Study on Olefins Yield from Methanol Conversion over Different Catalysts

Scientific ResearchStudy on Olefins Yield from MethanolConversion over Different CatalystsMunib Shahda; Yan Dengchao; Wang Zhihe; Wen Huixin(East China University of Science & Technology, Shanghai 200237)Abstract: Conversion of Methanol to Olefins (MTO) under different reaction conditions was ex-perimentally investigated over different catalysts, and comparison was made between the SAPO-34and GOR-MLC catalysts. Optimization of reaction conditions has been explored. Conversion ofmethanol to olefins over these catalysts under different reaction temperatures was experimentallystudied. In a fixed bed micro-reactor, the influence of temperature was found to be one of the majorfactors. For both catalysts the olefins yield was increased significantly when water was added tothe methanol feed. A temperature range of 460- 480 C appeared to be the optimum range suitablefor methanol conversion with appropriate catalyst activity and C,- C, olefins yield. Some otherhydrocarbons appeared during the MTO reaction in the presence of the SAPO 34 catalyst, while alot of dimethylether was formed when the GOR-MLC catalyst was used. In the course of the MTOreaction, the GOR-MLC catalyst was found to have a faster catalyst deactivation rate compared tothe SAPO-34 catalyst.Key words: MTO; methanol; olefins; SAP0-341 Introductioncrease the catalytic selectivity to olefins in the course of metha-nol conversion.The methanol to olefins (MTO) process provides an addi-tional route for producing ethylene and propylene for theThis paper focuses on the technology of MTO reaction withchemical industry. Low-carbon olefins such as ethene and different catalysts; however, the selectivity for ethene andpropene are the basic organic feedstocks used in modern propene over the GOR catalyst is not as high as that with thechemical industry.SAPO- 34 catalyst. Oligomerization and coke deposition arethe major factors affecting the selectivity for ethene andChina has imported a hefty amount of ethene to satisfy the propene. Addition of water to methanol can increase the C,-need of its economic development. At present, most light C yield. Nonetheless, in all our experiments, no aromatic hy-olefins are obtained from steam cracking of naphtha, and thisdrocarbons were detected.is considered to be a very energy consuming and costlyprocess. Natural gas is an alternative feedstock for the pro-2 Experimentalduction of light olefins. The catalytic conversion of methanolto olefins (MTO) is an interesting and promising way of con-2.1 Starting materials and apparatusverting natural gas and coal to chemicals via the intermediatemethanol. Coke deposition'"I is known to be the major cause2.1.1 Starting Materials and selected catalystsof deactivation during the MTO reaction over the SAPO-34catalyst. However, both catalytic selectivity and activity are The中国煤化工methanol, ethylene,influenced by the coke deposition on catalyst. Due to the propyn. The catalysts SAPO-HCNMHGPetroleum Processing,increasing demand for light olefins in the petrochemical 34 (prev ww wy owii l Jiindustry, several routes have been followed in order to in-SINOPEC) and GOR-MLC were used in the experiments.49China Petroleum Processing and Petrochemical TechnologyNo.1, March 20062.1.2 Apparatusanalyzed on-line by GC; the current in TCD was 170 mA, theinjector temperature was 200C, and the column temperature(1) Reactor: The whole series of experiments were conductedwas 90'C. Hydrogen was used as a carrier gas, and nitrogenin a fixed bed micro reactor provided with a stainless steelU- was used to purge the residual products from the tube. Thetube of 6mm(id.)x140mm(in length), loaded with 0.5g of GDX101 chromatograph was used for the separation ofC-catalyst; the reactor was operated under atmospheric C hydrocarbons, MeOH, DME, CO, CO, H,O, and olefins.pressure. At the end of each run, prior to replacement of the Figure 1 shows the sketch of the experimental unit.catalyst the U-tube was purged with nitrogen to eliminate theresidual products inside the tube.3 Results and Discussion(2) Gas chromatograph: A gas chromatograph