Catalytic Performance and Texture of TEOS Based Cu/SiO2 Catalysts for Hydrogenation of Dimethyl Oxal

CHEM. RES. CHINESE UNIVERSITES 2012, 28(1), 119- -123Catalytic Performance and Texture of TEOS Based Cu/SiO2Catalysts for Hydrogenation of DimethylOxalate to Ethylene GlycolWANG Shu-rong , YIN Qian-qian and LI Xin-baoState Key Laboratory of Clean Energy Uilization, Zhejiang University, Hangzhou 310027, P. R. ChinaAbstract Highly active and selective Cu/SiO2 catalysts for hydrogenation of dimethyl oxalate(DMO) to ethyleneglycol(EG) were successfully prepared by means of a convenient one-pot synthetic method with tetraethoxysi-lane(TEOS) as the source of silica. XRD, H2-TPR, SEM, TEM, XRF and N2 physisorption measurements wereperformed to characterize the texture and structure of Cu/SiO2 catalysts with different copper loadings. The activecomponents were highly dispersed on SiO2 supports. Furthermore, the coexistence of Cu' and Cu* contributed a lot tothe excellent performance of Cu-TEOS catalysts. The DMO conversion reached 100% and the EG selectivity reached95% at 498 K and 2 MPa with a high liquid hourly space velocity over the 27-Cu-TEOS catalyst with an actual cop-per loading of 19.0%(mass fraction).Keywords Cu/SiO2; TEOS; Hydrogenation; Dimethyl oxalate; Ethylene glycolArticle ID 1005-9040(2012)-01-119-051 Introductioninfluence of SiO2 carriers(such as HMS and SBA-15) and theEthylene glycol(EG) is an important chemical product,deposition-precipitation temperature on the texture, structurewhich is widely used in polyester manufacture. At present, EGand catalytic performance of Cu/SiO2 for the hydrogenation ofis mainly synthesized by ethylene oxide hydrolysis, a routedimethyl oxalate to ethylene.which is significantly influenced by the price of crude oil andAmong the researches on the hydrogenation of dimethylconstrained by the accompanying huge energy consumptionoxalate(DMO) to EG, CwSiO2 catalysts were rarely reported,and the huge requirement for water'l. As a result, there hassynthesized by convenient one-pot method with tetraethoxysi-been considerable searches for an altermative and sustainablelane(TEOS) as the source of amorphous SiO2 support, whichapproach, namely, the synthesis of EG from syngasl2.31.was a widely used cheap material. In the present work, one-potThe indirect synthesis process has two steps, the couplingsynthetic method was adopted to prepare the CwSiO2 catalystsof CO with nitrite esters to oxalates and then the hydrogenationwith different Cu loadings. And then the catalytic performanceof oxalates to EGI4.5]. However, the catalysts used for the oxa-of the gas-phase hydrogenation of DMO to EG over these ca-late hydrogenation have been considered to be the main restric-talysts was evaluated by means of a continuous-flow fixed-bedtion on the industrial synthesis of EG from syngas and a greatreactor. XRD, H2-TPR, SEM, TEM, XRF and N2 physisorptiondeal of work has been carried out to improve their performance.vere performed to characterize the textural and structuralIn the early technology, ruthenium-based homogeneous cata-properties of the catalysts.lysts were used for the hydrogenation of oxalatesh.1. However,owing to the dificulties of separating the catalyst from product2 Experimentaland serious corrosion resulting from the use of homogeneouscatalysts, Cu-based heterogeneous catalysts have been more2.1 Preparation of Catalystsreccently investigated for the hydrogenation of dialkyl oxalates.Cw/SiOz catalyst was prepared via a onc-pot synthesis. AnAlthough scientists have developed a series of Cu-Cr/SiO2appropriate quantity of Cu(NO3)2:3H2O(A.R. grade, Sinopharmbased catalysts for the hydrogenation of oxalatesl&9, theChemical Reagent Ltd.) was dissolved in deionized water andtoxicity of Cr has limited the wide application of them. There-ethanol, as a co-solvent, was added to it while strring. Then thefore, researches into oxalate hydrogenation catalysts have beenrequired amount of 28%(mass fraction) aqueous ammoniafocused on the development of Cr-free Cu based catalystsl0- -14.solution was added drop by drop with continuous siring until*Corresponding author. E-mail: srwang@zju.cdu.cnRceived June 15, 2011; accepted October 10, 2011.