SAPO-35分子筛的合成及其甲醇制烯烃反应性能

Chinese Journal of Catalysis 34 (2013) 798-807催化学报2013年第34卷第4期I www.chxb.cn值化管抓available at www. sciencedirect.comarabpolsScienceDirectEL SEVIERjournal homepage: www. elsevier.com/locate/chnjcArticleSynthesis of SAPO-35 molecular sieve and its catalytic properties inthe methanol-to-olefins reactionLI Bing, TIAN Peng, LI ]inzhe a, CHEN Jingrunab, YUAN Yangyang ab, SU Xiongab, FAN Dongab,WEI Yingxua, QI Yuea, LIU Zhongmina.*●Dalian National Laboratory for Clean Energy, National Engineering Laboratory for Methanol-to-Olefins, Dalian Institute of Chemical Physics, ChineseAcademy of Sciences, Dalian 116023, Liaoning, Chinab University of Chinese Academy of Sciences, Bejing 10049, ChinaARTICLEINFOABSTRACTArticle history:SAP0-35 molecular sieve samples with different Si contents were hydrothermally synthesized usingReceived 30 December 2012hexamethyleneimine as the template and characterized by XRD, XRF, SEM, MAS NMR, XPS and NzAccepted 30 lanuary 2013physisorption. Three SAP0-35 samples were tested as methanol-to-olefins catalysts. After the reac-Published 20 April2013tion, the evolution of coke species was investigated over SAPO-35 and SAPO 34 catalysts with simi-lar Si concentrations. A correlation between the cage size of the molecular sieves and the coke spe-cies was obtained.Keywords:SAPO-35o 2013, Dalian Institute of Chemical Physics, Chinese Academy of Sciences.Published by Elsevier B.V. All rights reserved.Methanol to olefnsSAP0-34Coke speciesDeactivation1. Introductionaminomethylation [4- _7]. Moreover, SAP0-35 is also exploredas an adsorbent for CO2/CH4 separation [8,9].Silicoaluminophosphate molecular sieves (SAPO-n) wereThe methanol-to-olefins (MTO) process is the key process tofirst synthesized and reported by Union Carbide Corporationproduce light olefins from coal or natural gas. The molecular(UCC) in 1984 [1]. Due to their mild acidity and special poresieves ZSM-5 and SAP0-34 with eight membered ring and CHAstructures, SAPO molecular sieves have widespread applica-cages exhibit excellent catalytic performance in the reactiontions in the chemical industry [2]. SAP0-35, as a member of the[6,7,10-12]. Other small pore molecular sieves with thefamily with the levyne-like crystal structure (LEV), is a small .eight-membered pore size such as SAP0-17, SAP0-18, andpore molecular sieve with a pore diameter of 3.6 nm x 4.8 nm.SAP0-35 have also been explored as catalysts for the MTO re-The framework of SAP0-35 comprises levyne cages connectedaction [7,13,14]. The SAPO-35 molecular sieve showed a fasterby single six- membered rings and double six-membered rings. coking deactivation rate compared with SAPO-34 [4.6]. At pre-There are two distinct T sites in the framework: one is in thesent, the correlations between the deactivation process (de-double six membered ring and the other is in the singleposited coke species) and the reaction conditions (such as timesix-membered ring The distribution of these two sites is in thend temperature) during the MTO reaction over the SAPO-34ratio of 2:1 [3]. Due to the special structure of SAP0-35, it hascatalyst are relatively clear [15,16]. However, there is no reportbeen employed as the catalyst in methanol conversion andso far on the deactivation process during methanol conversion*Corresponding author. Tel/Fax: +86-41 1-84379335; E-mail: liuzm@dicp.ac.cnDOI: 10.1016/S1872-2067(12)60557-9 I ht://www.sciencedirect.com/science/journal/18722067 I Chin.」