equipped with 3.1 Study on the olefins yield from methanola GDX101 (porapak QS) packed column, 2m in length and 3 conversion over the GOR-MLC catalystmm in inside diameter, was used with a Thermal ConductivityDetector (TCD) to analyze the reaction products.3.1.1 Efectof the reaction temperature on product yieldoverthe GOR-MLC catalyst with pure methanol feed2.2 Experimental procedureThe GOR-MLC catalyst showed a combined molar selectivityMethanol conversion was carried out in a flow reactor at to light olefins of more than 55%, and the low levels of low-varied temperature under atmospheric pressure in the pres- molecular paraffins could not be attributed to its shapeence of nitrogen. The partial pressure of N2 was equal to 0.2 selectivity.MPa, and the space velocity of methanol stream was4 h'.The reaction products from conversion of methanol were In these tests the formation of methane, CO, and CO (whichO1Nitrogen35Feed 707| 11EvacuationSampling analysisFigure 1 Sketch of the experimental unit1- Pressure reducing valve; 2- Valve; 3- Rotameter中国煤化工6- Reactor; 7--Measuring pipe; 8- Constant flow pur:MHCNM H Gi;10- Cooler; 11- Coling rap; 12- -Pessre maintaining valve50Scientific Researchis not defined in this paper) were partly atibuted to the reac- The experiments revealed that the catalyst deactivation wastion of methanol on the micro-reactor wall.very fast with remarkable coke formation on the surface ofcatalyst.Figure 2 shows the effects of reaction temperature on theproduct yield. The yield of propylene increased with rising 3.1.2 Efet of reactin temperature on product yield fromreaction temperature, while the yield of ethylene increased methanol/water mixed fed over the GOR-LMC catalystwith rising reaction temperature till approximately 460 C thenstarted to decrease, because the propylene was more reac- The test results demonstrated an increase in yield of lighttive than ethylene and coke formation on the catalyst was olefins products when a diluent was employed. Figure 3faster at higher temperaturesl?. A temperature range of 450- shows how the temperature could influence the product yield475C appeared tobe the optimum temperature range in terms with a very low ethyene yield appearing during this reaction.of oefins yield as shown in Figure 2. The dimethylether (DME) At the same time the proplene yield increased as the reac-yield during this reaction appeared to be very significant, as tion temperature increased.evidenced by the almost complete conversion of methanolover this kind of catalyst to dimethylether without being di-rectly converted to light olefins. The intermediate during de-50hydration of methanol to dimethylether ether over the solidacid catalyst was a protonated surface methoxyl radical2),and the remainder of the reaction products as protonatedsurface methoxyl radicals without hydrogen transfer could20be subjected to subsequent conversion to light olefins.As shown in Figure 2 the DME yield decreased with increas-375 400 425 450 475 500 525ing reaction temperature, because coke formation on the cata-Temperature, Clyst was faster at higher temperatures, and the ratio of DMEFigure 3 Effect of reaction temperature onconverted to light olefins increased too.product yield with pure methanolfeed(Reaction time=30 min, Space velocity=4h', Feed: 50%water with70.0050% methanol)60.00◆一Ethylene;■- -Propylene;▲-DME;50.00●- Ethylene+propylene日40.00 .For this case it was found that ethene comprised a small30.00 .proportion, so the reactor effluent consisted essentially of20.00propene and DME.10.00 IIn the early stages of reaction, when the catalyst was very375 400 42: ; 450 475 500 525active, complete conversion of the methanol took place at thetop of the catalyst bedBI. In the remaining part of the catalystFigure2 Effect of reaction temperature onbed there were only hydrocarbon interconversion reactions,leading to isomeric scrambling. It was further reasonably as-(Reaction time =30 min, Space velocity=4hr')sumed中国煤化工vity decine caused by◆- Ethylene; ■- -Propylene; ▲- -DME;coke foYCNMHG; decreasing number ofcatalytr: sIucs wule ue 1uMUIIIsI acuvity and intrinsic selec-51China Petroleum Processing