Simultaneous operation of dibenzothiophene hydrodesulfurization and methanol reforming reactions ove

第40卷第6期燃料化学学报Vol. 40 No.62012年6月Journal of Fuel Chemistry and TechnologyJun. 2012文章编号: 0253-2409(2012 )06-071407Simultaneous operation of dibenzothiophene hydrodesulfurization andmethanol reforming reactions over Pd promoted alumina based catalystsMuhammad Yaseen',2 , Muhammad Shakirullah'Imtiaz Ahmad'，Ata Ur Rahman'，Faiz Ur Rahman' , Muhammad Usman2，Rauf Raxzaq2(1. Institute of Chemical Sciences, Unirersity of Peshanar, Khrbrpukhtunkhua, Pakistan; 2. State Key Laboralory of MuliphaseComplex Systenzs, Institule of Process Enginering, Chinese Academy of Sciences , Bejing 100190, China)Abstract: In the current study simultaneous reactions of hydrdesufurization (HDS) of dibenzothiophene (DBT) and reforming ofmethanol in a micro-autoclave reactor were studied over bi- metlic ( Co-Mo/Al2O, and Ni Mo/Al203) and tri-metallic ( Pd-Co-Mo/AL2O, and Pd-Ni-Mo/Al2O, ) catalyst systems which were prepared by incipient impregnation method. In situ bydrogenuilization and low Pd loadings were the major targets of this study. For comparison purpose, calytic activity was separatelydetermined for both the methanol reforming and HDS of DBT reactions as well. Ni based catalysts were confirmned with betteractivity than Co ones for both the reactions with Pd promoted ones ranking at the top i. e. Pd-Ni-Mo/ Al2O2> Ni-Mo/Al2Oz> Pd-Co-Mo/ Al20,> Co-Mo/ Al2O, where Pd-Ni-Mo/ Al2O, showed 91% DBT conversion at 380 C and 12 h reaction time. Some of theselected organic addtives on cataytic activity were tested for their efect toward HDS reaction which was unique with close relationto their chemical nature. Reaction products were quantitatively and qualitatively analyzed via HPLC and GC-MS techniquesrespectively which helped in elucidating reaction mechanism.Key words: in situ hydrodesulfurization of DBT; methanol reforming; Pd promoted alumina based catalyst; organic additiveCLC number: 0643Document code: AOne of the serious problems associated with theof our previous work with remarkable increase in theuse of fuel oils at an alarming rate is their damage tocatalytic activity (97% DBT conversion )15，Thesethe environment，especially by sulfur (S) and its .successive improvements in HDS activity werederivatives. For the control of these issues ，attributed to the enhanced reforming properties of Pdenvironmental regulations are becoming stringentthan RuC16] indicating the strong dependence of thewhich has led to the creation of new and advancedcurrent in situ hydrogenation process on reformingdesulfurization techniques over the past fewreaction. Based on this assumption， we furtheryears'121. Among these, oxidative desulfurization,investigated the reforming portion of the currentbiodesulfurization and catalytic hydrodesulfurizationprocess by analyzing some other in situ hydrogen(HDS) are few to namel1,3-6]. HDS has been theproduction sources. For example , it is considered thatmost effective one atributed to its flexible naturemethanol reforms more easily than ethanol[7] for theleading to many advancements in catalyst chemistryproduction of pure hydrogen on industrial scale andand process variables'. This process mostly utilizesthus can produce better results in the current in situCo or Ni promoted Mo based Al203 supprtedHDS process. Therefore, in the current phase of thecatalysts while enhancing effect by noble metals likeproject, ethanol reforming was replaced by methanolPt, Pd or Ru has also been repored3.8,9], amongreforming reaction over similar catalyst systemswhich Ru and Pd are considered better promoters ofreported earlier 5. Moreover， some organicthe conventional HDS catalystsi10-131.additives