. Catal,Vol.34. No. 4, April 2013LI Bing et al. / Chinese lournal of Catalysis 34 (2013) 798 -807799catalyzed by SAPO-35. It is speculated that the activity differ-resonance frequency of 600.13 MHz by using a 4 mm probeences between SAP0-34 and SAPO-35 are due to their differenthead. One pulse program was used and the π/2 pulse lengthstructures. Research on the deactivation during the MTO reac-was 4.4 us. The recycle delay time is 10 s. The spin rate was 12tion over SAPO-35 will help understand the effect of the cagekHz with 32 times sampling frequency. Before the H MAS NMRstructures of the molecular sieves and the deposited coke spic-experiments, all samples underwent vacuum dehydration ates400 oC and < 10-3 Pa for 20 h to remove water and impurities inIn this work, SAPO-35 molecular sieves with different Sithe molecular sieve. The samples were transferred in a nitro-contents were hydrothermally synthesized by using hexameth-gen-flled glove box to the NMR rotator for testing. The quanti-yleneimine as the template. The effect of Si content on thetative processing of the 1H MAS NMR data was as follows. Aphysical and chemical properties of the molecular sieves wascalibration curve for the correlation between peak areas andstudied in detail. In addition, three SAP0-35 molecular sievesmass of sample were first made by using adamantine as thewith different Si contents were chosen as MTO catalysts to in-external standard. Then the 1H MAS NMR spectra of the samplevestigate the effects of the acidity on the reaction. The evolutionwith known mass were acquired under identical acquisitionof the coke species in the reaction was investigated over bothconditions. The peak areas of Bronsted acid sites was calculat-SAPO-35 and SAP0-34 catalysts with similar Si concentrations.ed by Gauss-Lowrance linear ftting to get the density ofThe correlation between the cage size of the molecular sievesBronsted acid sites of the samples from the calibration curve.and deactivation due to the coke species is discussed.2.3. Reaction evaluation2. ExperimentalThe catalytic properties of the molecular sieves were evalu-2.1. Synthesis of molecular sievesated using an atmospheric fixed bed equipment. 1.2 g calcinedsample (40-60 mesh) was packed in the reactor, which wasThe SAP0-35 molecular sieves were synthesized as report-then purged with nitrogen at 520 oC for 30 min. After purging,ed elsewhere [17] Pseudoboehmite powder (w([Al2O3) = 70methanol solution (40 wt%) was injected into the reactor byawt%), sol (w(SiO2) = 31.1 wt%) and phosphoric acid (w(H3PO4}pump. The products were analyzed online using an Agilent= 85 wt%] were the aluminum source, silicon source, and7890A gas chromatograph (CC) with PoraPLOT Q-HT capillaryphosphorus source, respectively. Hexamethyleneimine (HMI,column and FID detector.chemical pure) was used as the template agent The molar ratioof reactants was 0.96P20s:1.0Al20x:nSiO2:1.51HMT:55.47H2O2.4. Collections of coke deposits and analysis of coke species(n= 0.1, 0.2, 0.3, 0.5, 0.8, 1.0). The detailed procedure was asfollows. Into a beaker, deionized water, aluminum source, sili-The deposited coke species were collected by stopping thecon source, phosphorus source, and HMI were sequentiallyfeed after a pre-determined reaction time, unloading the cata-added with vigorous strring to make the initial gel homogene-lysts quickly and then quenching in liquid nitrogen.ous. This was then transferred into a 100 ml stainless steelThe Guisnet method was used for the qualitative analysis ofreactor, which was then sealed and heated at 2000C for 24 h.the coke species [18,19]. In a Teflon bottle, 50 mg of catalystAfter crystallization, the product was collected by centrifugalwas added to 1 ml HF water solution (20 wt%}. After the mix.separation, rinsed with water until the pH level was neutral,and then dried at 120C. The samples with different Si contentschloromethane was added into the solution. 