢀofꢀChineseꢀAcademyꢀofꢀSciences,ꢀBeijingꢀ100049,ꢀChinaꢀ
A R T I C L E ꢀ I N F O ꢀ
A B S T R A C T ꢀ
Articleꢀhistory:ꢀ
Thisꢀstudyꢀwasꢀperformedꢀforꢀtheꢀdevelopmentꢀofꢀaꢀgreenꢀandꢀpromisingꢀapproachꢀforꢀtheꢀsynthesisꢀ
ofꢀmethylꢀacrylateꢀandꢀacrylicꢀacidꢀbyꢀaꢀone‐stepꢀaldolꢀcondensationꢀreactionꢀofꢀdimethoxymethaneꢀ
andꢀmethylꢀacetateꢀoverꢀcesiumꢀoxide‐supportedꢀonꢀZSM‐35ꢀzeoliteꢀcatalysts;ꢀtheꢀeffectꢀofꢀbaseꢀsitesꢀ
asꢀwellꢀasꢀacidꢀsitesꢀonꢀtheꢀaldolꢀcondensationꢀreactionꢀwasꢀstudiedꢀinꢀdetail.ꢀItꢀwasꢀfoundꢀthatꢀbaseꢀ
sitesꢀwereꢀharmfulꢀforꢀaldolꢀcondensationꢀdueꢀtoꢀtheirꢀfailureꢀinꢀcatalyzingꢀtheꢀdecompositionꢀofꢀ
dimethoxymethaneꢀprecursorꢀintoꢀformaldehyde,ꢀwhereasꢀtheꢀacidꢀsiteꢀwasꢀindispensableꢀforꢀtheꢀ
reactionꢀ toꢀproceed.ꢀ Thisꢀreactionꢀ cannotꢀtakeꢀ placeꢀ withoutꢀanꢀ acidꢀ site.ꢀ Althoughꢀ acidꢀ sitesꢀ inꢀ
H‐formꢀofꢀtheꢀzeoliteꢀ(HZSM‐35)ꢀareꢀindispensableꢀforꢀtheꢀaldolꢀcondensationꢀreaction,ꢀnotꢀallꢀofꢀ
themꢀtendꢀtoꢀfavorꢀthisꢀreaction.ꢀAꢀstrongꢀacidꢀcatalyzesꢀmethanol‐to‐olefin‐likeꢀreactionsꢀresultingꢀ
inꢀhydrocarbonꢀbyproducts,ꢀwhichꢀareꢀfinallyꢀtransferredꢀtoꢀhardꢀcoke.ꢀMediumꢀstrongꢀacidsꢀandꢀ
weakꢀacidsꢀareꢀgreatꢀcandidatesꢀforꢀtheꢀtargetꢀaldolꢀcondensationꢀreactionꢀwithꢀhighꢀactivityꢀandꢀ
Receivedꢀ26ꢀFebruaryꢀ2018ꢀ
Acceptedꢀ2ꢀAprilꢀ2018ꢀ
Publishedꢀ5ꢀJuneꢀ2018ꢀ
Keywords:ꢀ
Aldolꢀcondensationꢀ
Dimethoxymethaneꢀ
Methylꢀacetateꢀ
Zeoliteꢀ
Acidityꢀ
selectivity.ꢀ Aꢀ γ‐Al
2
O ꢀ sampleꢀ withꢀ abundantꢀ weak‐strengthꢀ Lewisꢀ acidꢀ sites,ꢀ togetherꢀ withꢀ aꢀ fewꢀ
3
medium‐strong‐strengthꢀ acidꢀ sites,ꢀ performsꢀ wellꢀ withꢀ aꢀ highꢀ activityꢀ andꢀ considerableꢀ stabilityꢀ
duringꢀtheꢀsynthesisꢀofꢀmethylꢀacrylateꢀandꢀacrylicꢀacid.ꢀ
©ꢀ2018,ꢀDalianꢀInstituteꢀofꢀChemicalꢀPhysics,ꢀChineseꢀAcademyꢀofꢀSciences.
PublishedꢀbyꢀElsevierꢀB.V.ꢀAllꢀrightsꢀreserved.
1
.ꢀ ꢀ Introductionꢀ
andꢀfurtherꢀposesꢀaꢀriskꢀofꢀexplosionꢀdueꢀtoꢀtheꢀdirectꢀmixingꢀofꢀ
hydrocarbonsꢀ withꢀ oxygen/air,ꢀ whichꢀ limitsꢀ theꢀ widespreadꢀ
applicationꢀ ofꢀ thisꢀ route.ꢀ Therefore,ꢀ itꢀ isꢀ highlyꢀ necessaryꢀ toꢀ
developꢀaꢀnovelꢀgreenꢀrouteꢀforꢀMAꢀandꢀAAꢀsynthesis.ꢀ
Theꢀ synthesisꢀ ofꢀ MAꢀ andꢀ AAꢀ throughꢀ theꢀ one‐stepꢀ aldolꢀ
condensationꢀ reactionꢀ ofꢀ formaldehydeꢀ withꢀ methylꢀ acetateꢀ
(MAc)ꢀhasꢀattractedꢀsubstantialꢀinterestꢀandꢀconcernꢀfromꢀbothꢀ
academicꢀandꢀindustrialꢀcommunitiesꢀinꢀtheꢀpastꢀfewꢀdecadesꢀ
becauseꢀ ofꢀ itsꢀ simplifiedꢀ reactionꢀ routeꢀ andꢀ commonꢀ feed‐
stocks,ꢀwhichꢀcanꢀbeꢀreadilyꢀderivedꢀfromꢀnaturalꢀgas,ꢀcoal,ꢀandꢀ
Acrylicꢀacidꢀ(AA)ꢀandꢀmethylꢀacrylateꢀ(MA),ꢀwidelyꢀusedꢀforꢀ
paintings,ꢀcoatings,ꢀcarbonꢀfibers,ꢀandꢀadhesives,ꢀareꢀpresentlyꢀ
producedꢀbyꢀtheꢀtwo‐stepꢀoxidationꢀofꢀpropyleneꢀ[1–5].ꢀInꢀthisꢀ
process,ꢀ propyleneꢀ isꢀ oxidizedꢀ withꢀ oxygen/airꢀ toꢀ produceꢀ
acroleinꢀfollowedꢀbyꢀoxidationꢀofꢀacroleinꢀtoꢀAA.ꢀAfterꢀtheꢀes‐
terificationꢀ ofꢀ AAꢀ withꢀ methanol,ꢀ AAꢀ isꢀ convertedꢀ intoꢀ MA.ꢀ
However,ꢀ thisꢀ two‐stepꢀ processꢀ isꢀ accompaniedꢀ byꢀ theꢀ
over‐oxidationꢀofꢀpropylene,ꢀresultingꢀinꢀaꢀlowꢀselectivityꢀofꢀAA,ꢀ
*
ꢀ
ꢀCorrespondingꢀauthor.ꢀTel:ꢀ+86‐411‐84379998;ꢀFax:ꢀ+86‐411‐84379038;ꢀE‐mail:ꢀliuzm@dicp.ac.cnꢀ
Correspondingꢀauthor.ꢀTel:ꢀ+86‐411‐84379418;ꢀFax:ꢀ+86‐411‐84379038;ꢀE‐mail:ꢀwlzhu@dicp.ac.cnꢀ
#
DOI:ꢀ10.1016/S1872‐2067(18)63069‐4ꢀ|ꢀhttp://www.sciencedirect.com/science/journal/18722067ꢀ|ꢀChin.ꢀJ.ꢀCatal.,ꢀVol.ꢀ39,ꢀNo.ꢀ6,ꢀJuneꢀ2018ꢀ ꢀ ꢀ

1130ꢀ
ZhanlingꢀMaꢀetꢀal.ꢀ/ꢀChineseꢀJournalꢀofꢀCatalysisꢀ39ꢀ(2018)ꢀ1129–1137ꢀ
biomassꢀ[6,7].ꢀThisꢀreactionꢀhasꢀbeenꢀreportedꢀtoꢀbeꢀcatalyzedꢀ
byꢀ acids,ꢀ bases,ꢀ orꢀ base‐acidꢀ bifunctionalꢀ catalysts,ꢀ amongꢀ
whichꢀ acid‐baseꢀ bifunctionalꢀ catalystsꢀ attractꢀ mostꢀ attentionꢀ
owingꢀ toꢀ theirꢀ goodꢀ performanceꢀ inꢀ aldolꢀ condensationꢀ reac‐
tionsꢀ asꢀ aꢀ resultꢀ ofꢀ cooperationꢀ betweenꢀ theirꢀ acidꢀ andꢀ baseꢀ
activeꢀsites.ꢀForꢀexample,ꢀitꢀwasꢀreportedꢀthatꢀtheꢀyieldꢀofꢀme‐
thylꢀ methacrylateꢀ overꢀ supportedꢀ cesiumꢀ oxideꢀ catalystsꢀ wasꢀ
notꢀmoreꢀthanꢀ23.0%ꢀduringꢀtheꢀaldolꢀcondensationꢀreactionꢀofꢀ
methylꢀ propionateꢀ andꢀ formaldehydeꢀ [8,9],ꢀ butꢀ theꢀ yieldꢀ wasꢀ
doubledꢀ overꢀ cesiumꢀ catalystsꢀ dopedꢀ withꢀ otherꢀ metalsꢀ withꢀ
withꢀtheꢀperformanceꢀofꢀtheꢀaldolꢀcondensationꢀreactionꢀwasꢀ
illustrated.ꢀ
2.ꢀ ꢀ Experimentalꢀ
2.1.ꢀ ꢀ Catalystꢀpreparationꢀ
Zeolitesꢀ (ZSM‐35)ꢀ andꢀ γ–Al
2
O ꢀ wereꢀ purchasedꢀ fromꢀ theꢀ
3
ShanghaiꢀZhuoyueꢀChemicalꢀLimitedꢀCompany,ꢀChina.ꢀTheꢀzeo‐
liteꢀsampleꢀwasꢀconvertedꢀintoꢀaꢀH‐formꢀofꢀHZSM‐35ꢀbeforeꢀuseꢀ
byꢀconductingꢀionꢀexchangeꢀthriceꢀatꢀ80ꢀ°Cꢀforꢀ6ꢀhꢀinꢀ1ꢀmol/Lꢀ
ꢀ
acid‐baseꢀ bifunctionalꢀ properties [10].ꢀ Inꢀ addition,ꢀ V–Pꢀ oxideꢀ
catalystsꢀattractꢀmuchꢀacademicꢀattentionꢀowingꢀtoꢀtheirꢀsimi‐
larꢀacid‐baseꢀbifunctionalꢀproperties.ꢀSpiveyꢀetꢀal.ꢀ[11]ꢀevaluat‐
edꢀatꢀleastꢀ80ꢀcatalyticꢀmaterialsꢀandꢀfoundꢀVꢀoxidesꢀsupportedꢀ
onꢀamorphousꢀsilicaꢀtoꢀbeꢀtheꢀmostꢀeffective.ꢀAiꢀ[12–17]ꢀcon‐
