Microsoft Word LYX 论文 英+中 274.doc

Transcription

1 Chinese Journal of Catalysis 36 (215) 催化学报 215 年第 36 卷第 3 期 available at journal homepage: Article The catalytic effect of H2 in the dehydrogenation coupling production of ethylene glycol from methanol using a dielectric barrier discharge Jing Zhang a, Teng Li a, Dongjiang Wang a, Jialiang Zhang b, Hongchen Guo a, * a State Key Laboratory of Fine Chemicals, Department of Catalytic Chemistry and Engineering, School of Chemical Engineering, Dalian University of Technology, Dalian 11624, Liaoning, China b School of Physics and Optoelectronic Engineering, Dalian University of Technology, Dalian 11624, Liaoning, China A R T I C L E I N F O A B S T R A C T Article history: Received 17 August 214 Accepted 12 October 214 Published 2 March 215 Keywords: Dielectric barrier discharge plasma Hydrogen catalysis Ethylene glycol synthesis Hydrogen atom Carbon hydrogen bond activation The catalytic effect of H2 in the one step synthesis of ethylene glycol (EG) from methanol dehydrogenation coupling reaction using dielectric barrier discharge (DBD) was studied by in situ optical emission spectroscopy and online chromatographic analysis. The influence of discharge frequency, methanol and H2 flow rates as well as reaction pressure was investigated systematically. Results show that, in the non equilibrium plasma produced by DBD, H2 dramatically improved not only the conversion of methanol but also the selectivity for EG. Using the reaction conditions of 3 C,.1 MPa, input power 11 W, discharge frequency 12. khz, methanol gas flow rate 11. ml/min, and H2 flow rate 8 18 ml/min, the reaction of the CH3OH/H2 DBD plasma gave a methanol conversion close to 3% and a selectivity for EG of more than 75%. The change of the EG yield correlated with the intensity of the Hα spectral line. H atoms appear to be the catalytically active species in the reaction. In the DBD plasma, the stable ground state H2 molecule undergoes cumulative collision excitation with electrons before transitioning from higher energy excited states to the first excited state. The spontaneous dissociation of the first excited state H2 molecules generates the catalytically active H atom. The discharge reaction condition affects the catalytic performance of H2 by influencing the dissociation of H2 molecules into H atoms. The catalytic effect of H2 exhibited in the non equilibrium plasma may be a new opportunity for the synthesis of chemicals. 215, Dalian Institute of Chemical Physics, Chinese Academy of Sciences. Published by Elsevier B.V. All rights reserved. 1. Introduction Ethylene glycol (EG), the simplest diol, is used extensively all over the world [1] and is currently produced by the hydration of ethylene oxide. However, with the depletion of petroleum resources, the synthesis of EG from coal is more appealing [2,3]. The coal based EG production process involves dimethyl oxalate as an intermediate and requires the oxidative dehydrogenation of syngas, oxidative esterification, coupling of CO and hydrogenation of dimethyl oxalate to produce EG. Although many researchers have investigated this method, problems due to the multistep process, high capital input and complex technology remain as impediments [4,5]. We recently reported a one step synthesis of EG from methanol by dielectric barrier discharge (DBD) in the presence of H2. The 71.5% EG selectivity and 15.8% methanol conversion have been reached at atmospheric pressure [6]. The reaction pathway for the one step synthesis of EG from methanol is characterized by the breaking of the C H bond of methanol in the non equilibrium plasma of the DBD. The dissociation of one * Corresponding author. Tel/Fax: ; E mail: hongchenguo@163.com DOI: 1.116/S (14) Chin. J. Catal., Vol. 36, No. 3, March 215

