Catalytic Conversion of Methanol by Oxidative Dehydrogenation

Available online at www.sciencedirect.comScienceDirect|Journal of Natural Gas Chemistry 16(2007)1-5SCIENCE PRESSww.esvie.comomoltfarecpArticleCatalytic Conversion of Methanol by Oxidative DehydrogenationToshihito Ohtake*，Tohru Mori, Yutaka MorikawaDepartment of EBlectronic Chernistry, Interdisciplinary Graduate Schoo of Science and Engineering,Tokyo Institute of Technology, 4259 Nagatsuta, Midoriku, Yokohama 226 8502, Japan[Manuscript received October 16, 2006; revised November 30, 2006]A bstract: This study investigates the effects of addition of oxygen on the oxidative dehydrogenation(ODH) of methanol when a fluorotetrasilicic mica ion-exchanged with palladium (Pd2+. _TSM) was used asthe catalyst. The reaction proceeded at a very low temperature in the presence of oxygen, and HCOOCH3was obtained at high selectivity. By calculating the equilibrium conversion, it has been shown that substan-tial ODH took place for HCOOCH3 production. Consequently, this reaction would make dehydrogenationthe dominant reaction at equilibrium. Not all the H dissociated from CH3OH was converted to H2O byoxidation. It has been shown that the H2O was not produced from oxidative dehydrogenation by thedirect reaction of CH3OH and O2 when an attempt was made to carry out oxidative dehydrogenationusing an isotope oxygen trace method in the gas phase. Therefore, when CH3OH was converted to CO2and dehydrogenated to HCOOCH3, the C-O bonds were not dissociated.Key words: C1 chemistry; conversion; methanol; oxidative dehydrogenation; catalysis1. Introductionusing a palladium catalyst, and the activation processof methanol and the roles played by oxygen were in-Formaldehyde and acetaldehyde can be obtainedvestigated. The palladium catalyst was supported onfrom methanol and ethanol, respectively, by dehy-fluoro tetrasilicic mica (TSM), which has no activity,drogenation and oxidation. Different industrial tech-and is present as Na[Mg2.sSi4O1oF2] in layered silicateniques have been developed for these processes in theminerals [13,14].past. In recent years, there has been increasing in-terest in the oxidative dehydrogenation (ODH) of the2. Experimentalcorresponding alcohols. ODH of alcohols seems to bea promising alternative to partial oxidation because itPalladium catalyst was prepared using ion-involves a more simple, single-step reaction that canexchanged fluoro tetrasilicic mica (Na+-TSM) as atake place in tubular reactors, provided that high ac-precursor. The Na+-TSM powder was refined fromtivity and selectivity can be achieved under mild con-10 wt% sol made by Topy Industries. Pd2+. -TSMditions of temperature and pressure. There are sev-catalyst was prepared by dispersing the Na+-TSM ineral reports on the chemical adsorption of methanol3% ammonia solution to a concentration of 0.15 wt%.on various transition metals [1 -11]. Methanol reacts[Pd(NH3)4]Cl2-0.68H2O (N. E. Chemcat Corp.) waswith intermediates at low temperature [12], and atadded to the stirred Na+-TSM solution in a quan-higher temperature, it will react with bydrocarbons.tity中国煤化工exchange capacityThese reactions will proceed by the activation of the(2.54occurred rapidly.0- -H bond in methanol that forms surface methox-Pd2-:YHC N M H Gately after mixing,ide species and hydrogen atoms. ODH was attemptedbut was stirred for a further 24 h. The deposits* Corresponding author. Tel: +81-6-6850- 6239; Fax: +81-6-6850 6236; E-mail adress: ohtake@nakato.chemn.es osau.acjip2Toshihito Ohtake et al./ Journal of Natural Gas Chemistry Vol. 16 No. 1 2007were washed with water, removed by decantation, andobtained. This result suggests that HCOOCH3 wasdried at 60°C for a further 24 h.produced by the CH3OH dehydrogenation at aboutDehydrogenation of methanol was carried out u8-200 °C, as shown in Equation (2); furthermore, CH4ing a fixed bed flow reactor system. Methanol wasand CO2 could be observed after decarboxylation ofsupplied by a microfeeder (Furue Science, Ltd). AHCOOCH3, as shown in Equation (3). The selectivityreaction