ductedꢀtheꢀaldolꢀcondensationꢀreactionꢀofꢀpropionicꢀacidꢀwithꢀ
NH
4
NO ,ꢀfollowedꢀbyꢀwashingꢀwithꢀdeionizedꢀwater,ꢀdryingꢀatꢀ
3
100ꢀ°Cꢀovernight,ꢀandꢀcalcinationꢀatꢀ550ꢀ°Cꢀforꢀ4ꢀh.ꢀ
Toꢀ synthesizeꢀ theꢀ zeolite‐supportedꢀ alkalineꢀ metalꢀ cesiumꢀ
oxideꢀ catalysts,ꢀ theꢀ HZSM‐35ꢀ zeoliteꢀ precursorꢀ wasꢀ impreg‐
natedꢀwithꢀanꢀaqueousꢀsolutionꢀcontainingꢀvaryingꢀamountsꢀofꢀ
formaldehydeꢀ overꢀ V–P–Siꢀ samplesꢀ andꢀ obtainedꢀ aꢀ yieldꢀ ofꢀ
Cs
2
CO .ꢀ Theꢀ mixtureꢀ wasꢀ heatedꢀ withꢀ stirringꢀ atꢀ 50ꢀ °Cꢀ forꢀ atꢀ
3
ꢀ
70%ꢀatꢀtheꢀoptimumꢀexperimentalꢀconditions.ꢀGogateꢀetꢀal.[18]ꢀ
leastꢀ5ꢀhꢀuntilꢀaꢀpaste‐likeꢀmaterialꢀisꢀformedꢀandꢀthenꢀdriedꢀinꢀ
anꢀoven.ꢀTheꢀdriedꢀsampleꢀwasꢀcalcinedꢀinꢀflowingꢀairꢀatꢀ500ꢀ°Cꢀ
forꢀ4ꢀh.ꢀTheꢀsampleꢀwasꢀdenotedꢀasꢀZ–Cs–X,ꢀwhereꢀZꢀdenotesꢀ
zeoliteꢀ andꢀ Xꢀ refersꢀ toꢀ theꢀ weightꢀ percentageꢀ ofꢀ theꢀ alkalineꢀ
metal.ꢀBeforeꢀuse,ꢀtheꢀsampleꢀpowderꢀwasꢀpressedꢀandꢀsievedꢀ
toꢀ obtainꢀ particlesꢀ inꢀ theꢀ sizeꢀ rangeꢀ ofꢀ 20–40ꢀ mesh.ꢀ Weꢀ alsoꢀ
preparedꢀsilica‐supportedꢀalkalineꢀmetalꢀoxideꢀsamplesꢀusingꢀaꢀ
similarꢀmethod.ꢀ ꢀ
studiedꢀ theꢀ vaporꢀ phaseꢀ aldolꢀ condensationꢀ reactionꢀ ofꢀ for‐
maldehydeꢀwithꢀpropionateꢀderivativesꢀoverꢀaꢀseriesꢀofꢀV–P–Siꢀ
catalystsꢀ ofꢀ varyingꢀ atomicꢀ ratiosꢀ andꢀ obtainedꢀ aꢀ maximumꢀ
methacrylicꢀ acidꢀ yieldꢀ ofꢀ 56%.ꢀ V–P–Siꢀ catalystsꢀ alsoꢀ performꢀ
wellꢀinꢀtheꢀaldolꢀcondensationꢀ reactionꢀofꢀ formaldehydeꢀ withꢀ
otherꢀ carbonylꢀ compounds,ꢀ suchꢀ asꢀ acetoneꢀ orꢀ acetaldehydeꢀ
[19–21].ꢀ
Theꢀeffectꢀofꢀtheꢀacid‐baseꢀbifunctionalꢀpropertyꢀofꢀV–P–Siꢀ
ternaryꢀ oxideꢀ catalystsꢀ wasꢀ studiedꢀ inꢀ theꢀ caseꢀ ofꢀ theꢀ aldolꢀ
condensationꢀ ofꢀ aceticꢀ acidꢀ orꢀ methylꢀ acetateꢀ withꢀ formalde‐
hydeꢀ toꢀ prepareꢀ MAꢀ andꢀ AA.ꢀ Huꢀ etꢀ al.ꢀ [22]ꢀ preparedꢀ V–P–Siꢀ
catalystsꢀ withꢀ highꢀ acidꢀ andꢀ alkaliꢀ quantitiesꢀ byꢀ theꢀ incipientꢀ
wetnessꢀimpregnationꢀmethodꢀandꢀfoundꢀthatꢀtheꢀactivitiesꢀofꢀ
theseꢀ catalystsꢀ forꢀ theꢀ conversionꢀ ofꢀ formaldehydeꢀ increasedꢀ
withꢀanꢀincreaseꢀinꢀtheꢀalkaliꢀquantity,ꢀwhileꢀtheꢀselectivityꢀforꢀ
acrylicꢀacidꢀincreasedꢀwithꢀanꢀincreaseꢀinꢀtheꢀacidꢀquantity.ꢀThisꢀ
meansꢀ thatꢀ theꢀ acid‐baseꢀ propertiesꢀ ofꢀ theseꢀ catalystsꢀ playꢀ aꢀ
significantꢀ roleꢀ inꢀ enhancingꢀ bothꢀ precursorꢀ conversionꢀ andꢀ
productꢀselectivity.ꢀYangꢀetꢀal.ꢀ[23]ꢀfoundꢀthatꢀweakꢀbaseꢀsitesꢀ
wereꢀ essentialꢀ forꢀ aldolꢀ condensation,ꢀ whileꢀ weakꢀ acidꢀ sitesꢀ
wereꢀresponsibleꢀforꢀimprovementꢀinꢀtheꢀselectivityꢀofꢀacrylicꢀ
acidꢀoverꢀaꢀseriesꢀofꢀV–P–Siꢀcatalysts.ꢀFengꢀetꢀal.ꢀ[24]ꢀstudiedꢀ
theꢀcorrelationꢀbetweenꢀtheꢀcatalystꢀstructureꢀandꢀconstitutionꢀ
ofꢀaꢀV–Pꢀcatalystꢀandꢀitsꢀperformanceꢀinꢀtheꢀaldolꢀcondensationꢀ
reactionꢀofꢀaceticꢀacidꢀorꢀmethylꢀacetateꢀandꢀformaldehydeꢀandꢀ
foundꢀ thatꢀ catalystsꢀ containingꢀ medium‐strongꢀ acidꢀ sitesꢀ ofꢀ
highꢀdensityꢀperformedꢀmuchꢀbetter.ꢀInꢀourꢀpreviousꢀworkꢀonꢀ
theꢀ aldolꢀ condensationꢀ reactionꢀ ofꢀ formaldehydeꢀ withꢀ methylꢀ
acetateꢀoverꢀ aꢀ pureꢀ acidicꢀ H‐formꢀofꢀaꢀzeoliteꢀtoꢀ produceꢀ MAꢀ
andꢀAA,ꢀweꢀreportedꢀanꢀoptimumꢀyieldꢀasꢀhighꢀasꢀ61.1%ꢀ[25].ꢀItꢀ
seemsꢀthatꢀbaseꢀsitesꢀasꢀwellꢀasꢀacidꢀsitesꢀareꢀgreatlyꢀhelpfulꢀinꢀ
catalyzingꢀaldolꢀcondensationꢀreactions.ꢀ ꢀ
2.2.ꢀ ꢀ Characterizationꢀ
Theꢀ crystallinitiesꢀ ofꢀ theꢀ samplesꢀ wereꢀ characterizedꢀ byꢀ aꢀ
PANalyticalꢀX’PertꢀPROꢀX‐rayꢀdiffractionꢀ(XRD)ꢀinstrumentꢀwithꢀ
CuꢀK ꢀradiationꢀ(λꢀ=ꢀ1.51059ꢀÅ)ꢀatꢀ40ꢀkVꢀandꢀ40ꢀmA.ꢀ ꢀ
α
Theꢀ totalꢀ surfaceꢀ areaꢀ wasꢀ calculatedꢀ accordingꢀ toꢀ theꢀ
Brunauer‐Emmett‐Tellerꢀ (BET)ꢀ equation.ꢀ Theꢀ microporeꢀ vol‐
umesꢀ andꢀ microporeꢀ surfaceꢀ areasꢀ wereꢀ evaluatedꢀ usingꢀ theꢀ
t‐plotꢀ method.ꢀ Theꢀ mesoporeꢀ volumesꢀ wereꢀ calculatedꢀ usingꢀ
theꢀBarrett‐Joyner‐Halendaꢀ(BJH)ꢀmethod.ꢀ ꢀ
Theꢀbulkꢀacidityꢀofꢀtheꢀzeoliteꢀwasꢀdeterminedꢀbyꢀtheꢀtem‐
perature‐programmedꢀdesorptionꢀofꢀammoniaꢀ(NH ‐TPD)ꢀonꢀaꢀ
3
MicromeriticsꢀAutoChemꢀ2920ꢀinstrument.ꢀTheꢀsampleꢀ(0.2ꢀg)ꢀ
wasꢀloadedꢀinꢀaꢀU‐shapedꢀmicroreactorꢀandꢀpreheatedꢀatꢀ550ꢀ
°Cꢀforꢀ0.5ꢀhꢀinꢀaꢀheliumꢀatmosphere.ꢀAfterꢀcoolingꢀtoꢀ100ꢀ°C,ꢀtheꢀ
sampleꢀwasꢀsaturatedꢀwithꢀammonia,ꢀfollowedꢀbyꢀpurgingꢀwithꢀ
heliumꢀ toꢀ removeꢀ physicallyꢀ adsorbedꢀ ammoniaꢀ molecules.ꢀ
Ammoniaꢀ desorptionꢀ wasꢀ conductedꢀ inꢀ aꢀ flowingꢀ heliumꢀ at‐
mosphereꢀ(30ꢀmL/min)ꢀbyꢀheatingꢀfromꢀ100ꢀtoꢀ650ꢀ°Cꢀatꢀaꢀrateꢀ
ofꢀ 10ꢀ °C/min;ꢀ desorptionꢀ wasꢀ analyzedꢀ usingꢀ aꢀ thermalꢀ con‐
ductivityꢀdetectorꢀ(TCD).ꢀ ꢀ
Fourierꢀ transformꢀ infraredꢀ (FT‐IR)ꢀ spectroscopyꢀ wasꢀ con‐
–1
ductedꢀatꢀaꢀspectralꢀresolutionꢀofꢀ4ꢀcm ꢀonꢀaꢀBrukerꢀTensorꢀ27ꢀ
FT‐IRꢀ spectrophotometerꢀ equippedꢀ withꢀ aꢀ mercu‐
ry‐cadmium‐tellurideꢀ (MCT)ꢀ detector,ꢀ whichꢀ wasꢀ sensitiveꢀ toꢀ
–OHꢀ groupꢀ vibrations.ꢀ Theꢀ sampleꢀ wasꢀ pressedꢀ intoꢀ aꢀ
self‐supportingꢀdiskꢀwithꢀaꢀdiameterꢀofꢀ13ꢀmm.ꢀTheꢀdiskꢀwasꢀ
thenꢀputꢀinꢀaꢀquartzꢀcellꢀconnectedꢀtoꢀaꢀvacuumꢀsystem,ꢀsealedꢀ
Herein,ꢀweꢀpreparedꢀaꢀseriesꢀofꢀzeolitesꢀsupportedꢀbyꢀvary‐