2 Jing Zhang et al. / Chinese Journal of Catalysis 36 (215) methanol molecule gives one hydroxymethyl radical and one H atom; the coupling of two hydroxymethyl radicals gives the desired EG molecule. The simultaneous coupling of two H atoms leads to a H2 molecule. Theoretically, this reaction has an atom economy of 1%, as the EG and hydrogen are both value added products. Obviously, the one step synthesis of EG is much more attractive for commercial application. What is more, as the selective activation of the C H bond of methanol molecule is made difficult by conventional catalytic methods due to the bond energy of C H falling between that of the O H bond and the C O bond [7], the function of the non equilibrium plasma in the selective dissociation of the C H bond of methanol might help research into the selective activation of other chemical bonds. In the previous work [6], we observed hydrogen acting as a catalyst during the one step synthesis of EG from methanol in a DBD. In this paper, with in situ optical emission spectroscopy (OES) and online chromatographic analysis, we have studied the relationship between hydrogen dissociation activation and methanol dehydrogenation coupling to form EG under different discharge conditions. On this basis, the catalytic mechanism of H2 in the one step EG synthesis from methanol is discussed. 2. Experimental The schematic of the experimental setup and the reactor structure is shown in Fig. 1. H2 (99.99%, Dalian gas company, Ltd) was introduced into a vaporizing mixer and reactor under the control of a D7 19B mass flow controller (Beijing Sevenstar Electronics Co. Ltd., Beijing, China). When all the air in the reaction system had been fully replaced by H2, liquid methanol (AR, Sinopharm Chemical Reagent Co., Ltd, Shanghai, China) was introduced by a pump. It was vaporized and mixed with H2 in the vaporizing mixer, and this mixture was introduced into the plasma reactor through a thermally insulated pipeline. Then, the high voltage (HV) power supply (CTP 2 K cold plasma power supply, Nanjing Suman Electronics Co. Ltd) was turned on and the voltage output was regulated to generate a DBD. The voltage, current and power of the discharge were measured on an oscilloscope (DPO312 digital oscilloscope, American Tektronix Company, Beaverton, United States). The H2 dissociation in the plasma zone was observed with an in situ OES (SP2758, Princeton Instruments, United States; grating 3 G/mm, exposure time 1. s). The effluent of the reactor was introduced though a second insulated pipeline into a double detector online GC for compositional analysis (GC SP 689, Rainbow Chemical Instrument Co. Ltd. Shandong Lunan. FID for the detection of CH3OH, HOCH2CH2OH, C2H5OH, C3H7OH and CH4, etc., TCD for the detection of CH4, CO, H2O, etc.). The CH3OH conversion (X) and product selectivity (S) were calculated according to the following equations: X(CH3OH) = (n(ch3oh) n(ch3oh))/n(ch3oh) Si = ni(ch3oh)/(n(ch3oh) n(ch3oh)) where, n(ch3oh) and n(ch3oh) were the moles of CH3OH before and after the reaction, respectively, and ni(ch3oh) was the moles of CH3OH converted to product i. 3. Results and discussion 3.1. The promoting effect of H2 in the synthesis of EG from methanol under DBD The conversion of methanol using different discharge methods has been extensively studied, but the purpose of those studies was to produce hydrogen and/or syngas [8 17]. It is worth mentioning that Liu s group [18,19] once observed trace amounts of EG in the product of their methanol to hydrogen reaction using a corona discharge. In our previous work, we found two factors played key roles in the direct synthesis EG from methanol discharge: one was the discharge intensity of DBD, and the other was that methanol had to be fed into the DBD reactor together with the H2. A moderate discharge intensity was favorable for the production of EG. In this paper, the effect of the H2 flow rate on EG synthesis was investigated in a DBD reactor of moderate discharge intensity at a fixed methanol flow rate. As shown in Table 1, in the absence of H2, the conversion of methanol was only 9.6%. The major products were CO and CH4, making EG the minor product. The selectivity towards these products were 55.2%, 16.6% and 8.%, respectively. At a H2 flow rate of 8 ml/min, however, the conversion of methanol reached 3.5%, 2.2 times greater than without H2. The selectivity for EG reached 75.4%, which is 8.4 times greater than without H2. These results showed that the presence of H2 improved not only the conversion of methanol but also the selectivity for EG. This is extraordinary because if H2 served merely as carrier or dilution gas, as is expected, the conversion of methanol should have decreased with increasing H2 flow rate due to the decrease in the residence time of the reactant MFC MIXER Inlet Power supply Inlet HV Dielectric Pump.6 Reactor OES Oscilloscope Grounding electrode Heat tracing Outlet Methanol GC Outlet Fig. 1. Schematic of the experimental setup and the reactor structure.