and Petrochemical TechnologyNo.1, March 2006tivity of the remaining active sites were unchanged.1008In case of using the water/methanol mixed feed, the productDME appeared in the products mix, and a similar result was60observed in the case of using pure methanol feed.40The yield of DME was dependent on the crystal size of cata-20lyst and the amount of coke formation4. Because of the fast0catalyst deactivation, DME formation was identified.360 380 400420 440 460 480 500Temperature, CThe yield of dimethylether decreased with an increasing re-Figure 4 Effect of reaction temperature onaction temperature for both pure methanol feed and the wa-methanol conversion and product yield withter/methanol mixed feed while the yield of olefins increased.pure methanol fed.This was due to a higher rate of DME conversion to olefins at(Time on stream=30 min, Space velocity=4h')higher temperatures.■- CH;▲-C0;★- Ethylene; <- CH;● - Propylene;★- CH;O- _CH,OH;◆- Ethylene+propylene3.2 Study on the olefins yield from methanolconversion with the SAPO 34 catalystmethanol yield increasing. At higher temperatures the deacti-vation of catalyst was faster because of a faster coke forma-Two series of experiments were performed to investigate the tion on the catalyst.impact of reaction temperature. One series adopted the puremethanol feed; the other scries was performed with addition The presence of light paraffins was atributed to competitiveof water as the diluent. In all the experiments, the weight chemisorption over the strongest acid sites 151.hourly space velocity (WHSV) was fixed at 4h' and the reac-tion time was set at 30 min.At low temperature a lot of methanol and lower hydrocar-bons appeared owing to the catalyst deactivation caused by3.2.1 Effect of reaction temperature on product yield over the oligomers inside the catalyst channels 161. A temperaturethe SAPO-34 catalyst with pure methanol feedrange of 460-- 480C appeared to be an optimum range interms of the catalytic selectivity of hydrocarbons.Figure 4 shows the yield changes with reaction temperatureover the SAP0-34 catalyst with pure methanol fed. The pro- 3.2.2 Eftet of reaction temperature on product yield overpene yield was significantly increased but the yield of ethenethe SAPO-34 catalyst with methanol/water mixed feedcould increase with reaction temperature till approximately470 C then it started to decline, and the propylene yield was It has be proven that adding water to methanol would in-less than the ethylene yield. The propylene molecules could crease the olefins yield and decrease the rate of catalystundergo oligomerization more easily to formn bigger oligomers, dcactivation, so the study on reaction conditions for metha-which would be trapped in the cages of the SAPO-34 zeolite. nol conversion with mixed water/methanol feed was veryAs the reaction temperature continued to increase, these oligo- important. A feed composed of 50 mol% methanol and 50mers were cracked to form ethylene, and, as shown in Figure mol% water had been studied. Figure 5 shows the effect of4, the oligomers cracking might occur at around 400C. Dur- temperature on the products yield while the reaction time wasing this reaction a string of other hydrocarbons like ethane specif中国煤化工and propane were also identified. The propane yield increasedCNMHGwith reaction temperature then started to decline, while after The prouucls yien clauigu Wiul leaction temperature. Ad-480C the hydrocarbons yield started to decrease with the dition of water could favor the formation of olefins, and could52Scientific Research1003.2.3 Methanol conversion over the SAPO- 34 catalyst80 tFigures 6 and 7 show the changes of methanol conversion asa function of time on stream with different temperature using50pure methanol feed and water/methanol mixed feed,40respectively.20At the beginning of reaction the methanol conversion was100% till after 40- -50 minutes when the methanol conversion360 380 400 420 440 460 480 500decreased. The decreasing trend for methanol conversionTemperature, Cwas seen apparently with different temperature and was at-Figure 5 Effect of reaction temperature ontributable to the coke