中国煤化工，Supported by the National Science and Technology Supporting Plan ThrougAD22B06), theZhejiang Provincial Natural Science Foundation, China(No. 1110089), the Fundan:MYHC N M H Gcentral Univer-sities of China(No.2011FZA4012)， the Research Fund for the Doctoral Program of Higher Education of China(No.200901011 10034), and the Zhejiang Provincial Key Science and Technology Innovation Team, China(No.2009R50012).120CHEM. RES. CHINESE UNIVERSITIESVol.28the pH value of the mixture reached 12- -13, thus forming acopper ammonia complex solution. The required amounts of3 Results and DiscussionTEOS and copper ammonia complex solution were mixedand stired for 4 h at room temperature, and then aged for 24 h.3.1 Catalytic Activity of Prepared CatalystsThe suspension was transferred to an oil bath, preheated at 363The reaction scheme for the hydrogenation ofDM0 to EGK, for the evaporation of ammonia, water and ethanol, andis shown below:then filtered. The filtrate was washed with 500 mL of deionizedwater five times then dried at 393 K overmight. The catalystHOCH2COOCH;+2H2- +HOCH2CH2OH+CHzOH (1)precursors were calcined in static air at 723 K for 4 h, palle-HOCH2CH2OH+H2-→CHzCH2OH+H2Otized, crushed and seived to 40- -60 mesh. The prepared cata-The hydrogenation of DMO to EG proceeds via methyllysts were designated as w-Cu-TEOS，where w showed theglycolate(MG), and the EG can be further dehydrated to etha-nominal copper loading(mass fraction, %).nol.The catalytic performance2.2 Characterization of Catalystsin the gas-phase hydrogenation of DM0 to EG are summarizedN2 adsorption-desorption isotherms were recorded at 77 Kin Table 1. These tests were carried out under fixed conditions:)n a Quantachrom-Autosorb-l-C apparatus. Before thethe reaction temperature was 498 K, the system pressure was 2measurement, the sample was degassed at 573 K for 3 h. TheMPa and the H2/DMO molar ratio was 260. The actual Cuspecific surface area was calculatedloadings measured by XRF are ilustrated in Table 2. The expe~Brunauer-Emmett-Teller(BET) analysis of the N2 adsorptionrimental results show that the w-Cu-TEOS catalysts(w=16, 21,isotherms.27) had high reactive activity due to a similar amount of efcX-ray diffraction(XRD) patterms were recorded on a PA-tive Cu loadings(12.5%, 15.6%， 19.0%, mass fraction). WhenNalytical X'Pert PRO X-ray difractometer with a Cu Ka radi-the actual Cu loading was 19.0%(mass fraction), the conversionation source operated at 40 kV and 30 mA. The 20 range wasof DMO remained at 100% and the selectivity of EG was about10°- -90°0 and the scanning speed was 5 9/min.95%. However, the selectivity of EG decreased obviously withMorphologies of the catalysts were observed by scanningthe elevated selectivity of by-proucts(ethanol and 1, 2-BDO)electron microscopy(SEM, FEI Model SIRION-100).when the Cu loading increased to 26.5%(mass fraction) inThe particle size and distribution were determined by32-Cu-TEOS. This can be atributed to the improved activitytransmission electron microscopy(TEM, Philips-FEI, Tecnaaccording to higher copper loadings, while excessive activeG2 F30). .component loading may cause the aggregation of copper par-Temperature-programmed reduction(TPR) was conductedticles, leading to a lower actity for EG.on an Auto Chem II 2920 instrument. Calcined Cu/SiO2 sampleTable 1 Performance of w Cu/SiO2 catalysts for theof 30 mg was purged with Ar at 373 K for 30 min and thenhydrogenation of DMO to EGcooled to room temperature. The reduction was carried out withSelectivity(%)Catalyst10%H2-Ar(40 mL/min), and the sample was heated at a rate ofconversion(%) EG Ethanol MG__ 1.2-BDO10 K/min up to 873 K. The amount of H2 consumed was moni-16-Cu-TEOS959tored by a thermal conductivity detector.21-Cu-TEOS27-Cu-TEOS100The efective copper loadings of the prepared catalysts32-Cu-TEOSwere analyzed by X-ray fluorescence(XRF), which