5 min later, NaOHwere denoted as 0.1Si, 0.2Si, 0.3Si, 0.5Si, 0.8Si, and 1.0Si.was added and mixed well. The mixture was then transferredto a separatory funnel and shaken vigorously. The lower layer2.2. Characterizationliquid was collected into a sample vial for GC testing by an Ag.ilent 7890-5975C MSD gas chromatograph-mass spectrometerX-ray powder diffraction (XRD) measurements were per-with a HP-5 capillary column. Each compound was identifiedformed using a PANalytical X'Pert PRO X-ray diffractometerusing the NIST 08 library.with Cu anode, Ka radiation [ = 0.15418 nm), voltage of 40 kVand current of 40 mA. The elemental analysis was carried out3. Results and discussionusing a Phillips Magix 2424X X-ray Fluorescence Spectrometer[XRF). Morphology images were acquired on a Hitachi S-3400N3.1. Synthesis and characterization ofSAP0-35 with different Siscanning electron microscopy (SEM]. X-ray photoelectroncontentsspectroscopy (XPS) was determined employing a ThermoESCALAB 250Xi X-ray photoelectron spectrometer with Al KaThe XRD results of the samples are shown in Fig. 1. All theradiation. The peak of Al 2p at 74.7 eV from Al203 on the sur-SAPO-35 molecular sieves with different Si contents had theface of the samples was used as the internal standard. N2 ad-LEV framework structure, as shown by comparing to thesorption measurement was performed on a Micromeriticsstandard pattern [7,20]. The relative crystallinit, relative yield,ASAP 2010 volumetric adsorption analyzer.and chemical composition of the SAP0-35 samples with differ-1H MAS NMR experiments were performed on a Brukerent Si contents are listed in Table 1. The solid yield of the sam-Avance 1600 solid phase NMR spectrometer with a protonples increased with increasing amount of SiO2 in the initial gels.800LI Bing et al. / Chinese Journal of Catalysis 34 (2013) 798-807Table 1Relative crystallinity and chemical composition of the SAPO-35 sampleswith different Si contents.1.0SiRelativeSSampleerysallinity (%) yield (%) composition incorporation*_ wlltum L0.8Si0.1Si8462 SiolooPas1.820.2Si974 SogAlok49yP-420.5Si0.3Si10036 Sg2Alo485Pa4231.0194Sio. I2Alo 487Po391)195Sio.69Al0.465Pa3670.803500Sio175A0457P0.3680.70*Defined as the molar ratio of [/5(i++P)]oux/[S/(Si+Al+P)]silicon incorporation in the framework of SAPO-5 molecular02(405(sieves and found that highly dispersed and singly substituted20(°)silicon was the most stable, fllowed 5Si and 8Si islands. TheseFig. 1. XRD ptterns of the SAPO-35 samples with diferent Si contents.results suggested that the energy of Si incorporation in theframework increased with increasing content of subtitutedHowever, the crstallinity exhibited first an increase then asilicon, leading to the decreasing ability for silicon incorpora-decrease trend. 0.3Si sample showed the highest crystallinity.tion in the framework. Their conclusion is consistent with ourIn our previous research on the synthesis of SAP0-34, we alsoexperimental results.found that samples with 0.2-0.3 Si contents had the highestThe SEM images of the samples with different Si contentscrystallinity [21]. This suggestedthat the Si content in the ini-are shown in Fig 2. The SAP0-35 grains presented the typicaltial gel not only afected the chemical composition of the sam-rhombohedral morphology, but the change of silicon contentples but also the crytallinity. In addition, the XRF elementalhad a significant effect on the surface roughness of the molecu-analysis results indicated that the Si content of SAP0-35 in-lar sieve crystals. With a low silicon content, the molecularcreased with the increase of Si content in the initial gel. To bet-ter correlate the Si contents of the starting materials and prod-rough and irregular small holes appeared when the siliconucts, we introduce a concept of "silicon incorporation" definedcontent was increased to 0.8. When the silicon content was 1.0,as [Si/Si+A)P]produa/[Si/(Si+AI+P)]el (shown in Table 1). Athe surface of the crystal was wrapped by small particles and itdescending tendency was found for silicon incorporation inshowed a 'core-shell' structure, which presumably resultedSAP0-35 with the increase of Si content in the starting materi-from a secondary crystal growth on the rough surface of theals. The 0.1Si sample exhibited the highest silicon incorporationcrystal or the enrichment of excessive amorphous silica in theof 1.82. The silicon incorporation decreased to less than 1.0gel on the crystal surface. The surface compositions of the 0.5Siwhen the