ingꢀweightꢀpercentagesꢀofꢀcesiumꢀoxideꢀbyꢀanꢀincipientꢀwetnessꢀ
impregnationꢀmethodꢀinꢀorderꢀtoꢀstudyꢀtheꢀeffectꢀofꢀbaseꢀsitesꢀ
orꢀacidꢀsitesꢀonꢀtheꢀperformanceꢀofꢀtheꢀaldolꢀcondensationꢀre‐
action.ꢀ Dimethoxymethaneꢀ (DMM)ꢀ wasꢀ usedꢀ asꢀ theꢀ sourceꢀ ofꢀ
formaldehyde.ꢀ Theꢀ texturalꢀ andꢀ acid‐baseꢀ propertiesꢀ ofꢀ theseꢀ
samplesꢀ wereꢀ characterizedꢀ inꢀ detailꢀ andꢀ theirꢀ relationshipꢀ
withꢀCaF ꢀwindows,ꢀandꢀthenꢀheatedꢀupꢀtoꢀ450ꢀ°Cꢀforꢀatꢀleastꢀ4ꢀ
2
hꢀtoꢀremoveꢀanyꢀretainedꢀwaterꢀbeforeꢀcollectingꢀtheꢀspectra.ꢀ
Pyridineꢀ adsorptionꢀ wasꢀ analyzedꢀ byꢀ exposingꢀ theꢀ preheatedꢀ

ꢀ
ZhanlingꢀMaꢀetꢀal.ꢀ/ꢀChineseꢀJournalꢀofꢀCatalysisꢀ39ꢀ(2018)ꢀ1129–1137ꢀ
1131ꢀ
diskꢀtoꢀaꢀpyridineꢀvaporꢀatꢀroomꢀtemperature.ꢀAllꢀtheꢀspectraꢀ
wereꢀcollectedꢀatꢀtheꢀroomꢀtemperature.ꢀ ꢀ
Thermogravimetricꢀ analysisꢀ (TGA)ꢀ wasꢀ conductedꢀ onꢀ aꢀ
thermogravimetricꢀ analyzerꢀ (Q600ꢀ SDT/Omnistar).ꢀ Aboutꢀ 10ꢀ
1
60
20
Zeolite
ZCs1wt
ZCs5wt
ZCs10wt
1
mgꢀofꢀtheꢀdeactivatedꢀsampleꢀwasꢀplacedꢀinꢀanꢀAl
andꢀthenꢀheatedꢀfromꢀroomꢀtemperatureꢀtoꢀ800ꢀ°Cꢀatꢀaꢀrateꢀofꢀ
0ꢀ°C/minꢀwithꢀaꢀpurgingꢀairꢀflowꢀofꢀ100ꢀmL/min.ꢀ
2 3
O ꢀcrucibleꢀ
1
80
2.3.ꢀ ꢀ Catalyticꢀtestꢀ
TheꢀaldolꢀcondensationꢀreactionꢀofꢀDMMꢀandꢀMAcꢀwasꢀcar‐
40
riedꢀoutꢀatꢀ350ꢀ°Cꢀwithꢀ0.5ꢀgꢀofꢀtheꢀsampleꢀ(20–40ꢀmesh)ꢀonꢀaꢀ
fixed‐bedꢀreactor.ꢀTheꢀsystemꢀpressureꢀwasꢀsetꢀatꢀ3ꢀMPa.ꢀDMMꢀ
andꢀMAcꢀprecursorsꢀinꢀtwoꢀseparateꢀstainlessꢀsteelꢀtubesꢀheldꢀ
0.0
0.2
0.4
0.6
0.8
1.0
atꢀ20ꢀ°CꢀwereꢀbubbledꢀintoꢀtheꢀreactionꢀtubeꢀusingꢀanꢀN
2
ꢀflowꢀ
(30ꢀ mL/min).ꢀ Theꢀ obtainedꢀ productsꢀ wereꢀ analyzedꢀ byꢀ anꢀ
P/P0
onlineꢀgasꢀchromatographꢀ(GC)ꢀequippedꢀwithꢀaꢀflameꢀioniza‐
tionꢀdetectorꢀ(FID)ꢀconnectedꢀtoꢀaꢀFFAPꢀcapillaryꢀcolumn.ꢀTheꢀ
conversionꢀofꢀDMMꢀtoꢀMAcꢀwasꢀdefinedꢀinꢀtermsꢀofꢀtheꢀmolarꢀ
ratioꢀofꢀconsumedꢀprecursorꢀtoꢀtheꢀfeed.ꢀTheꢀyieldꢀwasꢀcalcu‐
latedꢀasꢀ(molesꢀofꢀMAꢀandꢀAA)/(molesꢀofꢀMAcꢀfed).ꢀ ꢀ
Fig.ꢀ 2.ꢀ Nitrogenꢀ adsorption‐desorptionꢀ isothermsꢀ ofꢀ ZSM‐35ꢀ zeolitesꢀ
withꢀdifferentꢀweightꢀpercentagesꢀofꢀcesium.ꢀ
microporousꢀmaterialꢀwithꢀaꢀweakꢀhystereticꢀloopꢀdueꢀtoꢀtheꢀ
presenceꢀ ofꢀ inter‐crystalꢀ mesopores.ꢀ Theꢀ BETꢀ surfaceꢀ areaꢀ
2
changesꢀslightlyꢀfromꢀ262ꢀtoꢀ258ꢀm /gꢀwhenꢀ1ꢀwt%ꢀcesiumꢀwasꢀ
3.ꢀ ꢀ Resultsꢀandꢀdiscussionꢀ
usedꢀasꢀtheꢀsupport.ꢀFurther,ꢀthereꢀoccursꢀaꢀsmallꢀdecreaseꢀinꢀ
3
theꢀ microporeꢀ volumeꢀ fromꢀ 0.105ꢀ toꢀ 0.100ꢀ cm /g,ꢀ indicatingꢀ
3.1.ꢀ ꢀ Characterizationꢀofꢀzeoliteꢀsamplesꢀwithꢀvaryingꢀweightꢀ
thatꢀtheꢀcesiumꢀoxideꢀsupportꢀisꢀdistributedꢀuniformlyꢀonꢀtheꢀ
zeoliteꢀsurfaceꢀwithoutꢀgreatlyꢀalteringꢀitsꢀtexturalꢀproperties.ꢀ
percentagesꢀofꢀcesiumꢀ ꢀ
2
TheꢀBETꢀsurfaceꢀareaꢀdecreasesꢀfromꢀ262ꢀtoꢀ231ꢀm /gꢀatꢀaꢀce‐
Asꢀ shownꢀ inꢀ Fig.ꢀ 1,ꢀ HZSM‐35ꢀ exhibitsꢀ aꢀ typicalꢀ crystallineꢀ
ferrieriteꢀ(FER)ꢀframeworkꢀwithꢀclearꢀdiffractionꢀpeaksꢀatꢀ9.2°,ꢀ
siumꢀweightꢀpercentageꢀofꢀ5ꢀwt%,ꢀwhileꢀtheꢀmicroporeꢀvolumeꢀ
3
decreasesꢀ fromꢀ 0.105ꢀ toꢀ 0.086ꢀ cm /gꢀ althoughꢀ theꢀ externalꢀ
25.2°,ꢀandꢀ25.7°.ꢀTheꢀmorphologyꢀofꢀtheꢀzeolite,ꢀasꢀcharacter‐
2
surfaceꢀ doesꢀ notꢀ decreaseꢀ muchꢀ (fromꢀ 47ꢀ toꢀ 45ꢀ m /g).ꢀ Mi‐
izedꢀbyꢀscanningꢀelectronꢀmicroscopyꢀ(SEM)ꢀ(Fig.ꢀS1),ꢀindicatesꢀ
typicalꢀflake‐packing.ꢀNoꢀnewꢀdiffractionꢀpeaksꢀappearꢀafterꢀtheꢀ
impregnationꢀofꢀcesiumꢀoxide,ꢀindicatingꢀthatꢀnoꢀfusedꢀcesiumꢀ
oxideꢀphaseꢀexistsꢀinꢀtheꢀpreparedꢀsamples.ꢀ ꢀ
Theꢀnitrogenꢀadsorption‐desorptionꢀisothermsꢀofꢀtheꢀsam‐
plesꢀareꢀshownꢀinꢀFig.ꢀ2ꢀandꢀtheirꢀtextualꢀpropertiesꢀareꢀlistedꢀ
inꢀ Tableꢀ 1.ꢀ HZSM‐35ꢀ zeoliteꢀ exhibitsꢀ anꢀ isothermꢀ typicalꢀ ofꢀ aꢀ
croporeꢀblockageꢀbecomesꢀmoreꢀsevereꢀwithꢀaꢀdecreaseꢀinꢀtheꢀ
microporeꢀ volumeꢀ fromꢀ 0.105ꢀ toꢀ 0.063ꢀ cm /gꢀ atꢀ aꢀ cesiumꢀ
3
weightꢀ percentageꢀofꢀ 10ꢀ wt%;ꢀthisꢀisꢀ accompaniedꢀ byꢀreduc‐
2
tionsꢀinꢀtheꢀBETꢀsurfaceꢀareaꢀ(fromꢀ262ꢀtoꢀ173ꢀm /g)ꢀandꢀtheꢀ
2
externalꢀsurfaceꢀareaꢀ(fromꢀ47ꢀtoꢀ38ꢀm /g).ꢀThisꢀsuggestsꢀthatꢀaꢀ
cesiumꢀweightꢀpercentageꢀgreaterꢀthanꢀ5ꢀwt%ꢀisꢀexcessiveꢀforꢀ
supportingꢀtheꢀzeoliteꢀandꢀwouldꢀcauseꢀaꢀsevereꢀblockageꢀofꢀtheꢀ
sample’sꢀtexturalꢀproperties.ꢀThisꢀisꢀespeciallyꢀtrueꢀinꢀtheꢀcaseꢀ
ofꢀmicropores.ꢀ ꢀ
ZCs10wt
ZCs5wt
ZCs1wt
Zeolite
Theꢀ resultsꢀ ofꢀ NH ‐TPDꢀ curvesꢀ ofꢀ samplesꢀ withꢀ differentꢀ
3
weightꢀpercentagesꢀofꢀcesiumꢀareꢀshownꢀinꢀFig.ꢀ3.ꢀInꢀtheꢀcaseꢀofꢀ
theꢀHZSM‐35ꢀzeolite,ꢀtwoꢀtypicalꢀdesorptionꢀpeaksꢀofꢀammoniaꢀ
areꢀobservedꢀatꢀ169ꢀandꢀ405ꢀ°C,ꢀsuggestingꢀthatꢀtwoꢀtypesꢀofꢀ
acidꢀ sites,ꢀ includingꢀ weakꢀ andꢀ strongꢀ ones,ꢀ areꢀ distributedꢀ inꢀ
theꢀ H‐formꢀ ofꢀ theꢀ zeolite.ꢀ However,ꢀ theꢀ numberꢀ ofꢀ acidꢀ sites,ꢀ
especiallyꢀtheꢀstrongꢀones,ꢀexperiencesꢀaꢀsharpꢀdecreaseꢀafterꢀ
impregnationꢀwithꢀ1ꢀwt%ꢀcesium,ꢀwhichꢀindicatesꢀthatꢀaꢀlargeꢀ
Tableꢀ1ꢀ
Texturalꢀpropertiesꢀofꢀsamplesꢀwithꢀdifferentꢀcesiumꢀweightꢀpercent‐
ages.ꢀ
A
Ext‐BETꢀ
V
microꢀ
V
mesoꢀ
V
totalꢀ
2
Sampleꢀ
A
BETꢀ(m /g)
2
3
3
3
(m /g)ꢀ
(cm /g)ꢀ (cm /g)
(cm /g)ꢀ
0.180ꢀ
0.177ꢀ
0.164ꢀ
0.135ꢀ
Zeoliteꢀ*ꢀ
Z‐Cs‐1wtꢀ
Z‐Cs‐5wtꢀ