wasmanufactured by ThermoFisher, Model-IntelliPowerTM 4200,Table 2 Physical properties of w-Cu-TEOS catalystswith an X-ray tube of thodium anode and a scintillationdetec-Cu loading"Vpn1to(% mass(mg) (cmgt) Dnw/m cx'/mfraction)12.524.90.0711.06.5.3 Catalytic Activity Test15.617.70.011.26.The reactions were carried out in a continuous-flow19.026.016.2s.fixed-bed reactor. In each experiment, 3 mL of catalyst was_32-Cu-TEOS20.0.1019.4sandwiched with quartz sand and packed in a steel tube reactora. Measured by XRF; b. calculated according to the TEM analysis.with an inner diameter of 8 mm. A thermocouple was insertedThe infuence of liquid hourly space velocity(LHSV) wasinto the center of the catalytic bed to monitor the reactioninvestigated over 27-Cu-TEOS catalyst(see Fig.1). It is foundtemperature. Before the reaction, the catalyst was activated bythat the DMO conversion was maintained at almost 100% whenpure H2 at 623 K for4 h at a ramping rate of 4 K/min. Afterthe feed flow rate ranged between 0.4 and 1.4 hr'. The EGcooling to the reaction temperature, a solution of 12.5%(massselectivity was about 95% for feed flow rate between 0.8 andfraction) DMO in methanol was fed into the preheater by1.2 hr', which iani i _the feed capacity中国煤化工syringe pump at a constant flow rate, and vaporized and mixedcompared withing that a highwith the required amount of H2 at 473 K. The liquid productsperformance catTYHC N M H Gntrast, the recentwere condensed and analyzed by an Agilent 6820 GC equippedwork of Chen et al.ts showed that the conversion and sclecti-with a flame ionization detector(FID).vity declined dramatically when the feed flow rate was largerNo.1WANG Shu-rong et al.121than 0.25 h-'.catalytic activities. The BET surface area is an important but100F 5 s7 100cannot explain solely the factor affecting catalytic activity.Previous work has proved that relatively low BET surface area30一-8could result in better activity than that with higher BET surfacer15,18,.1910F- 603.2.2 Morphology ofPrepared Catalysts0叶The TEM images(Fig.2) show that particles on the0一H 20reduced Cu-TEOS catalysts with different loadings areuniformly dispersed. The particle sizes in the catalyst samplesare ina range of5- -7 nm. The average particle size of Cu was0.4 0.6 0.8 1.0 1.2 1.4LHSVAh-1calculated asFig.1 Catalytic performance of the 27-Cu-TEOSEnd'catalyst with different LHSVdexu=En,d;(2)●Ethanol;▲MG; vEG.where n; is the number of particles with an corresponding3.2 Characterization of Prepared Catalystsdiameter d(within a given diameter range)k0. The average sizeof the different catalyst samples were 6.5, 6.1, 5.6, 6.9 nm,3.2.1 Physical Propertiescorresponding to different loadings shown in Table 2. These Cunanoparticle sizes were exactly in the ranges of activated car-measured by the N2 physisorption measurement and the resultsbon supported copper nanoparticles with a diameter of (6+2)are summarized in Table 2. It shows that all the Cu-TEOSnm and the Al2O3 supported copper particles with a diameter ofcatalysts have a similar BET surface area of about 20 m/g. The5.4- -7.5 mm122. The histograms of the particles distributionpore volume and average pore size are in the range offor diferent loading catalyst samples were placed below the0.05- 0.11 cm/g and 11-19 nm accordingly.TEM images correspondingly. These small particles resulted inFrom this point of view, the relatively low surface area ofextremely high catalysts activity.Cu-TEOS catalysts has ltte influence on their fabulousA)(B)回20nm[20m[20nm60B)|[(C)e 40-> 60食30-e 40f30-善20-g 2020-点1e 10-246810Particle size/mParticle size/nmParticle size/hmFig.2 TEM images of the reduced w-Cu-TEOS catalystsw: (A) 16; (B)21; (C)27; (D)32. (A)-(D) are the particle size dstributions of the samples in images (A)- -(D), rspectively.Cu20(111) can be detected in the TEM image of eachsample at a larger magnification, and the Cu20 existed in27-Cu-TEOS catalyst is marked in Fig.3. The interplanar crys-1-0.246 nm. C0111)tal spacing values of0.244 nm and 0.246 nm can be assigned tothe