Si content of the initial gel was more than 0.3. Siliconand 1.0Si samples were analyzed by XPS (Table 2). It was foundincorporation was only 0.7 when the Si content of the initial gelthat the crystal surface of these two samples comprised silica,was 1.0. Therefore, we believed that it is the high silicon incor-phosphorus oxide, and alumina. The relative silion content ofporation associated with the samples with low Si contents thatthe surface was higher than that of the bulk phase, and the en-resulted in their relatively low solid yields. Catlow et al. [22].richment of silicon on the high silicon content sample surfaceemployed lattice simulation to calculate the energy changes ofwas higher. It was thus speculated that the shell of the 1.0SiL.OSFFig. 2. SEM images of the SAPO-35 samples with diferent Si contents.LI Bing et al. / Chinese Journal of Cataysis 34 (2013) 798 -807801Table 23.2. Methanol conversion by SAP0-35 molecular sieve withBulk and surface compositions of the SAP0-35 samples with diferent Sidifferent Si contentscontents.Composition0.1Si, 0.3Si, and 0.5Si samples were selected for the study ofSampleBulkSurfaceSirface/Sibukmethanol conversion. The results are shown in Fig. 4 and Table0.5SiSio.122Alo.487Po.391Sio,178Alo.478Po.3351.464. The results of the MTO reaction on SAP0-34 (elemental1.0SiSio.15sA0o457P036Sioz275Alo.405P03201.58composition is Sio.o86Alo.493Po.421) are also listed in Table 4 forcomparison. During an initial period, the conversion of metha-sample grains was formed by a secondary crystal growth onnol on SAP0-35 with different Si contents was maintained atthe rough surface of the crystal. Meanwhile, from the siliconabove 99%. The conversion decreased as the reaction timeenrichment on the SAPO-35 crystal surface revealed by XPS, weincreased and the higher the Si content, the faster the conver-can suppose that the distribution of silicon in the SAP0-35sion of methanol dropped. The selectivity for ethylene of thesecrystals was non-uniform and there was a gradual increase ofmolecular sieves gradually rose as the reaction time increased,the silicon content from the inside outwards. We also found awhile the selectivity for propylene decreased. Compared withsimilar situation in the crsallization of SAP0-34 with di-SAP0-34, SAP0-35 exhibited the characteristic of fast deactiva-ethylamine as a template agent [23]. The main reason would beion.the gradual increase of the gel pH during the synthesis of SAPOThe MTO reaction is a typical acid catalyzed reaction, andmolecular sieves. The increase of pH promoted depolymeriza-the acidity of the molecular sieve plays an important role on thetion of the silicon source in the gel system, thus the ability forlifetime and the selectivity to light olefins. We determined thesilicon incorporation into the molecular sieve framework wasamount of Bronsted acid of the three SAP0-35 samples usingenhanced (the pH value of the gel in our synthesis system be-1H MAS NMR (shown in Table 5]. Generally, the density offore and after the crysallization of 0.5Si sample were 5.56 andBronsted acid sites increased in the samples with the increase8.20, respectively).of the Si contents. Both are almost in a linear relationship. TheFigure 3 and Table 3 show the N2 adsorption-desorptionhigher acid density in the samples with higher Si contentisotherms, specific surface areas, and pore volumes of the sam-caused serious side reactions, such as coke deposition and theples with different Si contents. All three samples displayed ahydrogen transfer reaction, which shortened the life time of thehigh micropore surface area and micropore volume. Moreover,catalyst and generated more methane and propane. Moreover,the outer specific surface area and mesopore volume of thethe deposited coke species gradually formed during the reac-samples increased with the increase of silicon contents, whichtion will decrease the cage size of the molecular sieves, reducewould be due to the rough grain surface of the samples withthe generation and diffusion rates of molecules with largerhigh silicon contents, as revealed by the SEM images.diameters, and consequently increase the selectivity for eth-ylene and decrease the selectivity for propylene.200-105 n180-160 f90- (a4)|75 t买12050 t(1)-0.8Si3)\宣80F45 F? 60-0一0.2Si30 t40F85 F2)、20(b)(2)0.00.2 0.4 0.60.81.0Relative pressure (p/po)75 taFig.3. N2 adsorption-desorption isotherms of the SAP0-35 samples70 twith dfferent Si contents.Table3要65Pore structure parameters of the SAP0-35 samples with diferent Si50。