were used in fixed amounts to check theirRecently we for the first time reported the HDSeffect on catalytic activity. Activity in terms of bothof DBT over Ru and Pd promoted Al2O3 basedHDS and methanol reforming reactions was separatelycatalysts through a novel mechanism of in situdetermined. Ni was more active metal than Co in allhydrogen utilization generated via ethanol reformingcases. Reactions products were analyzed by GC-MSreaction' I4.15. . Incorporating Ru in to the conventionaltechnique while fresh and spent catalysts were(Co or Ni)-Mo type catalysts was very interestingcharacterized for BET surface properties. The presentand fruitful results (75% DBT) were reported via ourin situ bhydrogenation mechanism based study can benovel process"s . To further evaluate and analyze theapplied as an altemative approach toward HDS ofprocess, we targeted the in situ hydrogen productionDBT中国煤化工to its simple andreaction by replacing Ru with Pd in the second phaseCOSTYHCNMHGReceived date: 2011-12 20; Received in revised form: 201204-28.Foundation items: Fundamnental Rescarch Foundation of Sinopee (X505015)。Corresponding author: Muhanmad Yaseen, Tel: +92-91 921652, Fax: +92-91-9216652, E-mail: myousafzi@e gmail. com,本文的英文电子版由Elsevier出版社在SciecDirct上出版(ttp://w. siencedirect. con/siec/jouma/18725813)。第6期.Muharmmad Yaseen et al: Sirultancous operation of dibenzothiophene hydrodesufurization and .....7151 Experimentalis present in bulk amount compared to the negligible1.1 Chemical reagents Chemical reagents usedamount of in situ produced hydrogen (0.35 mol atin the current work were of analytical grade and weremost) via methanol reforming， resulting in anused without further purification. DBT was suppliedundetectable and negligible pressure gradient by theby ACROS Organics, New Jersey, USA. Al,O,，later. Moreover, the in situ type operation withphenol, naphhalene ，tetrahydronaphthalene ( THN )paralel production and consumption of hydrogeni. e. tetralin and anthracene were purchased fromminimizes the factor of fractional hydrogen pressure ,Tian jin Guang-fu fine Chemical Institute, China. nunlike conventional ex situ HDS process whichOctane was purchased from Tianjin Kermel Chemnicaldemands for high hydrogen pressure for its dissolutionReagent Company, China. Ethanol, o-xylene, Coin reaction mture-18-D1.(NO3)2 -6H,O, pyridine , diethylene glycol (DEG)Liquid products from the autoclave reactor wereand n-heptane were purchased from Tianjin Shi-Fu-dragged out after definite time intervals and wereChen Chemical Reagent Factory, China. PdCl andanalyses by HPLC and GC-MS techniques. Completedecahydronaphthalene ( DHN) i. e. decalin weredetails of these analyses are reported elsewhere' 5.provided by Sinopharm Chemical Reagent Co. Lld. ，2 Results and discussionChina. Ammonium heptarmolybdate ( NH,)。Mo,O2&* .2.1 Support and catalysts characterization4H2O and Pb ( CH,COO)2 were purchased fromTextural data of fresh and spent catalysts are presentedBeijing Chemical Works, China.1.2 Preparation and characterization of thein Table 1 and 2 respectively.Table I reveais that there is a considerabieCatalysts HDS catalysts in the curent study weredecrease in surface area of the support byprepared via classical incipient impregnation methodimpregnation of Co, Ni, Mo or Pd which is atibutedthe details about which have been reported10 their deposition over support surface. Logically,elsewherel511.3 HDS activity determination and productthis decrease should be greater in case of three metalanalysis HDS activity for the presulfided catalystssystems i. e. Pd-Co-Mo/Al,O, and Pd-Ni-Mo/was determined in a 250 mL batch autoclave reactor.Al2O3. These trends of change in structural propertiesTypical，each experimental run used 100 mare similar to our previously reported studiests. Insample, consisting of 90 mL of 9x10- DBT modelHDS reactions, Pd promoted catalysts were morefuel, 10 mL water methanol mixture (5 :5 v/v,active ( though had the least surface area) indicatingthat catalytic activity is mainly dependent on thecorresponding to 0. 