tube of inner diameter of 15 mm and lengthfor CH4 and CO2 decreased at higher temperature.of 350 mm and made of Pyrex glass was flleld with2 CHzOHHCOOCH3+2 H2(2)0.5 g of catalyst. The Pd2+-TSM was treated byHCOOCH3CO2+CH4(3)heating in N2 flow immediately before the reaction.The pretreatment of the catalyst was carried out in aCO and H2 produced from CH3OH decomposition30 ml/min N2 stream at 400C for2 h. The dehydro-can yield CH4 by methanation, as shown in Equationgenation of methanol was carried out in the presence(4), and the decomposition products of CH3OH in-or absence of oxygen. The amount of methanol sup-creased at higher temperature, whereas the selectivityplied was 12.5 mol% under a N2 stream, and the totalfor CH4 decreased.gas flow was 34.3 ml/min. In the case when oxygen3H2+CO - →CH4+H2O(4)was present, the ratios were O2/CH3OH=0.14 and0.28. The gas analysis was carried out using a gas00 rchromatograph (Shimadzu Corporation GC-7A, GC-ConveConversion8A).80 L。CHThe oxidative debydrogenation was also carriedout using an oxygen isotope (18O2; Prochem Inc.) to60investigate the roles played by oxygen.1802 in the gasphase was measured by a pulse method using a GC-MAS (Hitachi Inc, M-80) at 50。C, O2/CH3OH=0.2840and 30 mg of catalyst. The measuring condition was5 pulses/ml, and the Porapak Q (2 m) GC column20was maintained at 150 °C.Temperature programmed desorption (TPD) wascarried out on the catalyst before and after pretreat-2503050ment. The 0V-1 (1 m) GC-MAS column was main-tained at 150 °C. The TPD condition was 10 °C/min,Figure 1. Conversion and selectivity for CO, CH4,and the flow rate of the He carrier was 30 ml/ min.and CO2 over Pd-TSM catalyst in the ab-sence of oxygen3. Results and discussion3.2. Dehydrogenation in the presence of oxy-3.1. Dehydrogenation in the absence of oxygengenDehydrogenation of methanol was carried outFigure 2 shows the dehydrogenation reaction ofwith Pd2+-TSM in the absence of oxygen, as shownmethanol in the presence of oxygen, and the activ-in Figure 1. The stability does not change with time,ity does not change with time.The O2 ratio isand no products were obtained except CO, CO2, CH4,O2/CH3OH=0.28, and it is about 1:5 for completeand HCHO. It has been shown that CO selectivity in-oxidation reaction of CH3OH. The products are onlycreased with the conversion. Following equation (1),HCOOCH3 and CO2, and the formation of hydrogenit was shown that the decomposition of methanol pro-is not clear. The reaction proceeded at a much lowerceeded at over 200 °C.temperature than in the absence of oxygen, and theCHzOH→CO+2H2(1)conversion was about 40% at 60 °C. The maximurm中国煤化工which was constantIt has been shown that CH3OH decompositionatFHCOOCH3 obtaineddid not proceed easily at about 200 °C because nob;YHCNMHGowedabout65%8e-significant amount of CO was produced as in Equa-lectivity at 60。C, and the CO2 selectivity increasedtion (1). However, 50% CH4 selectivity and CO2 was at higher temperature. Hence, the main difference isJournal of Natural Gas Chernistry Vol. 16 No.12007the high HCOOCH3 selectivity at a lower tempera-to HCOOCH3, and H2 is consumed by oxidation toture compared with the selectivity in the absence ofH2O. Therefore, the H2 conversion was calculatedoxygen.from the amount of H2 formnation, O2 consumption,and CO2 formation. The plots were constant at about30 r70% in the expanding temperature region. This indi-cates that about 70% of the hydrogen included in theCH3OH formed H2O by oxidation and about 30% hy-0Fdrogen underwent no change by desorption. This re-sult shows that HCOOCH3 is not formed by oxidativedehydrogenation by the direct reaction of CH3OH and40 FO2: because the added oxygen will consume the hy-drogen produced by ODH, as shown in equation (2)ConversionHCOOCH3 will be produced.0tHCOOCH,Figure 3 shows the conversion and selectivity withH conversionO2/CH3OH=0.14 at 35-400 °C. The conversion wasconstant in the range 35- -200 °C and increased in50 80 100 120 140160180 200220the range 200 - 400 °C. The tendency of selectivity forReaction temperature (C)Figure 2. Conversion and selectivity of HCOOCH3,HCOOCH3 and CO2, were the similar at 60-200 。