3 276 Jing Zhang et al. / Chinese Journal of Catalysis 36 (215) Table 1 Effect of the H2 flow rate on methanol conversion under DBD. H2 flow rate (ml/min) X(CH3OH)/ % S(Product)/% EG C2H5OH n C3H7OH CH4 CO Others Reaction conditions: CH3OH gas flow rate = 11.1 ml/min, pressure =.1 MPa, input power = 11 W, discharge frequency = 12. khz, discharge voltage = 16.8 kv, temperature = 3 C. molecule in the discharge zone. The fact that H2 enhanced both methanol conversion and EG selectivity suggested that H2 behaves as a catalyst. It accelerates the methanol reaction selectively for EG. The conversion of methanol and selectivity for EG did not increase further at H2 flow rates beyond 8 ml/min. This is most probably due to the dilution effect of H2 at higher flow rates exceeding the catalytic enhancement. Dilution decreases the opportunities for the coupling reaction for hydroxymethyl radicals but increases the recombination reaction between H atoms and hydroxymethyl radicals. This recombinational reaction is the reverse of the C H bond dissociation of methanol The dissociation of H2 in DBD reactor and the relationship between EG yield and hydrogen atom concentration The in situ OES analysis was used to study the DBD plasma of the CH3OH/H2 mixture at different flow rates of H2 ( 1 ml/min) to investigate the catalytic mechanism of H2 in the direct synthesis of EG from methanol in a DBD reactor. As shown in Fig. 2, all the active species detected in the DBD plasma involved hydrogen. They were the excited H2 molecule (corresponding to the continuous spectrum bands in and nm, respectively), and the excited H atom (corresponding to line in nm). Specifically, the continuous spectrum in nm belongs to the radiative dissociation continuum band of H2 molecule (H2 a 3 g + H2 b 3 u + ), the continuous spectrum band in nm belongs to the Fulcher transition band of the H2 molecule (H2 d 3 u + H2 a 3 g + ) [2], while the spectral line in nm belongs to the Hα line (H 3d 2 D p 2 P ) [21]. Generally, the signals of the excited state H2 molecule and H atom were intense in the presence of H2, and this increased as the H2 flow rate increased. In the absence of H2 the signals of the excited state H2 molecule and H atom were too weak to see the bands of the Fulcher transition and radiative dissociation continuum of H2. This means that methanol alone does not produce significant amounts of hydrogen via the dissociation of the C H bond in the DBD plasma. The intensity of the Hα spectral line is related to the concentration of the ground state H atom [22]. Although the electron temperature of the plasma, or the electron energy, has an influence on the intensity of the Hα spectral line, it also influences the concentration of the ground state H atom. Based on the relationship of the intensity of the Hα spectral line with the concentration of the ground state H atom, we believed that a large number of H2 molecules dissociate into ground state H atoms during the discharge of CH3OH/H2, and that the concentration of ground state H atom increases with the flow rate of H2 at least within the range investigated. The OES observations revealed two pathways by which the dissociation of H2 molecule into a ground state H atom proceeds in a DBD plasma [23]. (1) The ground state H2 molecule is excited to the a 3 g + state via a non elastic collision with an electron, the excited H2 molecule, in an a 3 g + state, transitions to the first excited state (b 3 u + ). This is accompanied by the dissociation continuous spectrum band of nm and is followed by the spontaneous dissociation of the b 3 u + state (repulsive state) into two ground state H atoms. This pathway can be expressed as H2 (a 3 g + ) H2 (b 3 u + ) H (1s) + H (1s). (2) The ground state H2 molecule collides with an electron as above but is excited to the d 3 u + state. The d 3 u + state H2 molecule transitions to the a 3 g + state, which is followed by a transition of the a 3 g + state, which gives rise to the continuous Fulcher transition band of nm, and finally transitions H2 (d 3 u + a 3 g + ) H2(a 3 g + b 3 u + ) Hα ml/min-h2 H intensity EG yield (%) Wavelength (nm) flow rate (ml/min) Fig. 2. OES of the CH3OH/H2 non equilibrium plasma, and the correlation between the intensity of the Hα line and EG yield at different flow rates of H2. Reaction conditions: CH3OH gas flow rate = 11.1 ml/min, pressure =.1 MPa, discharge frequency = 12. khz, input power = 11 W, discharge voltage = 16.8 kv, temperature = 3 C.