0 20406080100120140160180Surface area (m2/g)Pore volume (cm/g)Time on stream (min)Micropore External Total Micropore Mesopore TotalFig, 4. CH:0H conversion (a) and CzH+C3H6 selectivity (b) in the MTO0.2Si443.526.3469.7 0.220.01 0.21reaction over SAPO-35 and SAPO-34. (1) 0.1Si; (2) 0.3Si; (3) 0.5Si; (4]0.8Si497441.0 538.4 0.230.05 0.28SAPO-34. Reaction condition: 450 °C, 40% CH3OH solution (SAPO-35,45.80.220.070.29WHSV = 2 h-+; SAPO-34, WHSV=4h-).02u Bing et al./Chinese Journal ofCatalysis 34 (2013) 798-807Table 4MTO results over SAPO0-35 and SAP0-34 molecular sieves.TOSSelectivity (wt%)SampleCH4CzH4CzH6_C3HCaHaC4CcsCz+Gx0.1Si2.3737.800.2736.11.0510.4211.8773.96722.6643.851.0232.301.419.249.4376.160.3Si42.7133.940.340.551.3112.468.6974.505541.385 381 489.138.1076.750.5Si29.0434.482.9413.1316.0763.53。009.1675.0542.0932.958.904.0874. 0SAPO-34b0.9633.61.6540.384.6315.6418212131a43.320.6238.801.5411.03Reaction condition: WHSV= 2h-.450 oC, 40% CH:OH solution..The catalyst lfetime, which is defined as the reation duration with >99% CHzOH conversion.b Reaction condition: WHSV =4 h-t, 450 °C, 40% CH:OH solution (elemental composition of SAPO-34: SoAe.04Por4).3. Analysis of the deposited coke species formed in the MTOreaction over SAP0-35 and SAPO-34tives became the major coke species. By comparing the evolu-tion of the organic species in SAPO-35 and SAPO-34, a commonThe fast deactivation of SAP0-35 molecular sieve in thecharacteristic is that the main organic species generated in themethanol conversion reaction indicated that the deposited cokeinitial stage were methylated benzene compounds whichspeies and their evolution with reaction time on SAPO-35 isgradually changed to bulky aromatic hydrocarbons. The fnaldifferent from those on SAP0-34. We selected 0.3Si SAPO-35 tocoke molecules in SAPO-34 were significanty larger than thosestudy the deposited coke species fllowing the reaction timein SAPO-35.and compared these with those generated on SAPO-34. TheThe hydrocarbon pool mechanism is a widely recognizedorganic species extracted from the catalysts were analyzed byMTO reaction mechanism. Much experimental results indicatedGC-MS (Fig.5}.that the multiply methylated benzene (methyi substituents> 3)The organic species initally produced on SAP0-35 (15 min)were mostly toluene, xylene and trimethylbenzene. As the re-a)|creased to 32 min, the methanol conversion de-00 9creased from 98.49% to 76.37%, and organic species with alarger molecular size, such as naphthalene and methyl naph-thalene, started to appear in the deposited coke species, Whenthe reaction time was 83 min, the methanol conversion further会|decreased to 10.88%. Among the deposited coke species, thesignals of naphthalene and methyl naphthalene obviously in-=33.46%creased, and a smallX= 98.57%72 minlene and anthracene dihydride appeared in addition to toluene,xylene and trimethylbenzene. The organic species during the :X= 96% 21 minearly stage (21 min) of methanol conversion over SAPO-34molecular sieve were mainly trimethylbenzene, tetra-0152025303540 45 50methylbenzene, naphthalene, methyl naphthalene and dime-Retention time (min)thyl naphthalene. When the reaction time was 72 min, themethanol conversion was still maintained at a high levelOCoub)|(98.57%}. Coke species such as tetramethylbenzene, naphtha-lene, methylnaphthalene and dimethylnaphthalene were en-hanced, and new signals corresponding to trimethyinaphtha-L X= 10.88% 83 minlene, phenanthrene and pyrene appeared. At the stage of nearlycomplete deactivation of SAP0-34 (447 min, 33.46% methanol00[0conversion), methyl benzene compounds disappeared. A sig-nificant change in the relative amounts of the other organicMIL」__X= 38.60%，49 minspecies occurred, and the polycyclic aromatic hydrocarbonX= 76.37%。