Z‐Cs‐10wtꢀ
262ꢀ
258ꢀ
231ꢀ
173ꢀ
47ꢀ
50ꢀ
45ꢀ
0.105ꢀ
0.100ꢀ
0.086ꢀ
0.063ꢀ
0.075ꢀ
0.077ꢀ
0.078ꢀ
0.072ꢀ
10
20
30
40
50

2
 ( )
Fig.ꢀ1.ꢀXRDꢀpatternsꢀofꢀHZSM‐35ꢀzeoliteꢀsupportedꢀbyꢀdifferentꢀweight
38ꢀ
percentagesꢀofꢀcesium.ꢀ
*ꢀHZSM‐35ꢀzeolite.ꢀ

1132ꢀ
ZhanlingꢀMaꢀetꢀal.ꢀ/ꢀChineseꢀJournalꢀofꢀCatalysisꢀ39ꢀ(2018)ꢀ1129–1137ꢀ
notꢀ changeꢀ (stillꢀ 100%)ꢀ withꢀ 1ꢀ wt%ꢀ cesiumꢀ asꢀ theꢀ support.ꢀ
However,ꢀ itꢀ reducesꢀ toꢀ aboutꢀ 10%ꢀ withꢀ Z–Cs–5wtꢀ andꢀ evenꢀ
becomesꢀzeroꢀinꢀtheꢀcaseꢀofꢀZ–Cs–10wt.ꢀTheseꢀresultsꢀmakeꢀitꢀ
obviousꢀ thatꢀ cesiumꢀ oxideꢀ supportsꢀ areꢀ harmfulꢀ toꢀ DMMꢀ de‐
composition.ꢀThisꢀisꢀalsoꢀtrueꢀforꢀMAcꢀprecursors,ꢀwhoseꢀcon‐
versionꢀdropsꢀtoꢀ20%ꢀatꢀaꢀcesiumꢀweightꢀpercentageꢀofꢀmoreꢀ
thanꢀ5ꢀwt%.ꢀConsideringꢀtheꢀfactꢀthatꢀtheꢀalkalinityꢀofꢀtheꢀsam‐
plesꢀ isꢀ greatlyꢀ enhancedꢀ byꢀ theꢀ introductionꢀ ofꢀ cesiumꢀ oxide,ꢀ
weꢀ suggestꢀ thatꢀ alkalinityꢀ isꢀ notꢀ helpfulꢀ inꢀ enhancingꢀ eitherꢀ
DMMꢀorꢀMAcꢀconversion.ꢀ ꢀ
Zeolite
ZCs1wt
ZCs5wt
ZCs10wt
ConversionꢀofꢀeitherꢀDMMꢀorꢀMAcꢀonꢀtheꢀZSM‐35ꢀzeoliteꢀisꢀ
notꢀadverselyꢀaffectedꢀbyꢀtheꢀZ–Cs–1wtꢀsample.ꢀWeꢀknowꢀthatꢀaꢀ
largeꢀnumberꢀofꢀacidꢀsitesꢀinꢀtheꢀZ–Cs–1wtꢀsampleꢀareꢀcoveredꢀ
byꢀtheꢀcesiumꢀoxideꢀsupport,ꢀasꢀillustratedꢀinꢀFig.ꢀ3,ꢀresultingꢀinꢀ
anꢀobviousꢀdecreaseꢀinꢀtheꢀstrengthsꢀofꢀtheꢀammoniaꢀdesorp‐
tionꢀpeaksꢀatꢀbothꢀlowꢀandꢀhighꢀtemperatures.ꢀThisꢀmeansꢀthatꢀ
theseꢀacidicꢀsitesꢀcoveredꢀbyꢀcesiumꢀoxideꢀareꢀnotꢀnecessaryꢀforꢀ
theꢀconversionꢀofꢀDMMꢀandꢀMAc.ꢀTheꢀremainingꢀacidꢀsitesꢀinꢀ
Z–Cs–1wtꢀareꢀenoughꢀtoꢀtriggerꢀtheꢀconversionꢀofꢀtheꢀprecur‐
sor.ꢀHowever,ꢀifꢀtheꢀremainingꢀacidꢀsitesꢀofꢀtheꢀsampleꢀareꢀcov‐
ered,ꢀasꢀinꢀtheꢀcaseꢀofꢀ5ꢀwt%ꢀcesiumꢀorꢀevenꢀ10ꢀwt%ꢀcesium,ꢀ
DMMꢀ andꢀ MAcꢀ conversionꢀ fallsꢀ sharply,ꢀ asꢀ shownꢀ inꢀ Fig.ꢀ 4,ꢀ
whichꢀsuggestsꢀthatꢀacidityꢀisꢀindispensableꢀforꢀtheꢀconversionꢀ
ofꢀprecursorsꢀinꢀanꢀaldolꢀcondensationꢀreaction.ꢀButꢀnotꢀallꢀtheꢀ
acidsꢀinꢀtheꢀHZSM‐35ꢀzeoliteꢀtendꢀtoꢀfavorꢀaldolꢀcondensation.ꢀ
Forꢀinstance,ꢀanꢀobviousꢀinitialꢀperiodꢀofꢀaboutꢀ2ꢀhꢀisꢀobservedꢀ
inꢀtheꢀHZSM‐35ꢀzeolite,ꢀasꢀ shownꢀinꢀFig.ꢀ 4,ꢀ duringꢀ whichꢀtheꢀ
incipientꢀconversionꢀofꢀMAcꢀatꢀTOSꢀ=ꢀ0.5ꢀhꢀisꢀ78%ꢀandꢀthenꢀfallsꢀ
toꢀ50%ꢀrapidly.ꢀAꢀlargeꢀamountꢀofꢀcokeꢀisꢀquicklyꢀformedꢀinꢀthisꢀ
periodꢀ(firstꢀ2ꢀh),ꢀasꢀshownꢀinꢀFig.ꢀS2.ꢀThisꢀinitialꢀperiodꢀdisap‐
pearedꢀwhenꢀZSM‐35ꢀwasꢀsupportedꢀwithꢀvaryingꢀweightꢀper‐
centagesꢀ ofꢀ cesium.ꢀ Takingꢀ intoꢀ accountꢀ theꢀ factꢀ thatꢀ aꢀ largeꢀ
numberꢀofꢀstrongꢀacidꢀsitesꢀinꢀZSM‐35ꢀzeoliteꢀareꢀcoveredꢀbyꢀ
theꢀsupportingꢀcesiumꢀoxide,ꢀweꢀproposeꢀthatꢀstrongꢀacidsꢀinꢀ
theꢀH‐formꢀofꢀtheꢀzeoliteꢀcauseꢀdeactivation.ꢀ ꢀ
TheꢀeffectꢀofꢀaddedꢀcesiumꢀoxideꢀonꢀtheꢀyieldꢀofꢀMAꢀandꢀAAꢀ
isꢀshownꢀinꢀFig.ꢀ5.ꢀInꢀtheꢀinitialꢀperiodꢀ(firstꢀ2ꢀh),ꢀtheꢀyieldꢀofꢀMAꢀ
andꢀAAꢀwithꢀtheꢀHZSM‐35ꢀzeoliteꢀisꢀveryꢀlow,ꢀwhereasꢀaꢀlargeꢀ
quantityꢀofꢀhydrocarbonꢀbyproductsꢀareꢀproducedꢀdueꢀtoꢀtheꢀ
unavoidableꢀmethanol‐to‐olefin‐likeꢀprocessꢀatꢀsuchꢀhighꢀreac‐
tionꢀtemperatures.ꢀThisꢀperiodꢀdisappearsꢀimmediatelyꢀifꢀzeo‐
litesꢀ areꢀ supportedꢀ withꢀ cesiumꢀ oxide,ꢀ whichꢀ inꢀ turnꢀ provesꢀ
thatꢀanꢀexcessꢀofꢀstrongꢀacidsꢀinꢀtheꢀHZSM‐35ꢀzeoliteꢀcontrib‐
utesꢀ mainlyꢀ toꢀ by‐reactions.ꢀ Alkalinityꢀ prohibitsꢀ theꢀ
by‐reactionsꢀ causedꢀ byꢀ strongꢀ acidsꢀ inꢀ theꢀ HZSM‐35ꢀ zeolite.ꢀ
Similarly,ꢀ alkalinityꢀ limitsꢀ theꢀ productionꢀ ofꢀ MAꢀ andꢀ AA.ꢀ Forꢀ
instance,ꢀtheꢀyieldꢀofꢀMAꢀandꢀAAꢀwithꢀtheꢀZ–Cs–1wtꢀsampleꢀisꢀ
lowerꢀ thanꢀ thatꢀ ofꢀ theꢀ HZSM‐35ꢀ zeoliteꢀ inꢀ theꢀ sameꢀ reactionꢀ
time,ꢀasꢀshownꢀinꢀFig.ꢀ5.ꢀTheꢀyieldꢀfurtherꢀdropsꢀfromꢀ28%ꢀinꢀ
theꢀHZSM‐35ꢀzeoliteꢀtoꢀjustꢀ3%ꢀinꢀZ–Cs–5wtꢀandꢀevenꢀtoꢀzeroꢀinꢀ
Z–Cs–10wt,ꢀ indicatingꢀ thatꢀ theꢀ supportingꢀ cesiumꢀ oxideꢀ doesꢀ
harmꢀ ratherꢀ thanꢀ goodꢀ inꢀ theꢀ aldolꢀ condensationꢀ reactionꢀ ofꢀ
DMMꢀandꢀMAcꢀtoꢀproduceꢀMAꢀandꢀAA.ꢀ ꢀ
1
50
300
450
600
o
T ( C)
Fig.ꢀ3.ꢀAcidicꢀpropertiesꢀofꢀzeolitesꢀsupportedꢀbyꢀvaryingꢀweightꢀper‐
centagesꢀofꢀcesium,ꢀasꢀcharacterizedꢀbyꢀNH ‐TPD.ꢀ
3
proportionꢀofꢀtheꢀacidꢀsitesꢀisꢀcoveredꢀbyꢀalkalineꢀmetalꢀoxides.ꢀ
Ifꢀtheꢀweightꢀpercentageꢀofꢀcesiumꢀincreasesꢀtoꢀ5ꢀwt%,ꢀorꢀevenꢀ
0ꢀwt%,ꢀnoꢀpeaksꢀcorrespondingꢀtoꢀammoniaꢀdesorptionꢀcouldꢀ
beꢀdetectedꢀatꢀeitherꢀlowꢀtemperaturesꢀorꢀhighꢀtemperatures,ꢀ
whichꢀsuggestsꢀthatꢀnoꢀacidicꢀsitesꢀexistꢀonꢀtheseꢀtwoꢀsamples.ꢀ
Itꢀalsoꢀmeansꢀthatꢀthereꢀareꢀnoꢀacidꢀsitesꢀavailableꢀforꢀtheꢀaldolꢀ
condensationꢀreaction.ꢀWeꢀproposeꢀthatꢀtheseꢀacidicꢀsitesꢀareꢀ
coveredꢀbyꢀcesiumꢀoxideꢀafterꢀtakingꢀintoꢀaccountꢀtheꢀfactꢀthatꢀ
mostꢀofꢀtheꢀporesꢀareꢀblockedꢀatꢀsuchꢀhighꢀweightꢀpercentagesꢀ
ofꢀcesium,ꢀasꢀdescribedꢀinꢀTableꢀ1.ꢀ
1
3
.2.ꢀ ꢀ EffectꢀofꢀbaseꢀsitesꢀonꢀtheꢀaldolꢀcondensationꢀofꢀDMMꢀandꢀ
MAcꢀ ꢀ
DMMꢀdecomposesꢀreadilyꢀoverꢀacidicꢀzeoliteꢀcatalystꢀwithꢀaꢀ
conversionꢀofꢀ100%,ꢀasꢀdepictedꢀinꢀFig.ꢀ4.ꢀItsꢀconversionꢀdoesꢀ
1
00
100
80
60
40
20
0
Zeolite
8
6
4
2
0
0
0
0
0
ZCs1wt
ZCs5wt
ZCs10wt
1
2
3
4
5
TOS (h)