4 Jing Zhang et al. / Chinese Journal of Catalysis 36 (215) to the b 3 u + state, with the associated dissociation continuous spectrum band of nm as above. This pathway can be expressed as H2 (d 3 u + ) H2 (a 3 g + ) H2 (b 3 u + ) H (1s) + H (1s). Among the active hydrogen species, the lifetime of the excited state species of H2 molecule (d 3 u +, a 3 g + and b 3 u + ) and H atom (Hα) is too short for them to participate in any chemical reactions. However, the ground state H atom has a longer lifetime and is active enough to participate in chemical reactions. As shown by Fig. 2, the intensity of the Hα spectral line (corresponding to the concentration of H atom) and EG yield increased with increasing H2 flow rate. This suggests that the H atom was closely related to the selective dissociation of the methanol C H bond. In another word, this suggests that the H atom is the catalytic active species in the direct synthesis of EG from methanol by DBD The influence of the discharge reaction conditions on the catalytic performance of H2 In addition to the flow rate of H2, the discharge conditions were further investigated in terms of discharge frequency, methanol flow rate, and reaction pressure. In situ OES was used at different reaction conditions to observe the CH3OH/H2 plasma, and the intensity of the Hα spectral line was also correlated with the yield of EG. As shown in Figs. 3, 4 and 5, when the discharge frequency increased from 7 to 19 khz, the intensity of the Hα spectral line first increased and then decreased, with a maximum at about 12. khz. However, when methanol flow rate was increased from 11.1 to 55.4 ml/min and reaction pressure increased from.7 to.275 MPa, the intensity of Hα spectral line simply decreased. These results showed that discharge frequency, methanol flow rate, and reaction pressure all influence the CH3OH/H2 DBD plasma. In each case, however, the yield of EG was seen to correlate with the intensity of Hα spectral line. This phenomenon strongly supports our judgment that H atoms are the catalytic active species for the direct synthesis of EG with CH3OH/H2 in a DBD plasma. Additionally, it also shows that the discharge frequency, methanol flow rate, and reaction pressure have an influence on the catalytic effect of H2. Apparently, these effects act through the dissociation of the H2 molecule into H atoms. The influence of discharge frequency on the dissociation of H2 could be attributed to either the shortening of the electron acceleration distance with the increase in the discharge frequency or the increase of electron density as discharge frequency increases at a fixed input power. Shortening the elec H2 (a 3 g + b 3 u + ) H2 (d 3 u + a 3 g + ) khz Wavelength (nm) Hα H intensity Frequency (khz) EG yield (%) Fig. 3. OES of the CH3OH/H2 non equilibrium plasma and the correlation between the intensity of the Hα line and EG yield at different discharge frequencies. Reaction conditions: CH3OH gas flow rate = 11.1 ml/min, H2 flow rate = 6 ml/min, pressure =.1 MPa, input power = 17 W, discharge voltage = 16.8 kv, temperature = 3 C. 2 H2 (a H2 3 g + b 3 u + ) (d 3 u + a 3 g + ) Hα ml/min Wavelength (nm) H intensity CH3OH flow rate (ml/min) EG yield (%) Fig. 4. OES of the CH3OH/H2 non equilibrium plasma and the correlation between the intensity of the Hα line and EG yield at different CH3OH gas flow rates. Reaction conditions: H2 flow rate = 8 ml/min, pressure =.1 MPa, discharge frequency = 12. khz, input power = 17 W, discharge voltage = 16.8 kv, temperature = 3 C.