presence of Cu20(11), while 0.207 nm belongs to thecopper particles. It indicates the coexistence of Cu2O and Cu in1-0207 im (uI1I)the reduced catalysts.Fig.4 shows the SEM images of all the as-preparedCu-TEOS catalysts. The active components have high disper-中国煤化工sity and have intensive interactions with SiO2 supports. TheYHCNMHGparticles are apparently spherical, producing a high surface areaFig.3 HRTEM image of 27-Cu-TEOSfor each single paticle.catalystNo.1WANG Shu-rong et al.123hydrogenation of DMO to EG. Yin et al.!I8] have reported that9] Haruhiko M.. Kouichi H,. Taizou U, Yasuo N.. Seizou 1, Takanorithe synergetic efect of Cu2 and Cu* was an important factor toT. Preparation of Enhylene Gilycol, JP 57123127, 1982, 235reveal the optimal performance of Cu/SiOz catalyst prepared by10] Kouichi H., Taizou U, Yasuo U, Catalyst Compotition for Produc-evaporation ammonia method, obtaining a high EG selectivitying Elhylene Gilycol and Process for Producing the Catalyst Compoof 95% under higher presure(2.5 MPa). Wang et al."9 hasstion, US 4614728, 1986shown that the synergism between Cu' and Cu* affects the[1] Thomas D. J, Wehri J. T, Wainwright M. s, Trimm D. L, Cant N.hydrogenation of DMO to EG over Cu-TEOS, with the focusM., Appl Catal. A, 1992.86, 101on the water/ethanol ratio. In the present work, the outstanding12] Vandergrift C. J. G, Elberse P. A, Mulder A., Geus J. w. AppL.activity of Cu-TEOS catalysts is atributed to the coexistence ofCatal, 1990, 59, 275Cu° and Cu*.13] Kohler M. A, Lee J. C.. Trimm D. L, Cant N. w, Wainwright M.S., Appl Catal. 1987, 31, 3094 Conclusions14] ZhuY. Y, WangS. R, ZhuL. J, GeX.L,LiX. B., LuoZ Y,Catal. Lett. 2010, 135, 27515] ChenL F., Guo P. J.. Qiao M. H, YanS. R, Li H. Xx, Shen W, Xulysts for the gas-phase hydrogenation of DMO to EG have beenH. L.. Fan K. N.,J. Catal. 2008, 257, 172successfully prepared via a convenient one-pot synthetic me-16] YinA. Y., Guo Xx. Y, Dai W. L, Li H. X, Fan K. N, Appl. Catal. A,thod with TEOS as the source of amorphous silica. When the2008, 349, 91actual Cu loading was 19.0%, a DMO conversion of 100% and17] GuoX. Y., YinA. Y, Dai W. L., Fan K. N, Catal. Lett, 200, 132,an EG seletivity of 95% were obtained under optimized hy-drogenation conditions. In addition, a larger feed flow rate was18] Yin A Y, Guo x. Y, Dai W. L, Fan K. N, Catal Commur, 2011,found, compared with previous work, by means of the prepared12, 412catalysts. The presence of Cu' and Cu* makes great contribu-19] WangS. R, ZhuL. J.. Ge x. L, LiX. B.. Chin Chem Lett. 2011,tions to the enhanced catalytic performance of Cu-TEOS cata-22, 362lysts.[20] HouZ. Y. Gao J.,. GuoJ. z.. Liang D.. Lou H, Zheng x. M, J.Catal, 2007, 250, 331References21] Alonso F, Moglie Y, Radivoy .,. Yus M., Adr. Synth. Catal, 2010,352, 3208[1] Jaap w. H, John s. L, Zhang x. K., Catal. Today, 2007, 123, 3102] ZhangX. R, YaoC. z, WangL. C, Cao Y, Wu D. Sum Y. H, Dai[2] ZhaoX. G, LvX. L, Zhao H. G, ZhuY. Q.. Xiao W. D., Chin. J.W. L, Fan K N, Chem. J. Chinese Universities, 2005, 26(6), 1137Catal,. 2004,27, 12523] XiJ. Y, Wang Z. F, Lu G. X, Chem. Res. Chinese Unvrities,[3] Ishino M., Tanura M, Deguchi T, Nakamura s, J. Catal, 1992,2001, 17(4), 420133,32524] Marchi A J, FieroJ. L G. Santamaria J. Monzon A, Appl. Catal.[4] Bartley W. J, Williamn J, Charleston w. V, Process for the Prepa-A, 1996, 142,375ration of Eithylene Glbycol, us 4677234, 198725] Yin A. Y, GuoX Y, Dai W. L, Fan K N,J. Phys. Chem. C, 2009,[5] Misunobu M, Kouzou F., Keigo N, Masaoki M. Hiroyuki s,113, 11003Continwous Production ofOralie Ester,JP 54103817, 1979, 9526] Li z., HuangH. B, Xie K C. Chem. J Chinese Universities, 2008,6] Mattoli U, Bianchi M, Menchi G, Frediani P, Piacenti F, J. Mol.29(8), 1609Caral, 1985. 29, 26927] ui z, Wei C. M, Zheng H. Y, Xie K. C, Chem. J. Chinese[7] Teunissn H. T, Elsevier C.1, Chem Comman.[8] Zehner L. R, Lenton R w., Process for the Preparation of EthyleneUniversities, 2010, 31(1), 14528] He z, Lin HQ, He P., Yuan Y. z,.J Catal, 2011, 277, 54Glycol, US 412245,5 1978中国煤化工MHCNMH G