27 and 0. 12 mol respectively) n0active specie ( Pd active phase) and partly on surfaceadditive ( blank experiments) or 10 mmol of aspecific organic additive and 0.5 g of catalyst. Detailsarea. Moreover, active phase formation is highlyof the reactor operation along with experimentalaffected by metal loadings which should be cricallymonitored i. e. low and excessive metal loadings canconditions have been reported elsewhere'The operating pressure at constant 380 Clead to insufficient active phase formation and activereaction temperature varied about 12+1 MPa. Thisphase blockage/ inactive phase ( Al2( MoO2)3 for Mopressure was assumed to be caused by n-octane whichsupported Al2O,bascd catalysts) respectivelyTable 1 Textural properties of the support and sulfided catalysts (fresh)Catalys/ supportBET surface areaPore volume Average pore_ Metal loading over AL20, w/%Usage (g)A/(m2.g' )v/cm'diarmeter d /nm CMoNPALO,(Pure)257.00.40.57Co-Mo/Al2O,176.00.30.680.5Ni-Mo/AlO,175. 80.65Pd-Co-Mo/Al,0,148. 10.74Pd-Ni-Mo/ Al2O,161.20.60In case of spent catalysts ( Table 2), Pd-Ni Mo/extents result in the production of more coke and SAl20, were found with the least surface area in thecompounds, resnectivelv_ which pet deposit overseries while in activity test for both methanolcatalys中国煤化Ilecreased surfacereforming and HDS reactions , Pd based catalysts wereareaYHCNMHGhermore, similarthe most active, suggesting that these two reactionstrend o1 cAtuIas Uo1a waD LUIU in ODr previousoccur to maximum levels over these catalysts.reports as well for ethanol reforming based in situReforning and HDS reactions occurring to largerHDS process14.15 ; strongly supporting the accuracy716燃料化学学报第40卷of curient methanol based process.Table2 Textural properties of the used catalystsCatalystAddive Time h/min Temperaturet/CBET Surface areaPore volumeAverage poreA/(m'.g")v /cm'diameter d /nmCo-Mo/AI2Odecalin380147.80.84Pd- Co-Mo/Al2O，dccalin80158.9.0.30.71Ni-Mo/ ALO,41.20.83Pd-Ni-Mo/Al2O3 decalin4118.01.01Pd-Ni-Mo/A12O31290.81.252. 2 Catalytic activity measurements andcontribute differently toward the final target of theoptimization of reaction parametersCatalyticprocess，therefore, it is essential to study themactivity was determined in terms of both HDS andseparately.methanol reforming reactions separately. Three typesn order to confirm whether reforming reactionf parameters i. e. reaction time, temperature andhas produced enough hydrogen required for theeffect of additives were optimized one by one. Thesedesulfurization of DBT sarmple, blank experimentsare discussed as follow.(with no DBT) using 90 mL n-octane, 5:5 water-2.2.1 Performance of different catalysts in terms ofmethanol (v/v) and 10 mmol of decalin as additive ,methanol reformingThe basic principal of thewere carried out at 380 C and 4 h reaction time. Thecurrent study is the in situ bydrogenation process ofGC-MS ( procedure discussed in section 2.3) data forDBT which is composed of two reactions i. e.he products are presented in Table 3 while themethanol reforming and HDS. Both reactions occurchromatogram is shown in Figure 1.simultancously inside the same reaction media andTable 3 Experimental data determined for the totalmethanol conversion and amount of hydrogen created at 380 c and4h reaction timeReaction conditionsTotal methanolDME producedHydrogentime h/min temperature vCconversion