5 278 Jing Zhang et al. / Chinese Journal of Catalysis 36 (215) H2 (a 3 g + b 3 u + ) H2 (d 3 u + a 3 g + ) MPa Wavelength (nm) Hα H intensity Pressure (MPa) EG yield (%) Fig. 5. OES of the CH3OH/H2 non equilibrium plasma and the correlation between the intensity of the Hα line and EG yield at different pressures. Reaction conditions: CH3OH gas flow rate = 11.1 ml/min, H2 flow rate = 8 ml/min, discharge frequency = 12. khz, input power = 17 W, discharge voltage = 16.8 kv, temperature = 3 C. tron acceleration tends to decrease the electron temperature or electron kinetic energy. This decrease of electron energy decreases the dissociation of H2, whereas the increase of electron density increases the dissociation of H2. Thus when the discharge frequency was lower than the optimum frequency of 12. khz, the dissociation of H2 was most probably limited by the electron density, and when the discharge frequency was higher than 12. khz, the dissociation of H2 was limited by the energy of electrons. The methanol flow rate and reaction pressure simply decreased the dissociation of H2 molecule over the range examined here. Both of these observations are explainable as an increase in the methanol flow rate. An increase in the methanol flow rate will decrease the chance of collision between an electron and H2 molecule, and an increase of reaction pressure will decrease electron kinetic energy The catalytic cycle of H2 in the CH3OH/H2 DBD plasma According to the experimental results, the catalytic mechanism of H2 in the direct synthesis of EG from a CH3OH/H2 DBD plasma can be described by the following steps. First, the H2 molecule acquires energy via a non elastic collision with an electron, and is excited to either the d 3 u + or a 3 g + state. These states are not stable. They transition to lower energy states (as indicated by the observation of the Fulcher transition band and dissociation continuous spectrum band), and spontaneously dissociate into ground state H atoms from the b 3 u + repulsive state. H atoms then selectively dissociated the C H bond of methanol by collision. Previous reports support the selective dissociation of the C H bond of methanol by H atoms [24 26]. These reports discuss the collision of H atoms with methanol molecules, and that methyl H atoms are preferentially abstracted as opposed to alcoholic H atoms of methanol. By colliding with a H atom, the activation energy of the C H bond dissociation reaction of methanol molecule (H + CH3OH CH2OH + H2) is reduced to kcal/mol, far lower than for the direct dissociation of C H, or