Fig.ꢀ4.ꢀEffectꢀofꢀcesiumꢀweightꢀpercentageꢀonꢀDMMꢀandꢀMAcꢀconversionꢀ
overꢀzeoliteꢀsupports.ꢀCesiumꢀweightꢀpercentageꢀwasꢀbasedꢀonꢀthatꢀof
Ourꢀresultsꢀshowꢀthatꢀcooperationꢀbetweenꢀbasesꢀandꢀacidsꢀ
isꢀnotꢀhelpfulꢀforꢀtheꢀaldolꢀcondensationꢀreaction;ꢀthisꢀobserva‐
tionꢀ isꢀ inconsistentꢀ withꢀ theꢀ conclusionꢀ ofꢀ publishedꢀ papers,ꢀ
whichꢀreportedꢀthatꢀaꢀcombinationꢀofꢀbaseꢀandꢀacidꢀenhancedꢀ
cesiumꢀ element;ꢀ Zeoliteꢀrefersꢀ toꢀ HZSM‐35,ꢀ SiO
temperatureꢀofꢀ350ꢀ°C,ꢀsystemꢀpressureꢀofꢀ3ꢀMPaꢀwithꢀN
2
/Al
2
O
3
ꢀ=ꢀ 79,ꢀreactionꢀ
ꢀequilibriumꢀ
2
gas,ꢀnDMM/nMAcꢀ=ꢀ2/1.ꢀTheꢀsaturatedꢀvaporꢀpressuresꢀofꢀDMMꢀandꢀMAc
wereꢀ43.3ꢀandꢀ23.0ꢀkPa,ꢀrespectively.ꢀ

ꢀ
ZhanlingꢀMaꢀetꢀal.ꢀ/ꢀChineseꢀJournalꢀofꢀCatalysisꢀ39ꢀ(2018)ꢀ1129–1137ꢀ
1133ꢀ
tratedꢀinꢀFig.ꢀS3.ꢀThisꢀfactꢀshowsꢀthatꢀacidicꢀsitesꢀareꢀindispen‐
sableꢀforꢀtheꢀaldolꢀcondensationꢀreactionꢀofꢀDMMꢀandꢀMAcꢀtoꢀ
produceꢀacrylicꢀacidꢀandꢀmethylꢀacrylate.ꢀ ꢀ
Now,ꢀitꢀisꢀclearꢀthatꢀacidicꢀsitesꢀareꢀessentialꢀtoꢀtriggerꢀtheꢀ
aldolꢀcondensationꢀreactionꢀofꢀDMMꢀandꢀMAc.ꢀWithoutꢀacidicꢀ
sites,ꢀ thisꢀ reactionꢀ wouldꢀ notꢀ occur.ꢀ However,ꢀ aꢀ strongꢀ acidꢀ
contributesꢀ mainlyꢀ toꢀ theꢀ by‐reactionsꢀ ratherꢀ thanꢀ theꢀ aldolꢀ
condensationꢀreaction.ꢀItꢀseemsꢀthatꢀmediumꢀorꢀweakꢀacidsꢀareꢀ
40
30
20
10
0
Zeolite
Z-Cs-1wt
Z-Cs-5wt
Z-Cs-10wt
favorableꢀ forꢀ thisꢀ reaction.ꢀ Therefore,ꢀ aꢀ typeꢀ ofꢀ γ–Al
2
O ꢀ withꢀ
3
abundantꢀweakꢀacidꢀsites,ꢀtogetherꢀwithꢀaꢀfewꢀmedium‐strongꢀ
acidꢀsites,ꢀwasꢀappliedꢀasꢀaꢀcatalystꢀtoꢀverifyꢀthisꢀhypothesis.ꢀ ꢀ
2 3
3.3.ꢀ ꢀ Efficiencyꢀofꢀγ‐Al O ꢀinꢀtheꢀaldolꢀcondensationꢀreactionꢀ ꢀ
Theꢀγ‐Al ꢀsampleꢀconsistsꢀofꢀAlO
2
O
3
6
ꢀoctahedraꢀandꢀAlO ꢀtet‐
4
1
2
3
4
5
rahedraꢀ withꢀ aꢀ slightlyꢀ tetragonallyꢀ distortedꢀ face‐centeredꢀ
cubicꢀ latticeꢀ structure.ꢀ Itsꢀ Lewisꢀ acidityꢀ isꢀ generallyꢀ derivedꢀ
fromꢀtheꢀcoordinativelyꢀunsaturatedꢀaluminumꢀionsꢀ[26,27].ꢀInꢀ
TOS (h)
Fig.ꢀ5.ꢀVariationꢀinꢀMAꢀandꢀAAꢀyieldꢀwithꢀreactionꢀtimeꢀoverꢀzeolites
supportedꢀ withꢀ differentꢀ weightꢀ percentagesꢀ ofꢀ cesium.ꢀ Zeolite:
ꢀ=ꢀ79,ꢀ350ꢀ°C,ꢀ3ꢀMPaꢀwithꢀN ꢀequilibriumꢀgas,ꢀand
DMM/nMAcꢀ=ꢀ2/1.ꢀTheꢀsaturatedꢀvaporꢀpressuresꢀofꢀDMMꢀandꢀMAcꢀwere
thisꢀstudy,ꢀtheꢀacidicꢀpropertiesꢀofꢀγ‐Al
2
O ꢀareꢀstudiedꢀbyꢀFT‐IRꢀ
3
HZSM‐35,ꢀSiO
n
2
/Al
2
O
3
2
characterizationꢀ usingꢀ pyridineꢀ asꢀ theꢀ probeꢀ molecule,ꢀ asꢀ
shownꢀinꢀFig.ꢀ6.ꢀThreeꢀnegativeꢀpeaksꢀatꢀ3733,ꢀ3677,ꢀandꢀ3600ꢀ
43.3ꢀandꢀ23.0ꢀkPa,ꢀrespectively.
–1
cm ꢀ appearꢀ inꢀ theꢀ OHꢀ regionꢀ withꢀ aꢀ smallꢀ shoulderꢀ peakꢀ atꢀ
theꢀ aldolꢀ condensationꢀ reaction.ꢀ Toꢀ analyzeꢀ thisꢀ discrepancy,ꢀ
weꢀperformedꢀtheꢀ aldolꢀ condensationꢀofꢀDMMꢀ andꢀ MAcꢀoverꢀ
pureꢀalkalineꢀmetalꢀoxideꢀcatalysts,ꢀsuchꢀasꢀNa
–1
aroundꢀ 3751ꢀ cm ꢀ uponꢀ theꢀ adsorptionꢀ ofꢀ pyridineꢀ vaporꢀ atꢀ
–1
roomꢀ temperature.ꢀ Theꢀ peakꢀ atꢀ 3600ꢀ cm ꢀ disappearsꢀ afterꢀ
2
O,ꢀK
2
O,ꢀandꢀCs Oꢀ
2
desorptionꢀunderꢀvacuumꢀforꢀ30ꢀminꢀatꢀ150ꢀ°Cꢀfollowedꢀbyꢀtheꢀ
supportedꢀ onꢀ non‐acidicꢀ silicaꢀ preparedꢀ byꢀ aꢀ traditionalꢀ wet‐
nessꢀimpregnationꢀmethod,ꢀasꢀshownꢀinꢀTableꢀS1.ꢀSurprisingly,ꢀ
theꢀDMMꢀprecursorꢀhardlyꢀdecomposedꢀoverꢀtheseꢀpureꢀalka‐
lineꢀ metalꢀ oxideꢀ catalystsꢀ andꢀ couldꢀ notꢀ provideꢀ enoughꢀ for‐
maldehydeꢀtoꢀensureꢀtheꢀoccurrenceꢀofꢀaldolꢀcondensationꢀwithꢀ
theꢀMAcꢀprecursor.ꢀOnꢀtheꢀcontrary,ꢀDMMꢀdecomposesꢀimme‐
diatelyꢀ intoꢀ formaldehydeꢀ uponꢀ contactꢀ withꢀ acidicꢀ catalysts,ꢀ
forꢀinstanceꢀtheꢀH‐formꢀofꢀtheꢀzeolite,ꢀwhoseꢀmassꢀsignalꢀ(m/zꢀ=ꢀ
disappearanceꢀ ofꢀ theꢀ peaksꢀ atꢀ 1611,ꢀ 1591,ꢀ andꢀ 1443ꢀ cm–1,ꢀ
whichꢀ areꢀ bandsꢀ typicalꢀ ofꢀ physicallyꢀ absorbedꢀ pyridine,ꢀ butꢀ
–1
oneꢀtypicalꢀbandꢀatꢀ1452ꢀcm ꢀofꢀpyridineꢀabsorbedꢀonꢀLewisꢀ
acidsꢀcanꢀstillꢀbeꢀobservedꢀafterꢀdesorptionꢀ[28–31].ꢀTheꢀthreeꢀ
–1
bandsꢀatꢀ1624,ꢀ1617,ꢀandꢀ1579ꢀcm ꢀareꢀassignedꢀtoꢀthreeꢀtypesꢀ
3+
ofꢀLewisꢀacidꢀsitesꢀofꢀfive,ꢀfour,ꢀandꢀthree‐coordinateꢀAl ꢀions,ꢀ
–1
respectively.ꢀ Byꢀ integratingꢀ theꢀ peakꢀ atꢀ 1452ꢀ cm ,ꢀ whichꢀ
standsꢀforꢀtheꢀinteractionꢀofꢀpyridineꢀwithꢀtheꢀtotalꢀLewisꢀacidsꢀ
available,ꢀ weꢀ canꢀ calculateꢀ thatꢀ theꢀ totalꢀ acidꢀ amountꢀ ofꢀ theꢀ
29)ꢀ isꢀ clearlyꢀ recordedꢀ byꢀ massꢀ spectrometryꢀ (MS),ꢀ asꢀ illus‐
1443
ad-Py-RT
o
de-Py-150 C
o
de-Py-350 C
o
de-Py-450 C
1452
1
579
611
1591
1
1617
1624
4000
3800
3600
3400
3200 1650
1600 1550 1500
1450
1400
1
Wavenumber cm 
Fig.ꢀ6.ꢀFT‐IRꢀspectraꢀofꢀtheꢀγ‐Al
2
O
3
ꢀsampleꢀwithꢀpyridineꢀasꢀaꢀprobeꢀmolecule.ꢀAllꢀtheꢀaboveꢀspectraꢀdetractedꢀfromꢀthatꢀofꢀtheꢀpureꢀγ‐Al
2
O ꢀsample.ꢀ
3

1134ꢀ
ZhanlingꢀMaꢀetꢀal.ꢀ/ꢀChineseꢀJournalꢀofꢀCatalysisꢀ39ꢀ(2018)ꢀ1129–1137ꢀ
γ‐Al
2
O
3
ꢀ sampleꢀ isꢀ 0.14ꢀ mmol/g.ꢀ Theꢀ strengthꢀ ofꢀ theꢀ peakꢀ atꢀ
100
–1
1452ꢀcm ꢀdecreasesꢀgreatlyꢀafterꢀdesorptionꢀatꢀ350ꢀ°Cꢀforꢀ30ꢀ
DMM conversion
MAc conversion
minꢀwithꢀanꢀacidꢀamountꢀofꢀ0.02ꢀmmol/gꢀandꢀevenꢀdisappearsꢀ
atꢀ450ꢀ°C.ꢀAtꢀsuchꢀhighꢀdesorptionꢀtemperatures,ꢀonlyꢀinterac‐
tionsꢀbetweenꢀpyridineꢀwithꢀmedium‐strongꢀandꢀstrongꢀLewisꢀ