x/%m /molgenerated m /molCo Mo/AI2O,1000.008 40.309Pd-Co-Mo/Al2O,0.00450.333Ni-Mo/Al2O,100.00240.345Pd-Ni-Mo/AI2O,0.002 20.347(DME) was confirmed being resulted from the3.0dehydration of methanol:26-28]. From Table 3 it canbe seen that activity order for DME production was:Co-Mo/Al2O, > Pd-Co-Mo/ Al2O, > Ni-Mo/Al2O3 >Pd-Ni-Mo/A12O3，showing that Co based catalystsmethanol(beforeDME(arterrcaction)favor the unwanted dehydration reaction while Nireactiom)favorsthe reforming pathway which is in goodagreement with the reported literature'29. A 90 mLsample of 900x10° DBT solution needs a total 0).00. 005 mol hydrogen for complete desulfurization.0.5.1.0 1.5 202.5 3.0 3.5Table 3 shows that hydrogen generation order was:Time 1/minPd-Ni-Mo/ AI2O, >Ni-Mo/ Al2O, >Pd-Co-Mo/Al2Oz>Co-Mo/ AI2O; with net productions of0. 347, 0. 345,Figure 1 0C-MS chromatogram for the reforming0.333 and 0. 309 mol respectively. The activity of theproducts over Pd-Ni-Mo/ AL2O,current methanol reforming based approach looksFrom GC-MS analysis, no methanol was foundman中国煤化Ipreviously reportedetha;in the sense that thein the products over any of the four catalyst systemsCNMH(indicating 100 % mcthanol conversion in each casc.formici. piOuula lauly u1cs morein situMoreover in the product stream， dimethyl etherhydrogen'completely overcoming the hydrogendemands of the DBT sample. Considering the第6期Muhammad Yaseen et al: Simultaneous operation of dibenzothiophene hydrodesulfurization and ....717excessive amount of generated hydrogen, the netlower DBT conversion than ethanol based process. Indifference in HDS catalytic activity seems to behe current process， however,， HDS activity wasmainly dependent on the hydrogenation reaction ratherdirectly related to the amount of hydrogen generated,than on the reforming one. Moreover, it can bewith both following the order: Pd-Ni-Mo/A12O3> Ni-assumed that more hydrogen production in case of Pd-Mo/ Al2O,> Pd-Co-Mo/ Al2O3> Co-Mo/ Al2O3.Ni-Mo/ Al,O,causes better hydrogen dissolution in thereaction mixture due to comparatively high fractional75F 0C-Mo/AIL0，. Pd-Co-Mo/Al.O,pressure, resulting in higher HDS activity.。Ni-Mo/AI,O，Pd-Ni-Mo/AI,O,2.2.2Optimization of HDS reaction temperatureand time Effect of temperature on the HDS activity5tof the catalyst was determined in temperature range of320 ~ 400 C at 4 h constant reaction time usingdecalin as additive in the complete absence of externalhydrogen supply. The experimental data are shown in曾1s-「Figure 2.HDS being highly endothermic reaction is320340360380400favored by high temperature and thus DBT conversionTemperature t/Cshows a gradual increase up to 380 C (Figure 2).Figure 2 Influence of reaction temperature onThe usual HDS reaction temperature range is 350 ~DBT conversion at 4 h reaction time360 C[30~32], but in our case it reached up to 380 C,which may be a compromising result between thHDS reaction time was optimized in a range of 1reaction conditions for HDS and methanol reforming,~ 12 h at fixed temperature of 380 C using decalin asas shown in reaction 1.the additive. The results are presented in Figure 3CH,OH + H2O- +3H2+ CO2( Reaction 1)indicating that DBT conversion increases directly withThe normal reaction temperature for reaction 1 isthe reaction time supporting the slow nature of HDSabout 400 ~600 C and has been widely applied forprocess. Logically, at longer reaction time, there arehydrogen produtnion(,3. It demonstrates that ourbetter chances for both the reforming and HDSreaction, being a combination of HDS and reaction 1，reactions to occur up to optimum levels.must be kept in a temperature range that favors bothof them. A 380 C is thus the desired optimum10口Co-Mo/A1O，■ Pd-Co-Mo/A,.O,reaction temperature where an increase from it can瓷80■Ni-Mo/AI,O，日 Pd-Ni-Mo/Al.0,promote reforming reaction but inhibits the HDS oneas shown by the decline in HDS activity in Figure 2 at400 C.Figure 2 also shows that Ni is a better HDScatalytic metal compared to Co which is consistentwith reported external hydrogen gas based20 tprocsses351. Moreover, a 69 % DBT conversion by0. 