dissociation of the C O and O H bonds, which are 94.57, and 1.78 kcal/mol, respectively. These data seem to show that there is no fundamental difference between the catalytic effect of H2 and that of 2 CH 3 OH HOC C OH + the well known conventional catalysts. The reduction of the activation energy was observed for the H2 catalyst in the C H bond dissociation of methanol. Therefore, the H2 catalyzed methanol to EG reaction can be described as follows. The C H bond of the methanol molecule is catalytically dissociated by hydrogen with hydrogen atoms acting as the catalytic active species, and the hydroxymethyl radicals so produced react to form EG via a homocoupling reaction. Meanwhile, the H atoms dissociated from methanol react to form H2 via a coupling reaction as well. Overall, the methanol to EG reaction is a H2 producing reaction rather than a H2 consuming one. It means that more H2 molecules will leave the DBD reactor than are introduced. Based on the above discussion, we have proposed a catalytic cycle for H2, as shown in Fig. 6. It should be pointed out that the catalytic effect of H2 will only be present in the high voltage electric field or, in other words, in its plasma state. H2 will return to its stable molecular state as soon as it leaves the discharge zone. This is perhaps an obvious trait of the H2 catalyst. Knowing that the dissociation of H2 molecule is a high energy process (13.55 kcal/mol), one may wonder why H2 is so efficient in accelerating this reaction. The answer might lie in hydrogen s ability to undergo cumulative collision excitation by low energy electrons [27]. 4. Conclusions HOC -H H2 exhibited catalytic enhancement of the dehydrogenation coupling of methanol to produce EG in a DBD reactor. Based on Catalytic e -hν (b 3 u+ ) Cycle 2H H(1s) + H(1s) Fig. 6. Schematic of H2 catalytic cycle in the EG synthesis with CH3OH/H2 plasma. e

6 Jing Zhang et al. / Chinese Journal of Catalysis 36 (215) Graphical Abstract Chin. J. Catal., 215, 36: doi: 1.116/S (14) The catalytic effect of H2 in the dehydrogenation coupling production of ethylene glycol from methanol using a dielectric barrier discharge Jing Zhang, Teng Li, Dongjiang Wang, Jialiang Zhang, Hongchen Guo * Dalian University of Technology In the non equilibrium plasma generated by a dielectric barrier discharge, H2 exhibited a catalytic effect on the dehydrogenation coupling reaction of methanol, which permitted the direct synthesis of ethylene glycol with high selectivity. 