32 minTable5X- 98.49%，15 minConcentration of Bronsted acid sites in SAP0-35 molecular sieves cal-101520253035404550culated from 1H MAS NMR.B acid sites (mmol/g)0.54Fig 5. Coke species in SAPO-34 (间and SAPO-35 (b) in the MTO reac-0.97tion. Reaction condition: WHSV = 4 h-, 400 °C, 40% CHzOH solution; X1.20refers to the CH3OH conversion in the reaction.LI Bing et al. / Chinese lournal ofCatalysis 34 (2013) 798-807803Graphical AbstractChin. J. Catal, 2013, 34: 798-807 do: 10.1016/S1872-2067(12)60557-9Synthesis of SAP0-35 molecular sieve and its catalytic properties inSAPO-3Sthe methanol-to-olefns reaction00o CLI Bing, TIAN Peng, LI Jinzhe, CHEN jingrun, YUAN Yangyang su Xiong,FAN Dong, WEI Yingxu, QlYue, LIU Zhongmin*Dalian Instute ofChemical Physics, Chinese Academy of Sciences;University of Chinese Academy of Sciences出c91588SAPO4ICHA)SAP0-35 was hydrothermally synthesized using hexamethyleneimine asthe template. The coke species in the MTO reaction over both SAP0-35 andSAP0-34 were investigated and correlated with their cage size.is a hydrocarbon pool active center, on which the methylationhydrothermally synthesized. The yield of solid sample in-of methanol/diethyl ether occurred to form ethylene and pro-creased with the increase of Si content in the synthesis gel. Thepylene, while the resulting less methylated benzene iscrytallinity of the sample increased first and then decreased asre-methylated to start a new catalytic cycle [15,16,24-27]. Inthe Si content increased. 0.3Si sample exhibited the highestthe present experiments, in the initial stage of reaction the ex-crstallinity. The Si incorporation degree showed a droppingistence of some hydrocarbon pool active species (multiplytrend when the silica content rose. The surface of SAP0-35methylated benzenes) were observed in SAP0-34, whereas thecrystals with a high silica content was rough. In particular, 1.0Sicoke species was dominated by less methylated benzenes inshowed the characteristic of a core-shell structure, which couldSAPO-35. With increased reaction time, at the stage of nearlybe due to secondary crytallization on the rough surface.complete deactivation, the main coke species in SAP0-34 wereSAPO-35 showed fast deactivation than SAPO-34 in the MTObulky aromatic hydrocarbons such as phenanthrene and py-reaction, and a higher silica content will cause faster deactiva-rene (generated from ring condensation by hydrogen transfertion. The main reason is that the higher Bronsted acid concen-from methylbenzene and methylnaphthalene), while the coketration in the sample with higher silica content led to more sidespecies in SAP0-35 were methylbenzene, naphthalene andreactions such as coke deposition and hydrogen transfer. Bymethylnaphthalene. The difference in the coke species ofcomparing and analyzing the deposited coke species formed inSAP0-35 and SAP0-34 is probably related to their structures.SAP0-35 and SAP0-34 during the MTO reaction, we concludedThe CHA cage of SAPO-34 (0.67 nm x 0.67 nm x 1.0 nm) isthat the smaller cage of SAP0-35 limited the generation of hy-larger than the LEV cage of SAP0-35 (0.63 nm x 0.63 nm x 0.73drocarbon pool active species (polymethylbenzens) and mac-nm). The smaller cage of SAP0-35 restricted the formation ofromolecular coke species (polyaromatic hydrocarbons). More-the hydrocarbon pool active species (multiply methylated ben-over, the smaller cage of SAPO-35 had a lower capacity to ac-zene) and deactivated coke species (macromolecular fused-ringcommodate deposited coke species. These two reasons result-aromatic hydrocarbon). Also, it has a smaller