acidsꢀ couldꢀ beꢀ detectedꢀ inꢀ theꢀ FT‐IRꢀ spectrum.ꢀ However,ꢀ weꢀ
HZSM-35
80
60
40
-Al O3
2
observedꢀoneꢀsmallꢀabsorptionꢀpeakꢀinꢀtheꢀγ‐Al
suchꢀhighꢀtemperatures.ꢀItꢀmeansꢀthatꢀweakꢀLewisꢀacids,ꢀratherꢀ
thanꢀ strongꢀ acids,ꢀ existꢀ predominantlyꢀ inꢀ theꢀ γ‐Al ꢀ sampleꢀ
togetherꢀwithꢀaꢀsmallꢀquantityꢀofꢀmedium‐strongꢀLewisꢀacids.ꢀ
NH ‐TPDꢀ resultsꢀ ofꢀ theꢀ γ‐Al ꢀ sample,ꢀ asꢀ shownꢀ inꢀ Fig.ꢀ S4,ꢀ
2 3
O ꢀ sampleꢀ atꢀ
O
2 3
3
O
2 3
indicatedꢀanꢀobviousꢀdesorptionꢀpeakꢀofꢀweak‐strengthꢀacidsꢀatꢀ
aꢀlowꢀtemperatureꢀofꢀ200ꢀ°C,ꢀwhichꢀsupportedꢀourꢀhypothesis.ꢀ
TheꢀaldolꢀcondensationꢀreactionꢀofꢀDMMꢀandꢀMAcꢀwasꢀcon‐
20
MA+AA yield
ductedꢀoverꢀtheꢀweakꢀLewisꢀacidicꢀsampleꢀ(γ‐Al
whetherꢀtheꢀaldolꢀcondensationꢀreactionꢀcanꢀbeꢀtriggered.ꢀTheꢀ
reactionꢀ performanceꢀ overꢀ theꢀ γ‐Al ꢀ sampleꢀ isꢀ depictedꢀ inꢀ
Fig.ꢀ 7.ꢀ DMMꢀ isꢀ completelyꢀ convertedꢀ withꢀ aꢀ conversionꢀ ofꢀ
00%,ꢀindicatingꢀthatꢀsuchꢀaꢀweakꢀacidꢀisꢀcapableꢀofꢀcatalyzingꢀ
theꢀdecompositionꢀofꢀDMMꢀintoꢀformaldehyde.ꢀSimilarly,ꢀMAcꢀ
conversionꢀ isꢀ veryꢀ closeꢀ toꢀ thatꢀ obtainedꢀ withꢀ HZSM‐35.ꢀ Butꢀ
thisꢀprocessꢀproceedsꢀmuchꢀmoreꢀsmoothlyꢀwithoutꢀanꢀobviousꢀ
2
O
3
)ꢀtoꢀevaluateꢀ
0
1
2
3
4
5
TOS (h)
O
2 3
Fig.ꢀ8.ꢀComparisonꢀofꢀtheꢀaldolꢀcondensationꢀreactionꢀperformancesꢀofꢀ
γ‐Al 3ꢀandꢀHZSM‐35.ꢀReactionꢀconditions:ꢀ350ꢀ°Cꢀandꢀ3ꢀMPaꢀwithꢀN ꢀasꢀ
theꢀ equilibriumꢀ gas;ꢀ Theꢀ saturatedꢀ pressuresꢀ ofꢀ DMMꢀ andꢀ MAcꢀ wereꢀ
2
O
2
1
–1
–1
43.3ꢀandꢀ23.0ꢀkPa,ꢀrespectively,ꢀGHSVꢀ=ꢀ7.2ꢀLꢀg ꢀh .ꢀ
canꢀ beꢀ convertedꢀ muchꢀ moreꢀ smoothlyꢀ withꢀ γ‐Al
2
O ꢀ withoutꢀ
3
initialꢀperiod.ꢀItꢀisꢀevidentꢀthatꢀtheꢀγ‐Al
convertingꢀbothꢀDMMꢀandꢀMAcꢀprecursorsꢀtoꢀproduceꢀMAꢀandꢀ
AA.ꢀTheꢀyieldꢀofꢀMAꢀandꢀAAꢀisꢀ~15%ꢀoverꢀtheꢀγ‐Al ꢀsample.ꢀ
O
2 3
ꢀsampleꢀisꢀcapableꢀofꢀ
significantꢀfluctuationsꢀthanꢀwithꢀtheꢀHZSM‐35ꢀzeolite.ꢀForꢀex‐
ample,ꢀtheꢀinitialꢀconversionꢀofꢀMAcꢀinꢀHZSM‐35ꢀzeoliteꢀisꢀveryꢀ
highꢀandꢀthenꢀfallsꢀfastꢀinꢀtheꢀfirstꢀ2ꢀh.ꢀTheꢀinitialꢀhighꢀconver‐
sionꢀofꢀMAcꢀisꢀattributedꢀtoꢀtheꢀsevereꢀMTO‐likeꢀreactionꢀcata‐
lyzedꢀ byꢀ strongꢀ Brönstedꢀ acids,ꢀ asꢀ describedꢀ inꢀ Fig.ꢀ S5.ꢀ Thisꢀ
by‐reactionꢀisꢀmildlyꢀcatalyzedꢀbyꢀtheꢀweakꢀLewisꢀacidꢀofꢀtheꢀ
O
2 3
Thoughꢀ theꢀ yieldꢀ isꢀ lowerꢀ thanꢀ thatꢀ onꢀ HZSM‐35ꢀ zeolite,ꢀ theꢀ
productsꢀobtainedꢀareꢀmoreꢀstableꢀwithꢀaꢀshorterꢀinitialꢀperiod.ꢀ
Toꢀ excludeꢀ systematicꢀ errors,ꢀ weꢀ conductedꢀ theꢀ experimentꢀ
onceꢀagainꢀatꢀtheꢀsameꢀreactionꢀconditionsꢀ(blueꢀlineꢀinꢀFig.ꢀ7)ꢀ
andꢀ achievedꢀ anꢀ excellentꢀ repeatability.ꢀ Therefore,ꢀ itꢀ isꢀ clearꢀ
thatꢀaꢀweakꢀLewisꢀacidꢀisꢀenoughꢀtoꢀtriggerꢀtheꢀaldolꢀcondensa‐
tionꢀreactionꢀofꢀDMMꢀandꢀMAcꢀtoꢀproduceꢀMAꢀandꢀAAꢀwithꢀbothꢀ
goodꢀactivityꢀandꢀstability.ꢀ ꢀ
γ‐Al
MAꢀandꢀAAꢀwithꢀtheꢀHZSM‐35ꢀzeoliteꢀisꢀalmostꢀdoubleꢀofꢀthatꢀ
obtainedꢀwithꢀtheꢀγ‐Al ꢀsample,ꢀindicatingꢀthatꢀaꢀcooperationꢀ
2 3
O ꢀsampleꢀwithꢀaꢀmuchꢀshorterꢀinitialꢀperiod.ꢀTheꢀyieldꢀofꢀ
O
2 3
betweenꢀstrongꢀacidsꢀandꢀweakꢀacidsꢀisꢀhelpfulꢀforꢀaldolꢀcon‐
densationꢀreactions.ꢀ
Aꢀ comparisonꢀ betweenꢀ theꢀ aldolꢀ condensationꢀ reactionꢀ
Theꢀ TGAꢀ curvesꢀ andꢀ firstꢀ derivativesꢀ correspondingꢀ toꢀ
performancesꢀ ofꢀ theꢀ γ‐Al
2 3
O ꢀ sampleꢀ andꢀ HZSM‐35ꢀ zeoliteꢀ isꢀ
2 3
cokedꢀ γ‐Al O ꢀandꢀdeactivatedꢀHZSM‐35ꢀzeoliteꢀareꢀshownꢀinꢀ
depictedꢀinꢀFig.ꢀ8.ꢀDMMꢀdecomposesꢀreadilyꢀwithꢀbothꢀcatalystsꢀ
withꢀ100%ꢀconversion,ꢀwhichꢀprovesꢀthatꢀanꢀacidꢀisꢀnecessaryꢀ
forꢀDMMꢀdecomposition.ꢀInꢀtheꢀcaseꢀofꢀtheꢀ MAcꢀ precursor,ꢀitꢀ
Fig.ꢀ 9.ꢀ Theꢀ amountꢀ ofꢀ cokeꢀ inꢀ γ‐Al
2 3
O
ꢀ wasꢀ 8.3ꢀ wt%,ꢀ whichꢀ isꢀ
largerꢀ thanꢀ thatꢀ ofꢀ HZSM‐35ꢀ zeoliteꢀ (6.9ꢀ wt%).ꢀ However,ꢀ theꢀ
deactivationꢀofꢀHZSM‐35ꢀzeoliteꢀisꢀmuchꢀmoreꢀsevere.ꢀForꢀex‐
ample,ꢀtheꢀconsumptionꢀtemperatureꢀofꢀcokeꢀofꢀHZSM‐35ꢀwasꢀ
1
00
100
80
60
40
20
st
1
DMM
DMM
nd
0.06
0.04
1
00
96
92
2
2
st
392
-Al O
2
3
8
6
4
2
0
0
0
0
HZSM-35
st
1
MAc
MAc
nd
599
0.02
1 MA+AA
nd
2
MA+AA
4
0.00
0
1
2
3
5
1
50
300
450
Temperature ( C)
Fig.ꢀ9.ꢀTheꢀTGAꢀcurvesꢀ(theꢀsolidꢀline)ꢀandꢀderivativesꢀ(theꢀdottedꢀline)ꢀ
600
750
TOS (h)
o
Fig.ꢀ7.ꢀAldolꢀcondensationꢀreactionꢀperformanceꢀoverꢀtheꢀγ‐Al
pleꢀwithꢀaꢀgoodꢀrepeatability.ꢀReactionꢀtemperatureꢀofꢀ350ꢀ°C,ꢀ3ꢀMPa
withꢀ N ꢀ asꢀ theꢀ equilibriumꢀ gas,ꢀ theꢀ saturatedꢀ pressuresꢀ ofꢀ DMMꢀ and
MAcꢀwereꢀ43.3ꢀandꢀ23.0ꢀkPa,ꢀrespectively,ꢀGHSVꢀ=ꢀ7.2ꢀLꢀg ꢀh .ꢀ
2
O ꢀsam‐
3
2
ofꢀaꢀcokedꢀγ‐Al
reactionꢀinꢀair.ꢀ
2
O ꢀsampleꢀandꢀdeactivatedꢀHZSM‐35ꢀzeoliteꢀafterꢀ6ꢀhꢀofꢀ
3
–1
–1

ꢀ
ZhanlingꢀMaꢀetꢀal.ꢀ/ꢀChineseꢀJournalꢀofꢀCatalysisꢀ39ꢀ(2018)ꢀ1129–1137ꢀ
1135ꢀ
5
99ꢀ°C,ꢀwhichꢀisꢀhigherꢀthanꢀthatꢀofꢀtheꢀγ‐Al
2
O
3
ꢀsampleꢀ(392ꢀ°C).ꢀ
tionꢀreactionꢀofꢀDMMꢀandꢀMAcꢀtoꢀprepareꢀacrylicꢀacidꢀandꢀme‐
thylꢀacrylateꢀbecauseꢀtheꢀDMMꢀprecursorꢀdoesꢀnotꢀdecomposeꢀ
intoꢀformaldehydeꢀoverꢀbasicꢀsites.ꢀAcidꢀsitesꢀareꢀindispensableꢀ
forꢀcatalyzingꢀtheꢀreactionꢀwhenꢀusingꢀDMMꢀasꢀtheꢀsourceꢀofꢀ
formaldehyde.ꢀ However,ꢀ notꢀ allꢀ acidicꢀ sitesꢀ inꢀ theꢀ HZSM‐35ꢀ