2 CH 3 OH Catalytic e e -hν H (b 3 2 u+ ) HOC -H H(1s) + H(1s) HOC C OH + Cycle 2H the close relationship between EG yield and the Hα spectral line intensity of the excited H atoms, we speculate that ground state H atom is the catalytically active hydrogen species. In the DBD plasma, the stable ground state H2 molecule underwent cumulative collision excitation with electrons and transitions from higher energy excited states to the first excited state. The spontaneous dissociation of the first excited state H2 molecule generates the catalytically active H atom. The discharge reaction conditions affect the catalytic performance of H2 by influencing the dissociation of the H2 molecule into H atoms. The catalytic performance of H2 observed here in the non equilibrium plasma may be a new opportunity for the synthesis of chemicals. References [1] Yue H R, Zhao Y J, Ma X B, Gong J L. Chem Soc Rev, 212, 41: 4218 [2] Wen C, Li F Q, Cui Y Y, Dai W L, Fan K N. Catal Today, 214, 233: 117 [3] Ma X B, Chi H W, Yue H R, Zhao Y J, Xu Y, Lü J, Wang S P, Gong J L. AIChE J, 213, 59: 253 [4] Song H Y, Jin R H, Kang M R, Chen J. Chin J Catal ( 宋河远, 靳荣华, 康美荣, 陈静. 催化学报 ), 213, 34: 135 [5] Chen Q L, Yang W M, Teng J W. Chin J Catal ( 陈庆龄, 杨为民, 腾加伟. 催化学报 ), 213, 34: 217 [6] Zhang J, Yuan Q C, Zhang J L, Li T, Guo H C. Chem Commun, 213, 49: 116 [7] Bauschlicher C W J, Langhoff S R, Walch S P. J Chem Phys, 1992, 96: 45 [8] Futamura S, Kabashima H. IEEE Trans Ind Appl, 24, 4: 1459 [9] Yan Z C, Li C, Lin W H. Int J Hydrog Energy, 29, 34: 48 [1] Burlica R, Shih K Y, Hnatiuc B, Locke B R. Ind Eng Chem Res, 211, 5: 9466 [11] Rico V J, Hueso J L, Cotrino J, Gallardo V, Sarmiento B, Brey J J, Gonzalez Elipe A R. Chem Commun, 29: 6192 [12] Rico V J, Hueso J L, Cotrino J, Gonzalez Elipe A R. J Phys Chem A, 21, 114: 49 [13] Wang B W, Zhang X, Bai H Y, Lü Y J, Hu S H. Front Chem Sci Eng, 211, 5: 29 [14] Lü Y J, Yan W J, Hu S H, Wang B W. J Fuel Chem Technol ( 吕一军, 闫文娟, 胡爽慧, 王保伟. 燃料化学学报 ), 212, 4: 698 [15] Wang Y F, You Y S, Tsai C H, Wang L C. Int J Hydrog Energy, 21, 35: 9637 [16] Lee D H, Kim T. Int J Hydrog Energy, 213, 38: 639 [17] Bundaleska N, Tsyganov D, Saavedra R, Tatarova E, Dias F M, Ferreira C M. Int J Hydrog Energy, 213, 38: 9145 [18] Li H Q, Zou J J, Zhang Y P, Liu C J. J Chem Ind Eng (China) ( 李慧青, 邹吉军, 张月萍, 刘昌俊. 