formation, which caused the catalystproduct yield with methanolwater mixed feeddeactivation. The temperature of 460C appeared to be an(Reaction time =30 min, Space velocity=4h"', Feed: 50mol%wateroptimum one in terms of methanol conversion.with 50mol% Methanol.)00■-Ethylene; ▲- -Propylenc;●- Ethylenc +propylene;★- -Methanol;O- C,H,; O - Ethane50 Ialso alleviate the catalyst deactivation. The increase in ole-40 tfins yield and reduction in catalyst deactivation were caused20 tby the reduction of olefin conversion to higher molecularweight coke type products. With the presence of water in the06(8120feed, some acidic active sites on surface were occupied byTime on stream, minpolar water molecules and thus were unavailable to olefins. ItFigure 6 Effect of reaction temperature oncould prevent olefins from undergoing oligomerization andmethanol conversioncoking on the active sites, and might keep the channels in-(Space velocity=4h-' with pure methanol feed)side the catalyst unblocked. Therefore, more olefins could◆- -360C;■- 480C;▲- -380C;白- -400C;★- -420C;exit from the reactor as products and less coke was formed on●一440C; 0- _460Cthe catalyst.。10p本The propylene yield increased with the rising reaction tem-。80perature before it started to decrease, and apparently the yieldg 60of propylene was more than the ethylene yield. At lower andhigher temperatures, methanol appeared in the productmixture, indicating to the incomplete conversion of methanol,because of the faster deactivation at temperatures lower orhigher than the optimal range. At a higher temperature the80catalyst was deactivated faster, because the coke formationwas faster at higher temperatures. At a lower temperature theFigure 7 Effect of reaction temperature oncatalyst deactivation occurred due to the formation of oligo-prstonmers inside the catalyst channels along with some hydrocar-中国煤化工: with 50% methanol.)bons such as ethane and propane.MYHCNMHG_ 40C:●-420C .53China Petroleum Processing and Petrochemical TechnologyNo.1, March 2006proved to be a good choice for increasing the olefin yield.Methanol conversion rate was used in the present study toevaluate the catalyic properties of the SAPO-34 zeolite with 4 Conclusionsdifferent structures and acidities. It is well known that metha-nol conversion can proceed on molecular sieves either The conversion of methanol to olefins on two different cata-through simple dehydration reaction to yield dimethyl ether lysts showed different light olefins yields. The GOR-MLC(DME) and water as the only products, or through a deep catalyst had shown a faster catalyst deactivation than thedehydration process to produce hydrocarbons7.SAPO-34 zeolite. The GOR-MLC catalyst had shown a yieldof propylene higher than the ethylene yield.In the case when the methanol/water feed was used, the metha-nol conversion increased with time on stream as compared to The SAPO- 34 catalyst had shown a better light olefin yield inthe case using methanol as the feed. This could be inter- particular when water was used as a diluent. The olefins se-preted by two factors. Firstly, water could reduce the metha- lectivity was more than 90% with the SAPO 34 catalyst. With-nol partial pressure, and the reduced methanol partial pres- out addition of water, the catalyst was deactivated rapidly.sure could favor a higher methanol conversion. Secondly, The decrease in catalyst deactivation rate and increase inbecause of the presence of water, the catalyst deactivation olefins yield were attributed to the reduction of olefin con-was greatly suppressed [8-101 and the partially deactivated version to higher-molecular-weight coke products.catalyst could favor a higher methanol conversion.The GOR-MLC catalyst had not demonstrated a high selec-3.3 Comparison of products yield obtained with tivity to olefns. However, this catalyst was cheaper and moredifferent catalystsreadily available on the market than the SAPO -34 zeolite. Thepresence of water in the feed showed a favorable effect onAccording to the study on the MTO reaction, the olefins the products distribution, besides increasing the olefins yield.yield over the catalyst SAPO-34 was higher than