zeoliteꢀareꢀnecessaryꢀforꢀtheꢀreaction.ꢀStrongꢀacidsꢀcontributeꢀ
mainlyꢀ toꢀ anꢀ unavoidableꢀ methanol‐to‐olefin‐likeꢀ process,ꢀ re‐
sultingꢀinꢀanꢀinitialꢀperiodꢀofꢀtheꢀreactionꢀduringꢀwhichꢀaꢀlargeꢀ
amountꢀ ofꢀ hydrocarbonꢀ byproductsꢀ areꢀ generated.ꢀ Medi‐
um‐strongꢀacidsꢀandꢀweakꢀacidsꢀareꢀoptimalꢀcandidatesꢀforꢀtheꢀ
aldolꢀcondensationꢀreactionꢀofꢀDMMꢀandꢀMAcꢀtoꢀproduceꢀacryl‐
icꢀ acidꢀ andꢀ itsꢀ esterꢀ withoutꢀ causingꢀ severeꢀ by‐reactions.ꢀ Aꢀ
goodꢀ reactionꢀ performanceꢀ withꢀ aꢀ considerablyꢀ highꢀ activityꢀ
DeactivationꢀinꢀHZSM‐35ꢀzeoliteꢀisꢀviolentꢀwithꢀhardꢀcoke.ꢀUn‐
likeꢀtheꢀγ‐Al ꢀsample,ꢀHZSM‐35ꢀzeoliteꢀpossessesꢀaꢀlargeꢀper‐
centageꢀofꢀstrongꢀBrönstedꢀacids,ꢀwhichꢀseemꢀtoꢀbeꢀresponsibleꢀ
forꢀthisꢀsevereꢀdeactivation.ꢀ ꢀ
O
2 3
3.4.ꢀ ꢀ Discussionꢀ
Theꢀ aldolꢀ condensationꢀ reactionꢀ ofꢀ formaldehydeꢀ andꢀ me‐
thylꢀacetateꢀcanꢀbeꢀcatalyzedꢀbyꢀeitherꢀbasesꢀorꢀacidsꢀorꢀbothꢀofꢀ
them.ꢀInꢀmostꢀofꢀtheꢀpublishedꢀreports,ꢀattentionꢀwasꢀpaidꢀtoꢀ
catalystsꢀwithꢀacid‐baseꢀbifunctionalꢀproperties,ꢀsuchꢀasꢀV–P–Siꢀ
orꢀalkalineꢀmetalꢀoxideꢀsamplesꢀimpregnatedꢀwithꢀotherꢀmetals,ꢀ
owingꢀtoꢀtheirꢀbetterꢀperformanceꢀthanꢀpureꢀacidꢀorꢀpureꢀbaseꢀ
catalysts.ꢀHowever,ꢀinꢀthisꢀstudy,ꢀtheꢀZ–Cs–1wtꢀsample,ꢀwhichꢀ
possessedꢀbothꢀacidityꢀandꢀalkalinityꢀperformedꢀpoorlyꢀwithꢀaꢀ
lowerꢀyieldꢀofꢀMAꢀandꢀAAꢀthanꢀtheꢀpureꢀH‐formꢀofꢀacidicꢀzeolite.ꢀ
Thisꢀ resultꢀ showsꢀ thatꢀ alkalinityꢀ isꢀ notꢀ helpful,ꢀ inꢀ factꢀ itꢀ isꢀ
harmful,ꢀforꢀtheꢀaldolꢀcondensationꢀreactionꢀofꢀDMMꢀandꢀMAcꢀ
toꢀproduceꢀMAꢀandꢀAA.ꢀThisꢀisꢀinconsistentꢀwithꢀpublishedꢀre‐
ports.ꢀToꢀunderstandꢀthisꢀdiscrepancy,ꢀweꢀconductedꢀtheꢀreac‐
andꢀstabilityꢀareꢀobtainedꢀwithꢀtheꢀγ‐Al
2
O ꢀsample,ꢀwhichꢀpos‐
3
sessesꢀ abundantꢀ weakꢀ acidꢀ sitesꢀ withꢀ aꢀ fewꢀ medium‐strongꢀ
acidꢀ sites.ꢀ Furtherꢀ workꢀ onꢀ evaluatingꢀ theꢀ maximumꢀ acidityꢀ
andꢀamountꢀofꢀacidꢀforꢀtriggeringꢀtheꢀaldolꢀcondensationꢀreac‐
tionꢀisꢀrequired.ꢀ
Referencesꢀ
tionꢀoverꢀpureꢀalkalineꢀmetalꢀoxideꢀcatalysts,ꢀsuchꢀasꢀNa
2 2
O,ꢀK O,ꢀ
[1] M.ꢀM.ꢀBettahar,ꢀG.ꢀCostentin,ꢀL.ꢀSavary,ꢀJ.ꢀC.ꢀLavalley,ꢀAppl.ꢀCatal.ꢀA,ꢀ
1
996,ꢀ145,ꢀ1–48.ꢀ
2] W.ꢀFang,ꢀQ.ꢀJ.ꢀGe,ꢀJ.ꢀF.ꢀYu,ꢀH.ꢀY.ꢀXu,ꢀInd.ꢀEng.ꢀChem.ꢀRes.,ꢀ2011,ꢀ50,ꢀ
962–1967.ꢀ
3] W.ꢀ E.ꢀ Campbell,ꢀ E.ꢀ L.ꢀ McDaniel,ꢀ W.ꢀ H.ꢀ Reece,ꢀ J.ꢀ E.ꢀ Williams,ꢀ H.ꢀS.ꢀ
Young,ꢀInd.ꢀEng.ꢀChem.ꢀProd.ꢀRes.ꢀDevelop.,ꢀ1970,ꢀ9,ꢀ325–334.ꢀ
andꢀCs O,ꢀandꢀfoundꢀthatꢀtheꢀDMMꢀprecursorꢀdoesꢀnotꢀdecom‐
2
[
[
[
[
poseꢀ intoꢀ formaldehydeꢀ overꢀ theseꢀ alkalineꢀ sites.ꢀ Butꢀ DMMꢀ
decomposesꢀimmediatelyꢀintoꢀformaldehydeꢀoverꢀacidicꢀzeoliteꢀ
andꢀenhancesꢀtheꢀaldolꢀcondensationꢀreactionꢀofꢀDMMꢀandꢀMAcꢀ
toꢀproduceꢀMAꢀandꢀAA.ꢀThisꢀshowsꢀthatꢀacidꢀsitesꢀareꢀindispen‐
sableꢀ forꢀ theꢀ aldolꢀ condensationꢀ reactionꢀ ofꢀ DMMꢀ andꢀ MAc.ꢀ
Here,ꢀitꢀshouldꢀbeꢀpointedꢀoutꢀthatꢀthisꢀhypothesisꢀisꢀsuitableꢀ
onlyꢀwhenꢀDMMꢀisꢀusedꢀasꢀtheꢀsourceꢀforꢀformaldehyde.ꢀ ꢀ
Strongꢀ acidsꢀ andꢀ weakꢀ acidsꢀ existedꢀ inꢀ theꢀ HZSM‐35ꢀ asꢀ
1
4] A.ꢀX.ꢀXu,ꢀY.ꢀWang,ꢀH.ꢀQ.ꢀGe,ꢀS.ꢀChen,ꢀY.ꢀH.ꢀLi,ꢀW.ꢀM.ꢀLu,ꢀChin.ꢀJ.ꢀCatal.,ꢀ
2
013,ꢀ34,ꢀ2183–2191.ꢀ
5] W.ꢀ Fang,ꢀ Q.ꢀ J.ꢀ Ge,ꢀ J.ꢀ F.ꢀ Yu,ꢀ H.ꢀ Y.ꢀ Xu,ꢀ Chin.ꢀ J.ꢀ Catal.,ꢀ 2011,ꢀ 32,ꢀ
022–1026.ꢀ ꢀ
1
[6] P.ꢀCheung,ꢀA.ꢀBhan,ꢀG.ꢀJ.ꢀSunley,ꢀD.ꢀJ.ꢀLaw,ꢀE.ꢀIglesia,ꢀJ.ꢀCatal.,ꢀ2007,ꢀ
245,ꢀ110–123.ꢀ
shownꢀ inꢀ theꢀ NH
‐TPDꢀ results,ꢀ withꢀ twoꢀ desorptionꢀ peaksꢀ ofꢀ
3
ammoniaꢀatꢀdifferentꢀtemperatures.ꢀAfterꢀtheꢀadditionꢀofꢀ1ꢀwt%ꢀ
cesium,ꢀ mostꢀ ofꢀ theꢀ acidsꢀ inꢀ theꢀ HZSM‐35,ꢀ especiallyꢀ strongꢀ
acids,ꢀareꢀcoveredꢀbyꢀbaseꢀsites,ꢀfollowedꢀbyꢀtheꢀdisappearanceꢀ
ofꢀtheꢀinitialꢀperiod.ꢀInꢀthisꢀperiod,ꢀaꢀlargeꢀquantityꢀofꢀhydro‐
carbonꢀ byproductsꢀ areꢀ formedꢀ becauseꢀ ofꢀ theꢀ unavoidableꢀ
methanol‐to‐olefin‐likeꢀprocessꢀatꢀsuchꢀhighꢀreactionꢀtempera‐
tures.ꢀDuringꢀtheꢀinitialꢀreactionꢀperiodꢀofꢀtheꢀHZSM‐35ꢀzeolite,ꢀ
theꢀstrongꢀacidꢀsitesꢀmightꢀfirstꢀcatalyzeꢀtheꢀprecursorsꢀtoꢀun‐
dergoꢀaꢀhydrogenꢀtransferꢀprocess,ꢀresultingꢀinꢀaꢀlargeꢀamountꢀ
ofꢀ hydrocarbonꢀ byproducts.ꢀ Later,ꢀ theꢀ remainingꢀ medi‐
um‐strongꢀandꢀweakꢀacidsꢀcontinueꢀtoꢀcatalyzeꢀtheꢀaldolꢀcon‐
densationꢀreaction.ꢀThisꢀsuggestsꢀthatꢀnotꢀallꢀtheꢀacidicꢀsitesꢀofꢀ
theꢀzeoliteꢀparticipateꢀinꢀtheꢀaldolꢀcondensationꢀreaction.ꢀ ꢀ
Medium‐strongꢀ andꢀ weakꢀ acidꢀ sitesꢀ seemꢀ toꢀ beꢀ theꢀ primeꢀ
candidatesꢀ forꢀ catalyzingꢀ theꢀ aldolꢀ condensationꢀ reactionꢀ ofꢀ
DMMꢀ andꢀ MAcꢀ toꢀ produceꢀ MAꢀ andꢀ AAꢀ withoutꢀ causingꢀ anyꢀ
drasticꢀ by‐reactions.ꢀ Thisꢀ hypothesisꢀ isꢀ validatedꢀ byꢀ theꢀ con‐
[7] J.ꢀL.ꢀLiu,ꢀH.ꢀF.ꢀXue,ꢀX.ꢀM.ꢀHuang,ꢀP.ꢀH.ꢀWu,ꢀS.ꢀJ.ꢀHuang,ꢀS.ꢀB.ꢀLiu,ꢀW.ꢀJ.ꢀ
Shen,ꢀChin.ꢀJ.ꢀCatal.,ꢀ2010,ꢀ31,ꢀ729–738.ꢀ
[8] M.ꢀAi,ꢀAppl.ꢀCatal.ꢀA,ꢀ2005,ꢀ288,ꢀ211–215.ꢀ
[9] B.ꢀLi,ꢀR.ꢀY.ꢀYan,ꢀL.ꢀWang,ꢀY.Y.ꢀDiao,ꢀZ.ꢀX.ꢀLi,ꢀS.ꢀJ.ꢀZhang,ꢀInd.ꢀEng.ꢀ
Chem.ꢀRes.,ꢀ2014,ꢀ53,ꢀ1386–1394.ꢀ
[
[
[
10] B.ꢀLi,ꢀR.ꢀY.ꢀYan,ꢀL.ꢀWang,ꢀY.ꢀY.ꢀDiao,ꢀZ.ꢀX.ꢀLi,ꢀS.ꢀJ.ꢀZhang,ꢀCatal.ꢀLett.,ꢀ
2013,ꢀ143,ꢀ829–838.ꢀ
11] J.ꢀJ.ꢀSpivey,ꢀM.ꢀR.ꢀGogate,ꢀJ.ꢀR.ꢀZoeller,ꢀR.ꢀD.ꢀColberg,ꢀInd.ꢀEng.ꢀChem.ꢀ