C,CO2，and H2 conversion in the presenceas shown in Figure 2. The CO selectivity increasedof oxygen at the rate of O2/CHgOH=0.28.with the conversion, which is similar to the reactionThe H2 conversion was calculated by thein the absence of oxygen, as shown in Figure 1. Theamounts of the H2 formed, O2 consumed,H2 conversion was not very high, although the CO2and CO2 formedselectivity was high in the range 120- -250 °C. TheseTable 1 shows equilibrium constants and calcu-results suggest that the role played by oxygen is notlated equilibrium conversions to form HC0OCH3 forjust the simple oxidation of methanol.the dehydrogenation of methanol in the presence ofoxygen: it was 10.4% at 60 °C. Because HCOOCH310was not produced in the absence of oxygen (Figure 1)but was produced with high selectivity at low temper-ature in the presence of oxygen (Figure 2), the oxygen区would make the dehydrogenation to be the dominantreaction when the reaction reached an equilibrium/●HCOOCH,state for the oxidative dehydrogenation of methanol.COCO,Table 1. Equllibrium constant and equilibriumH2 conversionconversion to form HCOOCHg for dehydrogenationin the presence of oxygenReactionEquilibriumtemperature (°C)constantconversion (%)604.83x10-110.41002003004009.80x10-110.915(1.5311.5Figure 3. Conversion and selectivity for HCOOCH3,2.0111.9CO，CO2， and H2 conversion in thepresence of oxygen at the rate of2502.4212.3O2/CHgOH=0.142.7712.73503.0113.0Methanol debydrogenation with O2/CHgOH=0.283.34at 100 °C was carried out with Pd-TSM pretreatedund中国煤化工rwn in Figure4. TheNext, H2 conversion was calculated to investi-prod=d CO2, and the for-gate the rate of formation of H2O by oxidation ofmatiYHC N M H GThe conversion wasH2, as shown in Figure 2. H2 is formed by bothvery low at 100- -200 °C. As shown in Figure 5, TPDCH3OH decomposition and CHgOH dehydrogenationmeasurements were carried out at 10 。C/min to de-Toshihito Ohtake et al./ Journal of Natural Gas Chemistry Vol. 16 No.12007termine the reason for the absence of activity. A peakladium at 40° and 47°. Thus, we found the product ofat 100 °C is obtained for H2O desorption, and peaksmetal palladium after the 400 °C pretreatment. Theat 250- -350 °C were observed for NH3 desorptiontwo peaks of the catalyst after reaction are somewhatbased on the [Pd(NH3)4]Cl2. Because N2 was alsobroad, which is attributed to redispersion.desorbed simultaneously, part of the NH3 ligand wasdecomposed to N2 and H2 on palladium. Hence, Pdcatalysis not including the NH3 ligand will give theactivity, and the Pd(NH3)42+ catalysis including theNH3 ligand does not find the activity.00 r●HCOOCH,80 t， CO22)B H2 conversion50 t1)5001(°)Figure 6. XRD spectra of the Pd-TSM catalyst for20 tbefore pretreatment (1), after pretreatment(2), and after reaction (3)000 150 200 250 3035404503.3. Reaction of oxygen isotopePretreatment temperature (C)Figure4. Plots of conversion and selectivity forHCOOCHs, CO2, and H2 conversion versusPulse reaction was carried out over the catalystpretreatment temperature. The reaction isusing 1802 in gas phase, and the gas composition wasat 100 °C and O2/CHsOH=0.28O2:CH3OH:Ar=0.28:1:6.7. Table 2 shows isotopic la-beling O2, CO2, and H2O measured by mass spec-trometer. Peeaks of 18O2, 1602, C160180, and Cl8O2used each mass (m/z= 32, 36, 44, 46) and those ofH28O and H2O applied m/z=17 and 19 to avoidAr2+. The HCOOCH3 formed was HC16O16OCH3,and HCOOCH3 containingl8O did not exist. Theseresults indicate that HCOOCH3 is formed by dehy-(5) x2畲|~drogenation and condensation from CH3OH, and 0(4)atoms in HCOOCH3 do not come from O2 in the gasphase. The ratio of 16O2:1802=1:3 was obtained in(3)the prepared gas although 160180 (m/z= -34) was not(2)detected at all. Therefore, isotope exchange for 16O2(1) x5and 1802 will hardly occur on the catalyst although1602 will be mixed in the gas in the preparation.10000300160180 could have been desorbed through the iso-tope mixture on the catalyst surface if O2 in