accommodationed in the faster deactivation of SAP0-35.capacity for deactivated coke species, which thus resulted inthe rapid deactivation. The lack of hydrocarbon pool activeReferencesspecies also led to the relatively low ethylene and propyleneselectivity with SAP0-35. Moreover, it is worth noting that[1] Lok B M, Messina C A, Patton R L, Gajek R T, Cannan T R, Flanigenduring the reaction, the organic compounds deposited inEM. uS Patent 4 440 871.1984SAPO-35 always contained a small amount of saturated cyclo-[2]FuY,WangLF,TanYx,JiHB.ChemReactEngTechnol(付晔,王alkane adamantane compounds. In our recent research on the乐夫,谭宇新，纪红兵化学反应工程与工艺)，2000, 16(1): 55coke species in SAPO-34 in the MTO reaction [28], we reported[3] ht://ww.iza-structure.org/databases/that adamantane compounds are the coke species in the low[4] Zhu Zh D, Hartmann M, Kevan L. Chem Mater, 2000, 12: 2781temperature reaction (≤350 oC), which were gradually trans-[5] Jeon H Y, Shin C H, Jung H I, Hong S B. Appl CatalA, 2006, 305: 70formed into naphthalene and substituted naphthalene species[6] Djieugoue M A, Prakash A M, Kevan L / Phys Chem B, 2000, 104:at elevated temperatures. In this experiment, the reaction6452temperature was 400 C, so the existence of adamantane com-[7] Venkatathri N, Yoo I w. Appl CatalA, 2008, 340: 265pounds in SAP0-35 once again demonstrated that the different[8] Cheung 0, Liu Q L, Bacsik z, Hedin N. Microporous Mesoporouscage size of the molecular sieve has a large influence on theMater, 2012, 156: 90generation and stability of the coke species.9] FalconerIL, Li Sh G, Noble R D. us Patent 20050204916A1.2005[10] Liu GY, Tian P, Liu zh M. Progr Chem(刘广宇田鹏，刘中民化学4. Conclusions进展), 2010, 22: 1531[11]ZhangICh,ZhangHB,YangXy,HuangZh,CaoWL.JNatGasSAP0-35 molecular sieves with different Si contents wereChem, 2011, 20: 266804L1 Bing et al / Chinese Jourmal ofCatalysis 34 (2013) 798- 807[12] Liu G Y, Tian P, Xia Q H, Liu Zh M.] Nat Gas Chem, 2012,21: 431[20] Prakash A M, Hartmann M, Kevan L Chem Mater, 1998, 10: 932[13) Nawaz s, Kolboe s, Sticker M. stud SurfSci Catal, 1994, 81: 393[21] Zhang D zh. [PhD Diseraion]. Dalian: Dalian Institute of Chemi-[14] Chen I Sh, Wright P A, Thomas I M, Natarajan s, Marchese L, :cal Physics, CAS (张大治。[博士学位论文]大连:中国科学院大连Bradley S M, Sankar G, Catlow C R A. Gai-Boyes P LJ Phys Chem,化学物理研究所),20071994, 98 10216[22 Sastre G, Lewis D w, CatlowC RA/ Phys Chem, 1996, 100: 6722[15] Yuan cY, Wei Yx, lij Zh, Xu Sh T, Chen I R, Zhou Y, Wang QY,Xu[23] LiuGY, Tian P, ZhangY,Li] zh, Xu L, Meng Sh H, Liu Zh M.L, Liu zh M. ChinJ Catol (袁翠峪魏迎旭，李金哲,徐舒涛,陈景Microporous Mesoporous Mater, 2008, 114: 416润，周游,王全义，许磊,刘中民催化学报), 2012, 33: 367[24] Wang I D, YuX B, Liu Y, Yang Y R Progr Chem (王靖岱，虞贤波,刘[16] Hereigers BP C, Bleken F, Nilsen M H, Svelle S, Lllrud K P,烨,阳永荣化学进展), 2009,21: 1757Bjprgen M, Weckhuysen B M, Olsbye U.JCatal, 2009, 264: 77[25] Arstad B, Kolboe S.JAm Chem Soc, 2001, 123: 8137[17] ZhangF M, Kuang X H, uiL sh, ShuXT, Wang W D, QinF M (张凤[26] Arstad B, Kolboe s. Catal Lett, 2001,71: 209美，匡晓辉,李黎声,舒兴田，王卫东，秦凤明), CN Patent[27] Lesthaeghe D, De Sterck B, Van Speybroeck V, Marin G B,101 397 143. 2009Waroquier M. Angew Chem, Int Ed, 2007, 46: 1311[18] Guisnet M.J Mol Catal A, 2002, 182-183: 367[28]WeiYX,LiJZh,YuanCY,XuShT,ZhouY,ChenIR,WangQY,[19] Guisnet M, Costa L, Ribeiro F R.J Mol Catal A, 2009, 305: 69Zhang Q Liu Zh M. Chem Commun, 2012, 48: 3082SAPO- 35分子筛的合成及其甲醇制烯烃反应性能李冰，田鹏"，李金哲"，陈景润“，袁扬扬叨，苏雄叻，樊栋”,魏迎旭"，齐越"， 刘中民”“中国科学院大连化学物理研究所洁净能源国家实验室(筹),甲醇制烯烃国家工程实验室,辽宁大连116023中国科学院大学,北京100909摘要:以六亚甲基亚胺为模板剂,采用水热法合成了不同硅含量的磷酸硅铝分子筛SAPO-35,并利用X射线衍射、X射线荧光光谱、扫描电镜、固体核磁、X射线光电子能谱和N:吸附脱附等方法对样品进行了表征.研究了不同硅含量的SAPO-35分子筛在甲醇转化制烯烃反应中的催化行为,同时对比分析了具有相近硅含量的SAPO-35和SAPO-34分子筛在甲醇转化反应过程中积炭物种随反应时间的演变特征，尝试将分子筛结构和其积炭失活行为进行了关联关键词: SAPO0-35;甲醇制烯烃; SAPO-34;积炭物种;失活收稿日期: 2012-12-30.接受日期: 2013-01-30.出版日期: 2013-04-20.*通讯联系人.电话/传真: (0411)84379335;电子信箱: liuzm@dicp.ac.cn本文的英文电子版由Elsevier出版社在ScienceDirect上出(tp://w siedre.com/scicnce/joual18722067.0筛，如SAPO-17, SAPO-18和SAPO-35也被尝试用作MTO1.