Res.,ꢀ1997,ꢀ36,ꢀ4600–4608.ꢀ
12] M.ꢀAi,ꢀAppl.ꢀCatal.,ꢀ1990,ꢀ63,ꢀ365–373.ꢀ
[13] M.ꢀAi,ꢀBull.ꢀChem.ꢀSoc.ꢀJpn.,ꢀ1990,ꢀ63,ꢀ3722–3724.ꢀ
[14] M.ꢀAi,ꢀAppl.ꢀCatal.,ꢀ1989,ꢀ48,ꢀ51–61.ꢀ
[15] M.ꢀAi,ꢀAppl.ꢀCatal.,ꢀ1989,ꢀ54,ꢀ29–36.ꢀ
16] M.ꢀAi,ꢀJ.ꢀCatal.,ꢀ1988,ꢀ112,ꢀ194–200.ꢀ
17] M.ꢀAi,ꢀJ.ꢀCatal.,ꢀ1987,ꢀ107,ꢀ201–208.ꢀ
18] M.ꢀ R.ꢀ Gogate,ꢀ J.ꢀ J.ꢀ Spivey,ꢀ J.ꢀ R.ꢀ Zoeller,ꢀ Catal.ꢀ Today,ꢀ 1997,ꢀ 36,ꢀ
[
[
[
2
43–254.ꢀ
19] E.ꢀ Dumitriu,ꢀ V.ꢀ Hulea,ꢀ N.ꢀ Bilba,ꢀ G.ꢀ Carja,ꢀ A.ꢀ Azzouz,ꢀ J.ꢀMol.ꢀCatal.,ꢀ
993,ꢀ79,ꢀ175–185.ꢀ
[
[
[
1
siderablyꢀ highꢀ activityꢀ andꢀ stabilityꢀ ofꢀ theꢀ γ‐Al
2
O ꢀ sampleꢀ inꢀ
3
20] R.ꢀTanner,ꢀP.ꢀGill,ꢀR.ꢀWells,ꢀJ.ꢀE.ꢀBailie,ꢀG.ꢀKelly,ꢀS.ꢀD.ꢀJackson,ꢀG.ꢀJ.ꢀ
Hutchings,ꢀPhys.ꢀChem.ꢀChem.ꢀPhys.,ꢀ2002,ꢀ4,ꢀ688–695.ꢀ
21] M.ꢀAi,ꢀJ.ꢀCatal.,ꢀ1987,ꢀ106,ꢀ273–279.ꢀ
[22] J.ꢀHu,ꢀZ.ꢀP.ꢀLu,ꢀH.ꢀB.ꢀYin,ꢀW.ꢀP.ꢀXue,ꢀA.ꢀL.ꢀWang,ꢀL.ꢀQ.ꢀShen,ꢀS.ꢀX.ꢀLiu,ꢀJ.ꢀ
Ind.ꢀEng.ꢀChem.,ꢀ2016,ꢀ40,ꢀ145–151.ꢀ
whichꢀ weakꢀ acidsꢀ existꢀ predominantlyꢀ alongꢀ withꢀ aꢀ smallꢀ
numberꢀmedium‐strongꢀacidꢀsites.ꢀItꢀseemsꢀthatꢀweakꢀacidsꢀareꢀ
powerfulꢀenoughꢀtoꢀcatalyzeꢀtheꢀaldolꢀcondensationꢀreaction.ꢀ ꢀ
4.ꢀ ꢀ Conclusionsꢀ
[23] D.ꢀYang,ꢀD.ꢀLi,ꢀH.ꢀY.ꢀYao,ꢀG.ꢀL.ꢀZhang,ꢀT.ꢀT.ꢀJiao,ꢀZ.ꢀX.ꢀLi,ꢀC.ꢀS.ꢀLi,ꢀS.ꢀJ.ꢀ
Zhang,ꢀInd.ꢀEng.ꢀChem.ꢀRes.,ꢀ2015,ꢀ54,ꢀ6865–6873.ꢀ
[24] X.ꢀZ.ꢀFeng,ꢀB.ꢀSun,ꢀY.ꢀYao,ꢀQ.ꢀSu,ꢀW.ꢀJ.ꢀJi,ꢀC.ꢀT.ꢀAu,ꢀJ.ꢀCatal.,ꢀ 2014,ꢀ ꢀ
Baseꢀsitesꢀareꢀnotꢀhelpfulꢀforꢀcatalyzingꢀtheꢀaldolꢀcondensa‐

1136ꢀ
ZhanlingꢀMaꢀetꢀal.ꢀ/ꢀChineseꢀJournalꢀofꢀCatalysisꢀ39ꢀ(2018)ꢀ1129–1137ꢀ
GraphicalꢀAbstractꢀ
Chin.ꢀJ.ꢀCatal.,ꢀ2018,ꢀ39:ꢀ1129–1137ꢀ ꢀ ꢀ doi:ꢀ10.1016/S1872‐2067(18)63069‐4
One‐stepꢀaldolꢀcondensationꢀreactionꢀofꢀdimethoxymethaneꢀandꢀ ꢀ
methylꢀacetateꢀoverꢀsupportedꢀCs/ZSM‐35ꢀzeoliteꢀcatalystsꢀ
ZhanlingꢀMa,ꢀXiangangꢀMa,ꢀHongchaoꢀLiu,ꢀWenliangꢀZhuꢀ*,ꢀXinwenꢀGuo,ꢀ
ZhongminꢀLiuꢀ*ꢀ
DalianꢀUniversityꢀofꢀTechnology;ꢀ ꢀ
DalianꢀInstituteꢀofꢀChemicalꢀPhysics,ꢀChineseꢀAcademyꢀofꢀSciences;ꢀ ꢀ
UniversityꢀofꢀChineseꢀAcademyꢀofꢀSciencesꢀ
ꢀ
ꢀ
ꢀ
AcidityꢀisꢀindispensableꢀforꢀtheꢀaldolꢀcondensationꢀreactionꢀofꢀDMMꢀwithꢀ
MAcꢀtoꢀprepareꢀacrylicꢀacidꢀandꢀitsꢀester,ꢀwhereasꢀalkalinityꢀisꢀharmful.ꢀ
314,ꢀ132–141.ꢀ
2016,ꢀ519,ꢀ155–164.ꢀ
[25] Z.ꢀL.ꢀMa,ꢀX.ꢀG.ꢀMa,ꢀH.ꢀC.ꢀLiu,ꢀY.ꢀL.ꢀHe,ꢀW.ꢀL.ꢀZhu,ꢀX.ꢀW.ꢀGuo,ꢀZ.ꢀM.ꢀLiu,ꢀ
Chem.ꢀCommun.,ꢀ2017,ꢀ53,ꢀ9071–9074.ꢀ
[29] Z.ꢀG.ꢀZhu,ꢀH.ꢀXu,ꢀJ.ꢀG.ꢀJiang,ꢀY.ꢀJ.ꢀGuan,ꢀP.ꢀWu,ꢀJ.ꢀCatal.,ꢀ2017,ꢀ352,ꢀ
1–12.ꢀ
[
26] X.ꢀS.ꢀLiu,ꢀR.ꢀE.ꢀTruitt,ꢀJ.ꢀAm.ꢀChem.ꢀSoc.,ꢀ1997,ꢀ119,ꢀ9856–9860.ꢀ
27] C.ꢀ Morterra,ꢀ V.ꢀ Bolis,ꢀ G.ꢀ Magnacca,ꢀ Langmuir,ꢀ 1994,ꢀ 10,ꢀ
[30] Z.ꢀC.ꢀYang,ꢀY.ꢀQ.ꢀLiu,ꢀJ.ꢀC.ꢀZhao,ꢀJ.ꢀX.ꢀGou,ꢀK.ꢀA.ꢀSun,ꢀC.ꢀG.ꢀLiu,ꢀChin.ꢀJ.ꢀ
Catal.,ꢀ2017,ꢀ38,ꢀ509–517.ꢀ
[
1
812–1824.ꢀ
[31] L.ꢀZ.ꢀYang,ꢀZ.ꢀY.ꢀLiu,ꢀZ.ꢀLiu,ꢀW.ꢀY.ꢀPeng,ꢀY.ꢀQ.ꢀLiu,ꢀC.ꢀG.ꢀLiu,ꢀChin.ꢀJ.ꢀ
Catal.,ꢀ2017,ꢀ38,ꢀ683–690.
[
28] Z.ꢀG.ꢀZhu,ꢀH.ꢀXu,ꢀJ.ꢀG.ꢀJiang,ꢀX.ꢀLiu,ꢀJ.ꢀH.ꢀDing,ꢀP.ꢀWu,ꢀAppl.ꢀCatal.ꢀA,ꢀ
Cs/ZSM-35 分子筛催化甲缩醛和乙酸甲酯发生一步法羟醛缩合反应
a,b,c
b
b
b,#
a
b,*
马占玲
, 马现刚 , 刘红超 , 朱文良 , 郭新闻 , 刘中民
a
大连理工大学化工与环境生命学部化工学院精细化工国家重点实验室, 辽宁大连116024
b
中国科学院大连化学物理研究所甲醇制烯烃国家工程实验室, 辽宁大连116023
c
中国科学院大学, 北京100049
摘要: 丙烯酸及其酯是重要的化工原料, 广泛应用于涂料、粘结剂、纤维等领域, 目前工业上常采用丙烯两段氧化法进行
制备. 然而该方法以石油基原料丙烯为源头, 采用 V/Mo/Bi 等金属催化剂, 不符合可持续发展理念, 且存在环境污染及氧
气下产物易过度氧化等问题. 因此, 如何高效、安全、大规模工业化制备丙烯酸及其酯是研究者追求的目标. 以乙酸甲酯
(
MAc) 和甲醛为原料, 通过羟醛缩合一步制备丙烯酸及其酯是一条完全不同于丙烯氧化法的合成路径, 原料均可由煤基甲
醇得到, 符合我国“富煤、贫油、少气”的基本能源结构, 且该方法碳原子利用率为 100%, 副产物仅为水, 属于绿色环保合成
路径.
羟醛缩合是典型的碳链增长反应, 可在酸性催化剂、碱性催化剂、以及酸碱双功能催化剂存在下发生. 碱性催化剂一
2
般为负载型碱金属氧化物, 例如以 SiO 为载体的负载型 Na, K, Cs 氧化物催化剂等, 但都存在活性组分流失的问题, 进而导
致催化剂的失活, 难以实现工业化. 酸碱双功能催化剂是目前研究的热点, 由于具有酸催化剂的高选择性和碱催化剂的高
活性, 其反应性能要远优于单一酸性催化剂和单一碱性催化剂, 广大研究者对此进行了深入广泛的研究, 目前基本处于实
验室阶段. 相对而言, 目前酸性催化剂上通过羟醛缩合反应制备丙烯酸及其酯的研究工作较少, 特别是以固体酸为催化剂
进行乙酸甲酯和甲醛气固相反应研究非常少见.
我们以甲缩醛为甲醛源, 创新性地采用固体硅铝分子筛为酸性催化剂, 催化甲缩醛 (DMM) 和 MAc 发生羟醛缩合反应
来制备丙烯酸. 硅铝分子筛具有较高的活性, 可高效地催化羟醛缩合反应, 且由于分子筛催化剂具有很好的再生性能, 即
使催化剂寿命较短, 也可采用流化床或移动床等反应器进行工业化, 因此存在良好的工业化前景. 为了进一步深入研究酸
性位和碱性位各自对 DMM 和 MAc 羟醛缩合反应的影响, 本文以 HZSM-35 分子筛为载体, 采用浸渍法制备不同碱金属铯
3
氧化物含量的催化剂, 利用氮气吸附/脱附方法和化学程序升温 (NH -TPD) 方法对其孔结构和酸性质进行表征, 并进一步

ꢀ
ZhanlingꢀMaꢀetꢀal.ꢀ/ꢀChineseꢀJournalꢀofꢀCatalysisꢀ39ꢀ(2018)ꢀ1129–1137ꢀ
1137ꢀ
考察催化剂的性能. 结果表明, 微孔体积随着碱金属 Cs 负载量的增加而逐渐减小, 当 Cs 负载量增加至 10 wt% 时, 样品微
3
3
孔体积从初始 0.105 cm /g 降至 0.063 cm /g. NH -TPD 结果显示, 当 Cs 负载量为 1 wt%, 酸性催化剂载体上的强酸和弱酸
3
活性位被大量碱性氧化物占据; 当负载量超过 5 wt% 时, 所有的酸性位均被覆盖. 随后考察负载不同碱金属含量分子筛的
羟醛缩合反应性能, 发现碱金属氧化物的引入不利于羟醛缩合反应的进行, 这主要是由于作为甲醛源的 DMM 只有在酸中
心上才能进行分解产生甲醛, 促使羟醛缩合反应顺利进行. 当采用 DMM 为甲醛源时, 体系中必须有酸性位存在. 同时得
知, 分子筛 HZSM-35 中强酸和弱酸均是羟醛缩合反应的有效酸性位, 但强酸同时催化原料发生类甲醇制烯烃过程, 致使大
量烃类副产物生成, 产生较重的积炭物种. 羟醛缩合反应在含有大量弱酸催化剂上 (如γ-Al
的活性和稳定性.
2 3
O ) 也可顺利进行, 且具有较高
关键词: 羟醛缩合反应; 甲缩醛; 乙酸甲酯; 分子筛; 酸性
收稿日期: 2018-02-26. 接受日期: 2018-04-02. 出版日期: 2018-06-05.
*
通讯联系人. 电话: (0411)84379998; 传真: (0411)84379038; 电子信箱: liuzm@dicp.ac.cn
#
通讯联系人. 电话: (0411)84379418; 传真: (0411)84379038; 电子信箱: wlzhu@dicp.ac.cn
本文的电子版全文由Elsevier出版社在ScienceDirect上出版(http://www.sciencedirect.com/science/journal/18722067).