the gasFigure 5. Temperature programmeddesorption(TPD) traces obtained in 30 ml/min Hephase could adsorb by dissociation, and the O atomnsgas flow at m/z=15 (1)，16 (2), 17 (3), 18would be at equilibrium with the O2 in the gas phase(4), and 28 (5)through adsorption, as shown in Equation (5) givenThese pretreated Pd-TSM catalysts were mea-belon中国煤化工sured by XRD at 400 °C before and after pretreat-YH，CNMH G+160(a)(5)ment, as shown in Figure 6. In the case of the afterpretreatment catalyst, two peaks exhibited metal pal-160180(g)←18O2(g)←→180()+180(a)(6)Journal of Natural Gas Chemistry Vol.16No. 1200In addition, the rate of oxygen consumption willdehydrogenation to HCOOCH3 was highly seletive.be suficiently rapid not to return to the gas phaseBy calculating the equilibrium conversion, it has beenbecause 160180 cannot be detected in the gas phase.shown that the debydrogenation to HCOOCH3 tookTable 2. lon intensity rate of products when usingplace and that the added oxygen would make dehy-oxygen isotope by mass spectrometrydrogenation the domninant reaction at equilibrium.PulseIon intensity rate of productsThe mass balance was calculated for H and 0 fornumber 1802/1602 CI60180/Ci8O2 H0/H°0methanol dehydrogenation in the presence of oxygen.3.40.330.14The result showed that not all the H dissociated from22.60.370.052CH3OH was converted to H2O by oxidation but a part2.40.300.016of the H was desorbed as H2. Therefore, it can be said2.80.230.0094HCOOCH3 was not produced from oxidative dehydro-3.20.190.042genation by the direct reaction of CH3OH and 02.Furthermore，a pulse reaction was carriedWhen CO2 was measured in the same manner,out using 1802. The HC0OCH3 formed wasonly Cl6O2 and C160180 were obtained, and C18O2HCI6Ol6OCH3, and HCOOCH3 that contained 180could not be detected by mass spectroscopy. Table 3did not exist. Hence, the HC0OCH3 would be formedshows the ratios of Cl6O2, C160180, and C1802.by dehydrogenation and condensation of CHyOH, andC18O2 was calculated to be about 0.75% -2.5%, al0 atoms in HCOOCH3 were not produced from O2though we could not detect it. Because these valuesin the gas phase. CO2 was also formed in the reac-are suficient to be detected by mass spectroscopy iftion and consisted of Cl6O2 and C18O2, but Cl60180ClI8O2 exits, we conclude that it is not produced.was not detected. Consequently, the C-O bonds inHence, the C-0 bond in CH3OH could not be disso-CH3OH were not dissociated by the dehydrogenationciated by the dehydrogenation.and the CO2 conversion.Table 3.Pomposition of C10O2, C10180, andReferencesC18O2 when using oxygen isotopeComposition (%)Pulse numberCl6O2Cl608O[ Wachs IE, Madix R J. Surf Sci, 1978, 76: 531173.624.42.0[2] WachsIE, Madix RJ J Catal, 1978, 53: 20870.926.62.5[3] Franaszczuk K, Herrero E, Zelenay P, Wieckowski A,376.022.31.6Wang J, Masel RI J Phys Chem, 1992, 96: 850980.81.04 Brogan M s, Cairns J A, Dines T J, Rochester CH.583.515.80.75Spectrochim Acta A, 1997, 53: 9435] Wang J, Masel RI. Surf Sci, 1991, 243: 1996] WangJ, MaselRI. J Vac Sci Technol A, 1991, 9:4. Conclusions1879[7 Gibson K D, Dubois L H. Surf Sci, 1990, 233: 59Methanol dehydrogenation was carried out over[8 Ehlers D H, Spitzer A, Luth H. Surf Sei, 1985, 160:a Pd-TSM catalyst. Methanol conversion was car-ried out in the absence of axygen. CO selectiv-[9] Shustorovich E, Bell AT. J Catal, 1988, 113: 341ity was increased by raising the reaction tempera-[10] Kua J, Goddard II W A. J Am Chem Soc, 199,9 121:10928ture. Methanol decomposition was the main reaction[11] Ishikawa Y, Liao M, Cabrera CR. Surf Sci, 200,0 463:at temperatures greater than 250 °C, and CH4 and66CO2 were produced at lower temperatures. This re-12] Vllegas I, Weaver MJ. J Chem Phys, 1995, 103: 2295sult suggests that CH3OH dehydrogenation yielded[13] Morikawa y, Takagi K, Moroka Y, IlawaT. J ChemHCOOCH3 and that this HCOOCH3 was decarboxy-Soc Chem Commun, 1983, (15): 845lated to CH4 and CO2. In the presence of oxygen,14] MorikawaY fiote F Moronoka y, IkawaT. Chem Ltt,the reaction proceeded at very low temperature. The中国煤化工YHCNM HG