that overthe catalyst GOR-MLC; and the tests results revealed that In all our experiments, no aromatic hydrocarbons were de-the catalyst SAPO-34 was more active than the GOR-MLC tected in the reactor effluent, which was atributable to thecatalyst.shape selectivity of the SAPO- 34 and GOR-MLC catalysts.Upon using the GOR-MLC catalyst the DME was the mainAcknowledgmentsproduct, which was known as an intermediate product in theMTO reaction, and a lot of methanol was not converted to The authors acknowledge the support of the Petroleum Pro-hydrocarbons over this catalyst.cessing Research Center of the East China University of Sci-ence and Tchnology.The deactivation of the GOR-MLC catalyst was much fasterthan the SAPO- 34 zeolite, because the oligomers were formedReferencesinside the catalyst pores, and these oligomers could not leavethe cages, leading to pore blokagel"l.[1]Andres T Aguayo,Ana Trrio, Javier Bilbao. Chem TechnolBiotechnol, 1999,74:315- -321During the MTO reaction, the GOR-MLC catalyst was found [2] Wu X, Anthony RG. Appl Cat A: Gen, 2001,218:241- -250to have a faster catalyst deactivation rate compared to the [3] Dahl IM, Kolboe s. J Catal, 1994,149: 458- -464SAPO- 34 catalyst. The methanol conversion over the SAPO- [4] Chd中国煤化工orous and Mesoporous34 zeolite using the methanol/water mixed feed could be car-MateridMHCNMH Gried out completely for a longer time than the case using pure [5] Stephen WiuIson, raul Barger. MIcroporous Materials, 1999,methanol feed. The addition of water as a diluent had been 29:117- -12654Scientific Research[6] Wu X, Anthony R G. Appl Cat, A: Gen, 2004, 260:63- 69 Olefins and Dimethyl Ether. Stud Surf Sci Cata, 2001, 133:[7] Venuto P B. Microporous Materials, 1994, 2 :297211- -218.[8] Wu X,Anthony RG. ApplCat,A: Gen, 2001,218:241- -250 [10] Van MJ N . Fletecher JCQ, O'ConnorCT Appl Cat, A:[9] Abraha M G, Wu X, Anthony R G. Effects of Particle Size Gen, 1996, 138: 135- -145.and Modified SAPO 34 on Conversion of Methanol to Light [1] CampeloJ M. Appl Cat, A: Gen, 2000, 192:85- 96The First in China 10-kt/a-class DME UnitCame on Stream in AnhuiThe 10 kt/a unit for manufacture of dimethyl ether via gas this unit totals 9 million plus RMB with fund savings amount-phase dehydration of methanol at Anhui Xin' ao Gas ing to more than 30% as compared to the investment in anCompany, which adopts the technology licensed by the imported similar unit, and the investment is expected to beSouthwest Chemical Research Institute and Sichuan Tianyi paid off within two years. The report on the engineering com-Science & Technology Co., Ltd. has been fficially put on missioning checkup and acceptance of this unit has revealedstream in January 2006 and has passed the 72-hr-long accep- that the DME purity can reach 99.99% and the steam con-tance test with all techno-economic indicators reaching or sumption on one ton of DME produced is reduced by 0.5-exceeding the design targets. The operation of this unit is 1.5 ton as compared to the similar overseas units.easy without waste pollutants discharge. The investment inProject for Recycling of Wastewater fromSaponification of Propylene Oxide Put onStream at Shandong Dongda Co.The project for recycling of wastewater from propylene oxide cal Co., Ltd. and Guangzhou Zhonghuan Wandai Environ-production unit has been put on stream at Shandong Dongda mental Engineering Co., Ltd. and the intellectual propertyChemical Co., Ltd. Through the multiple effee evaporation rights on this process have been filed in the relevant authori-of wastewater from saponification of propylene oxide the cal- ties of China and other 13 developed countries. Currently thecium chloride in wastewater is concentrated to 65%- -70%， construction of the phase I project, in which the investmentwhich is then processed and sold as a product. The con- totals 30 million RMB, has been completed, and this projectdensed water obtained during evaporation is returned back can produce 30 kUa of calcium chloride and recover 660 ktato the production unit for reutilization. This project is jointly of condensed water.invested, designed and constructed by the Dongda Chemi-中国煤化工MYHCNMHG55