前言反应的催化剂713.4.研究表明,与SAPO-34相比，磷酸硅铝系列分子筛(SAPO-n)由美国联合碳化物SAPO-35分子筛在MTO反应中表现出较快的积炭失公司(UCC)于1984年合成并报道"由于SAPO分子筛具活14.6].目前, SAPO-34分子筛上甲醇转化失活过程(积炭有温和可调的酸性和不同的孔道结构，在多个化工行业物种)随反应时间和反应温度等条件变化的规律和特点展现出广阔的应用前景(21. SAPO-35作为其中的一员,晶已经相对清楚I5.16,但是到目前为止,还未见到SAPO0-35体结构为插晶菱沸石型(LEV),孔径0.36 nm x 0.48 nm,分子筛上甲醇转化失活过程的研究.两者催化性能的差属小孔分子筛. SAPO-35的骨架结构可看作是LEV笼通别推测与自身结构的差异有关.对SAPO-35积炭失活过过单六元环和双六元环连接而成,它含有两种不同的T程的研究分析将有助于深入理解分子筛笼结构变化对原子位置,分别在双六元环和单六元环中,两者的分布比反应过程和积炭物种的影响.例为2:13).依据SAPO-35分子筛独特的结构特点研究者本文以六亚甲基亚胺为模板剂，采用水热法合成出尝试将其用作甲醇转化反应和氨甲基化反应的催化不同硅含量的SAPO-35分子筛,详细研究了硅含量变化剂14-7以及用于CO/CH吸附分离8.1.对最终产品物理化学性质的影响，并选取了3个具有不甲醇转化制烯烃(MTO)是以煤或天然气为原料制同硅含量的SAPO-35分子筛作为MTO反应的催化剂,考取低碳烯烃的关键过程. ZSM-5分子筛及具有八元环孔察了酸性质变化对反应结果的影响.另外,还对比研究道和cha笼的SAPO-34分子筛在MTO反应中表现出优异了具有相近硅含量的SAPO-35和SAPO-34在MTO反应的催化性能710121.1此外，其它八元环孔道的小孔分子过程中积炭物种随反应时间的变化，尝试将分子筛结构LI Bing et al. / Chinese Journal ofCatalysis 34 (2013) 798- 807805和其积炭失活行为进行关联.反应温度,然后停止通载气,并采用微量泵泵进甲醇溶液(w= 40%).产物采用气相色谱仪(Agilent7890A型)进行2.实验部分在线分析, PoraPLOT Q_HT毛细管色谱柱, FID检测器.2.1.分子筛的合成2.4.积炭样 品的收集及积炭物种分析SAPO-35分子筛的合成参照文献[17].以拟薄水铝反应过程同2.3节,在反应--定时间后停止进料,快石粉(w(Al2O3) = 70%，质量分数,下同)、硅溶胶(w(SiO2)=速卸出催化剂至液氮中急冷后保存，得到相应反应时间31.1%)和磷酸(w(H;PO4) = 85%)分别为铝源、硅源和磷的积炭样品.源,以六亚甲基亚胺(HMI,化学纯)为模板剂.各组分摩积炭物种的定性分析采用Guisnet法18.191.分别称取尔配比为0.96P2Os:1.0Al203;:nSiO2:1.51HMT:55.47H2O 50 mg催化剂装入聚四氟乙烯瓶中,加入1 ml的HF水溶(n= 0.1,0.2,0.3,0.5,0.8, 1.0).具体配料过程如下:依次向液(20%),摇匀后静置1 h,样品溶解后,再加入0.5 ml二氯烧杯中加入去离子水、铝源、磷源、硅源和有机胺,搅甲浣,静置5min后,加入NaOH溶液并摇匀,将混合物转移拌均匀后，将初始凝胶移至100 ml不锈钢合成釜中,密封到分液漏斗中振荡静置,将下层萃取液取出滴入微量进后加热至200 °C恒温晶化24 h.晶化完成后将产物离心，样 瓶中备分析用，采用安捷伦色质谱(Agilent固体样品用水洗至中性后120。C烘干备用.不同Si含量7890-5975C MSD)进行分析, HP-5毛细管色谱柱.积炭物的样品记为nSi (n= 0.1,0.2,0.3,05,0.8, 1.0).种定性采用Nist08数据库.2.2.分子筛的表征X射线粉末衍射(XRD)物相分析在PANalytical3.结果与讨论X'Pert PRO型X射线衍射仪上进行，Cu靶K&辐射源(2=3.1.不同硅含量 SAPO-35分子筛的合成与表征结果0.15418 nm),电压40 kV,电流40 mA.采用Philips公司的合成SAPO-35样品的XRD谱见图1.由图可知,所制Magix 2424X型射线荧光(XRF)光谱仪对样品进行元素系列不同硅含量的样品均为具有LEV结构的SAPO-35分析.样品形貌在Hitachi S-3400N型扫描电子显微镜分子筛[.20.表1列出了它们的相对结晶度、相对收率和(SEM)上观察.X射线光电子能谱(XPS)采用Thermo元素组成.可以看出,随着初始凝胶中SiO2用量的增加,ESCALAB 250Xi型X射线光电子能谱仪进行测定(以单所得样品的固体收率逐渐上升;而样品的相对结晶度则色化AlKa为激发源),以样品表面Al2O3的A12p=74.7先升高后降低，其中以0.3Si样品的最高，我们在eV为内标来校正样品表面的荷电.N2吸附_脱附实验在SAPO-34的合成研究中也发现,初始凝胶中硅含量在0.2美国麦克ASAP2010型物理吸附仪上进行.到0.3时所得样品的相对结晶度最高121.这说明凝胶中'H固体核磁共振(H MAS NMR)谱在Bruker Avance的硅 含量不仅影响产品的组成,同时对分子筛的结晶度II-600型固体核磁共振谱仪上测定，使用4 mm探头. 'H也有影响.另外，XRF结果显示，SAPO-35中硅含量随着的共振频率为600.13 MHz,采用单脉冲(one pulse)程序,初始凝胶中硅含量的增加而上升.为了进-步关联投料π/2脉宽为4.4 us,弛豫延迟为10s,转速为12 kHz,采样次硅含量与所得样品中硅含量的关系,这里提出硅进入率数为32次.实验前,所有样品在400 °C,低于10-3 Pa真空的概念,将其定义为[Si/(Si+Al+P)]=*/[Si/(Si+Al+P)]初始凝权脱水20 h以上，以脱除吸附在分子筛中的水和杂质.样品(见表1).可以看出，随着投料硅含量的增加，所得在N2手套箱中转移到核磁转子中待测.'HMASNMR谱SAPO-35样品中的硅进入率逐渐下降.其中0.1Si样品的的定量分析方法如下:以金刚烷为外标物，首先获得标硅进入率最高为1.82;当初始凝胶中硅含量大于0.3时,样峰面积与质量关系的标准曲线.在相同的采谱条件下硅进入率开始小于1;至1.0时仅为0.7.因此,可以认为正获得已知质量样品的'HMASNMR谱，根据是由于低硅样品中高的硅进入率导致了其较低的固体Gauss-Lowrance线型拟合获得B酸峰面积，再根据标准收率.文献[22]运用晶格模拟技术Lttce simulation曲线获得样品的B酸密度.techniques)计算了SAPO-5分子筛中硅进入骨架能量的2.3.催化性能评价变化,发现分散度高的单取代硅是最稳定的，其次是5Si采用常压固定床装置评价分子筛样品上甲醇转化和8Si岛,即随着硅取代量的增加,硅进入骨架所需能量反应性能.将1.2 g焙烧后的催化剂样品(40- 60目)装填在也相应增大,从而导致硅进入骨架的能力降低.这与本反应器中,于N2气氛下升温至520 °C吹扫30 min后降至文结果- -致.u Bing etal. /Chinese /ourmal ofCatays 34 (2013)798 807807通过氢转移产生的环缩合反应), SAPO-35上则主要是4.结论甲基苯、蔡和甲基恭.这可能与它们自身的结构有关.SAPO0-34分 子筛中的CHA笼(0.67 nm x 0.67 nmx1.0利用水热法合成了不同硅含量的SAPO-35分子筛，m)要大于SAPO-35中的LEV笼(0.63 nmx 0.63 nmx发现随着合成凝胶中硅含量的增加,样品的固体收率逐0.73 nm). SAPO-35中较小的笼体积限制了烃池活性物渐上升;样品的结晶度则先升高后降低,其中0.3Si样品种多甲基苯和大分子积炭失活物种稠环芳烃的生成，同具有最高的相对结晶度;硅进入骨架的能力随硅含量的时它对积炭失活物种也具有较弱的容纳能力,因而失活增加而下降.高硅SAPO-35的晶体表面比较粗糙,尤其是较快，烃池活性物种的缺乏也使得乙烯、丙烯选择性相1.0Si样品的晶体形貌呈核壳结构，推测为粗糙晶体表面对较低.另外,值得注意的是,反应过程中SAPO-35催化的二次晶体生长所致.不同硅含量的SAPO-35样品在剂上沉积的有机物中始终含有少量饱和环烷烃结构的MTO反应中都呈现较快速失活的特征,且硅含量越高，金刚烷类化合物.我们近期对SAP0-34在MTO反应的积失活速度越快.这主要是由于高硅样品中较高的B酸中炭物种研究中，首次发现报道了金刚烷类化合物是低温心密度导致了严重的积炭和氢转移反应等副反应.对比反应(≤350 °C)的积炭物种,升高温度后其逐渐转变为分析SAP0-35与SAPO-34在MTO反应过程所产生的积蔡和取代萘类物种2号本文实验的反应温度为400。C,金炭物种,发现SAPO-35中较小的笼体积限制了烃池活性刚烷类化合物在SAPO-35中的稳定存在再次说明分子物种多甲基苯和大分子积炭失活物种多苯环化合物的筛笼尺寸的变化对积炭物种的生成和稳定具有重要的生成，同时其对积炭失活物种也具有较弱的容纳能力,从影响.而导致了较快速的失活.