5% loaded Pd promoted catalyst is quitechallenging to the high Pd loadings (1. 9% ) reportedTime I/hliteratureAn important fact deducted from theFigure 3 Influence of reaction time oncurrent study was that due to better reforming natureDBT conversion at 380 C reaction temperatureof methanol than ethanol; we expected higher DBTconversion by the former at similar operatingIn Figure 3, catalysts followed the same activityconditions. This was not the case though as almostorder as discussed in section 2.2.2. A 91% DBTsimilar HDS activities can be seen for the methanolconversion at 380 C within 12 h proved that Pd-Ni-( current study ) and ethanol based processMo/Al2O, having low Pd loading (0. 5%) is a( previously reported!'s). This can be explained onpror中国煤化工suls (using Pd asexcessively high occurrence of methanol reformingpronz1an our previouslyreaction. It is assumed that 100% methanolrepo|Y片CN M H Gased in situ HDSconversion ( in current study ) utilizes more catalyticprocessli4, mainly attributed to the better reformingactive sites, leaving lesser chances for/ thactivity of Pd than Ru. On the other hand ，almosthydrogenation reaction, resulting in a comparatively第6期Muhammad Yaseen et al: Simultaneous operation of dibenzothiophene hyrodesulfurization and ....Table 4 Products and their relative abundance fordifferent catalysts used at 380 C and dfferent reaction times having decalin as additives. NoCatalystTime h/minTemperature 1/CBP peak (%)DBT peak (% )ACo-Mo/Al2O,38025. 374.7BPd-Co-Mo/. ALO,3268CNi-Mo/Al2O349.550.5Pd-Ni-Mo/Al2O,65.134.9Co-Mo/AL2O,60. 939. IPd-Co-Mo/ AL2O,1262.137.9Ni-Mo/Al2O,8079.420.6HPd-Ni-Mo/ AL0291.5.8.5Al2O, catalysts studied in the current project,13 5900 t3 ConclusionsSimultaneous reactions of HDS of DBT anCreforming of methanol were performed over (Co/Ni)14.18 20.64-Mo-Al20, catalysts. Incorporation of even a verylow Pd metal loading i.e.0. 5% ( by weight) into theconventional catalysts greatly increased the catalytic2021.45activity. Ni based catalyst were always more active10.27A17.36| 21.86 28.47，2136708775 1than Co based ones for both the methanol reforming5101525and HDS reactions. It was concluded that methanolTime 1/minthough produces more hydrogen than ethanol viaFigure 5 Product distribution indicated by GC-MSreforming in the current process, but the excessiveusing Co-Ma/ Al2O2( conditis listed in Table5. S.No. A)utilization of catalytic active sites by the methanolIn Table 4, the DBT peak area decreases whilereforming prevents it from a remarkable increase inthat of BP increases as we move along the series fromHDS activity compared to the ethanol based process.A-D, suggesting that catalytic activity increcases in theAmong the additives tested, decalin, tetralin andsame pattem. The same trend is true for series E-H,naphthalene enhanced the HDS activity while the restwhere experiments were performed at 12 h reactioninhibited it The current desulfurization process basedtime. Briefly speaking， from GC-MS data， theon the simultaneous operation of methanol reformingcatalytic activity order was: Pd-Ni Mo/Al20,> Ni-and HDS of DBT reactions avoids the use of externalMo/Al2O,> Pd-Co-Mo/ Al2O,> Co-Mo/ Al2O3 whichhydrogen and may be applied as an altemativenot only confirms our experimental data but alsoapproach to the conventional DBT desufurizationshows that Pd is a good promoter ofNi( Co) - Mo/process on industrial scale.Referencesi1] Yu G, us, CHEN H, zHU z. Diesel fuel dulfirzionn with hydrogen peroxide ponoted by fomic acid and calyed by activaled carbon[J]. Carbon, 2005 , 43(11): 2285-2294.[2] QUz, YANN, JIAJ, wU D. Removal of dibenzothiophene from snoulated petroleum by iegraled riradiaion and Zx/ alumina catalyst[J]. App Catal B, 2007, 71(1/2): 108-115.5[3] BRUNE M, REMERT R. Dsufurization of liquid fucelel via fractional evaporation and subsequent hydrdeslfurizaion upstream a fucl cellsystem[J]. Ind Eng Chem Res, 2005, 44(25): 9691-9694.[4] FURUYA T, KIRIMURA K, KINO K, USAMI s. Thermophic bodsurzarion of dbencotiophene and is drivives by Mcobacrimphlei WU-FI[J]. FEMS Microbiol Llt, 2001, 204(1): 19-133.[5] MACHSOUDI s, VOssouGhi M, KHEIROLOMOOM A, TANAKA E, KATOH s. Bicdufuizatin of hyrocarbons and diesed fuels byRhodococus sp. strain P32CI[J]. Biochem Eng J, 2001, 8(2): 151-156.[6] MEESALA L, BALOMAJUMDER C, CHATTERJEE s, ROY P. Bicsufluiation of dnbnenthiophene using recombinant Pseudomonesstrain[J]. J Chem Technol Biotechnol, 2008 , 83(3): 294-298.soLus D, AGUDO A, RAMREZ J, KLIMOVA T. Hydsufuration of hindered dibenothiophenes on bifunctional NiMo caalysspported on zeolie alumina composite[J]. Catal Today, 2006, 116(4):4中国煤化工iophene compounds[J.[8] CHO A, MOON s. Development of highly active Co (Ni) Mo catalystsCatal Surv Asia, 2010, 14(2): 64-74.CNMHGi9] KANDA y, TEMMA c, NAKATA K, KOBAYASHI T, SUGIOKA M, LEMICHI Y. Preparation and promance of noble metalphosphides spprted on sli ass new hyrdosulurizatinn calys[J1., Appl Cal A, 2010, 386(1/2): 171-178[10] MOCHIDA I, SAKANISHI K, MA x, NAGAO s, ISODA T. Deep hydrdsulfrization of diesel fucl; Design of reaction proces and720燃料化学学报第40卷catalys[J]. Catal Today, 1996, 29(1/4): 185-189.[11] MAGYAR s, HANCSOK J, KALL0 D. Hydsuluirzation and hytroconvrsion of heary FCC gasoline on PdH-USY zolie[]. PuelProces Technol, 2005, 86(11); 1-1164.[12] NAVARRO R, CASTAROP, ALVAREZ-CALVAN M, PAWELEC B. Hydrdesuturization of dincthipene and a SRCO on sulide Ni(C) Mo/Al2O3 calyts: Efet of Ru and Pd pomoio[]. Cal Today, 2009, 143(1/2): 18.-14.[13] PAWELEC B, NAVARRO R, FERRO J. Hyrdsulurization over PdMo/HY zeolie calyt$[J]. Fuel, 1997, 76(1): 6171.(14] MUHAMMAD Y,LUY , SHEN C, u C. Dibenothioph.ne hyodsufuriation over Ru promoted alumina basedd calyts using in situgenerated bhydrogen[J]. Energy Convers Manage, 2011, 52(2): 1364-1370.[15] MUHAMMAD Y, u C. Dibenohipe hyosurizationi using in situ geaed hydngen over Pd promoled alumina based ealysts[J]. Fuel Proces Technol, 2011, 92(3): 624630.[16] IWASA N, TAKEZAWA N. New spored Pd and P alyly calysts for scam reoming and dhyrognation of mchanol[J]. Top Caral,2003, 22(3): 215-224.[17] LAOSIRIPOIANA N, ASSABUMRUNGRAT s. Calyie steam reorming of methane, methanol, and charol over Ni/YsZ: The psibleU8 of these fucls in intemal reforming SOFC[J]. J Power Soures, 2007, 163(2): 943-951.[18] SINGHAL G, ESPINO R, SOBEL J. Hyoiulfurzation of sufur herxeli compounds: Rexcion mechais[J]. J Cal, 1981, 67(2): 46456.[19] CIDR, gl LLAMBIAS F, GONZALEZ M，AGUDO A. Catalytic activity of Co, Mo and CoMo spore on NaY zeoli:Hydrodesulfurization of gas oil at high pesure[U]. Catal Lett, 1994, 24(1): 147-157.[20] VRADMAN L, LANDAU M, HERSKOWII M, ZRSKY V, TALIANKER M, NIKITENKO s, KOLTYPIN Y, GEDANKEN A. Hgh laingorshort ws; slabis inide SBA 15: Promotion wih nicke and prormane in byosutuiuzion and hyogeaion[J]. J Cal, 203, 213(2); 163-175.[21] GIORDANO N, BART I, VAGHI A, CASTELLAN A, MARTINOTTI C. Structue and calytie acivty of MoO3. AlO, systrms: I Solid.state properties of oxidized catalyts[J]. J Catal, 1975, 36(1): 81-92.[22] KLOUZ V, FERRO V, DENTON P, KAT7. H, LISSE 1, BOUVOT-MAUDUTT s, MTRODATOS c. Ethanol reorming for hydrogenproduction in a hybrid etric vehicl: Process optimisatio[J]. J Power Sources, 2002, 105(1): 26-34.[23] uU Y, HAYAKAWA T, SUZUKI K, HAMAKAWA s, TSUNODA T, ISHI T, KUMAGAI M. Hghly acive cpperceie calys forsteam reforming of methanol[J]. Appl Catal A , 2002, 223(1/2): 137-145.[24] FAUNGNAWAKIU K, KIKUCHI R, EGUCHI K. Thermodynamic evluation of methanol seam reforming for hydrogen producion[J]. J .Power Sources, 2006, 161(1): 87-94.[25] TRIMM D. Coke fomnation and mininisation during steam rfoming ractions[J]. Catal Today, 197, 37(3): 23238[26] VISHWANATHAN V, JUN K, KIM I, ROH H. Vapour phase debydnation of crude rmethanol to dimebyl ether over Ne moifed H-zSM.-5catalys[J]. Appl Catal A, 2004, 276(1/2): 251-255.[27] JUN K, LEE H, ROH H, PARK s. Catalytic dehydraton of methanol to dinehlyl ether (DME) over sliacid clys[J, Bul-KorChem Soe, 2002, 23(6) : 803-806.[28] KIMs, BAEK s, LEE Y, JUN K, KIMM, Y0O L Efet of yralumina content on calyie performance of moifed ZSM.5 for dehydation .of crude methanol to dimethy1 ether[J]. Appl Catal A, 2006, 309(1): 139-143.[29] FRENT s, CAVALLARO s, MONDELLO N, SPADARO L, FRUSTERI F. Production of hydrogen for MC fuel cll by scaom roming ofchanol over MgO spprted Ni and Co catalysts[J]. Catal Commun, 2003 , 4(6): 259-268.[30; CHEN J, RINGZ Hds rativtis of dexiopheic eompounds in aLC-fner LGO and HSANH, inbion efe([]t. Fael, 204, 8(3): 35-3.[31] AL ZEGHAYER Y, SUNDERLANDP, AC-MASRY w, AL-MUBADDEL F, IBRAHIM A, BHARITYA B, JIBRIL B. Activity of CoMa/rAl2O, as a ealyst in hyorusuluzatio:; Efetes of Co/ Mo ratio and drying cmdio[]. Appl Catal A, 2005, 282(1/2): 163-171.[32] SOLIS D, KLMOVA T, RAMfREZ J, CORTEZ T. NiMo/ AL2O2-MgO (x) calyts: The efet of the prolonged exposure to ambien airon the lextural and calytic poperte[J], Catal Today, 2004, 98(1/2): 99-108.[33 BOUKIS N, DIEM v, HABICHrT w. DINUS E Mchanol rfoning in sprcrtical wer[J]. Ind Eng Chem Res, 203, 42(4); 78-735.[34] LEET, BAE J, KIM JY. Mecthanol rcorming over PT ctalysts for polymer etrolyte membrane[J]. fuel clls[ Cj// ASME 2006 4hIntermationd Conferncc on Fael Cll Science , Egineeing and Technology( FUELELL2006). Iruine . Califrnia, UsA; 2006: 143-1147.[351 LECRENAY E, SAKANISHI K. MOCHIDA I, SUZUKA T. Hyrousufurization acivity of CoMo and NiMo calys sppord on someacidic binary oxides[J]. Appl Catal A , 1998, 175<1/2): 237-243.[36] BARRIO v, ARIAS P, CAMBRA I, GOEMEZ M, PAWELBC B, FERRO I. Hydodesurzaion and hyrogeaion of moel compoundson sicaumina suppored bimetallil systems[J1. Fuel, 2003, 82(5): 501 -509.[37] ISHARA A, KABE T. Deep desfrization of ligt oil: 3 Efctes of solvents on hydsurzaion of dororiphe[J], Ind EngChem Res, 1993, 32(4): 753-755.[38] GOL 0, ATANUR 0, ARTOK L, ERBATUR 0. The efiet of adiveis on yrdousuriation of dbenothiophene over buUk molydenumsulfd: Increased catalytic ativity in the presence of phno[. J]. Fuel Pocess Tchnol, 2008 , 89(4): 419 423.[39] SCIR s, MINIC s, CRISAFULL C. Selctive hydrogenaion of penol to cyclohexanone over sppred Pd and Pd-Ca calys: An .investigation on the influence of dfferett sports and Pd precursors[]. Appl Catal A, 2002, 235(1/2): 21-31.[40] CHEN Y, LIAW C. LEE L. Selective byrogention of phenol to cyelol中国煤化工calcined Mg/AI hyoalcile[J]. Appl Catal A, 199, 177(1):18.[41] ECOROVA M, PRINS R. Munal inftuence of the HDS of dibenothiopheneCNMHGI,202121)111.[42] BABICHI, MOUIINJ. Sience and lcandogey of novel proeses for dep dslinion of al reinry seams: A riew.J]. Foue, 2B, 8&(6): 0761.43] PRINS R, ECOROVA M, ROHTHLSBERGER A. ZHAO Y, SIVASANKAR N, KUKULA P. Mchanisms of byrdsufurization andhydndeitroenatio[J]. Catal Today , 2006, 11(1/2); 84-93.