化工学报 ), 24, 55: 1989 [19] Li H Q, Zou J J, Zhang Y P, Liu C J. Chem Lett, 24, 33: 744 [2] Fantz U, Schalk B, Behringer K. New J Phys, 2, 2: 71 [21] Petrovic Z L, Phelps A V. Phys Rev E, 29, 8: 1648/1 [22] Worsley M A, Bent S F, Fuller N C M, Dalton T. J Appl Phys, 26, 1: 8331/1 [23] Liu X M, Johnson P V, Malone C P, Young J A, Kanik I, Shemansky D E. Astrophys J, 21,716: 71 [24] Lendvay G, Berces T, Marta F. J Phys Chem A, 1997, 11: 1588 [25] Chuang Y Y, Radhakrishnan M L, Fast P L, Cramer C J, Truhlar D G. J Phys Chem A, 1999, 13: 4893 [26] Han Y, Wang J G, Cheng D G, Liu C J. Ind Eng Chem Res, 26, 45: 346 [27] Horacek J, Cizek M, Houfek K, Kolorenc P, Domcke W. Phys Rev A, 26, 73: 2271/1 介质阻挡放电甲醇脱氢偶联一步合成乙二醇反应中氢气的催化作用 张婧 a, 李腾 a, 王东江 a, 张家良 b a,*, 郭洪臣 a 大连理工大学化工学院催化化学与工程系, 精细化工国家重点实验室, 辽宁大连 b 大连理工大学物理与光电工程学院, 辽宁大连 摘要 : 利用原位发射光谱表征和在线色谱分析, 研究了甲醇介质阻挡放电脱氢偶联一步合成乙二醇反应中氢气的催化作用, 考察了放电频率 甲醇和氢气进料量以及反应压力的影响. 结果表明, 在介质阻挡放电产生的非平衡等离子体中, 不但能显著提高甲醇转化率, 而且能显著提高乙二醇的选择性. 在 3 C,.1 MPa, 反应器注入功率为 11 W, 放电频率为 12. khz, 甲醇气体进料量为 11.1 ml/min, 氢气进料量为 8 18 ml/min 的条件下, 甲醇转化率接近 3%, 乙二醇选择性大于 75%. 乙二醇收率与激发态氢

7 28 Jing Zhang et al. / Chinese Journal of Catalysis 36 (215) 原子的 H α 谱线强度之间存在同增同减关系. 由此推测, 氢原子是起催化作用的活性氢物种. 活性氢物种的生成途径是 : 基态氢分子通过与电子碰撞变成激发态, 激发态氢分子通过第一激发态氢自动解离为基态氢原子. 放电反应条件通过影响氢分子解离来影响氢气的催化作用. 氢气在非平衡等离子体中显示的催化作用有可能为开辟新的化学合成途径提供重要机遇. 关键词 : 介质阻挡放电等离子体 ; 氢气催化作用 ; 乙二醇合成 ; 氢原子 ; 碳 - 氢键活化 收稿日期 : 接受日期 : 出版日期 : * 通讯联系人. 电话 / 传真 : (411) ; 电子信箱 : hongchenguo@163.com 本文的英文电子版由 Elsevier 出版社在 ScienceDirect 上出版 ( 1. 前言乙二醇 (EG) 是最简单的二元醇, 用途十分广泛 [1] ; 目前, 普遍采用环氧乙烷直接水合法生产. 然而, 随着石油资源的日益枯竭, 近年来煤制乙二醇技术广受关注 [2,3]. 煤制乙二醇须由合成气制草酸二酯中间体, 其工艺过程包括合成气氧化脱氢 氧化酯化 CO 偶联和草酸二酯加氢等单元. 目前, 虽然不少研究机构都在开发该工艺, 但存在流程长 投资大和工艺技术复杂等问题 [4,5]. 最近, 我们报道了在 存在下, 采用甲醇介质阻挡放电一步合成乙二醇的研究工作. 在常压条件下, 甲醇转化率和乙二醇选择性分别达 15.8% 和 71.5% [6]. 该方法的基本原理是 : 甲醇分子在介质阻挡放电产生的非平衡等离子体的作用下解离一个 C H 键, 生成羟甲基自由基 (C OH) 和氢原子 ; 两个羟甲基自由基 (C OH) 偶联生成乙二醇, 同时两个氢原子复合生成. 由于 也是高附加值产物, 所以上述甲醇脱氢偶联合成乙二醇反应在理论上原子经济性可达 1%, 因而在应用上更具吸引力. 不仅如此, 由于常规催化法在选择性活化甲醇分子的 C H 键方面难度很大 ( 键能顺序 : O H > C H > C O) [7], 因此非平衡等离子体所表现出的选择性解离甲醇 C H 键现象, 对于化学键的定向活化研究会有启发作用. 我们已经发现, 在甲醇介质阻挡放电一步合成乙二醇的反应过程中起到了催化剂的作用. 本文借助于发射光谱原位表征和在线色谱分析手段, 重点研究了 在不同放电反应条件下的解离活化及其与甲醇脱氢偶联反应的关系. 在此基础上讨论了 对甲醇一步合成乙二醇反应的分子催化作用机制. 2. 实验部分装置流程和反应器如图 1 所示. 实验时, 先将 ( 纯度为 99.99%, 大连气体有限公司 ) 通过质量流量计 (D7-19B, 北京七星华创电子股份有限公司 ) 精确控制流量后进入汽化混合器, 然后进入反应器. 待反应系统被 充分置换后, 将 CH 3 OH 液体 ( 分析纯, 国药集团化学试剂有限公司 ) 经高压恒流泵 (P23, 大连依利特分析仪器 有限公司 ) 精确控制流量后进入汽化混合器. 汽化后的 CH 3 OH 在汽化混合器中与 充分混合后, 经保温管线进入等离子体反应器. 打开高压电源 (CTP-2K 低温等离子电源, 南京苏曼公司 ), 调节电源电压产生介质阻挡放电. 放电电压 电流及功率等参数通过示波器 (DPO312 数字示波器, 美国 Tektronix 公司 ) 进行测量. 在等离子体区的解离情况通过原位发射光谱仪 (SP2758 型, 美国 Princeton 仪器公司, 3 G/mm 光栅, 1. s 曝光时间 ) 进行监测. 从反应器出口流出的反应产物经保温管线进入双检测器在线色谱进行组成分析 (SP-689 型气相色谱仪, 山东鲁南瑞虹化工仪器有限公司, 其 FID 用于测定 CH 3 OH, HOC C OH, C 2 H 5 OH, C 3 H 7 OH 和 CH 4 等, TCD 用于测定 CH 4, CO, O 等 ). CH 3 OH 转化率及各种产物的选择性由下述公式计算 : X(CH 3 OH) = (n (CH 3 OH) n(ch 3 OH))/n (CH 3 OH) S i = n i (CH 3 OH)/(n (CH 3 OH) n(ch 3 OH)) 其中, n (CH 3 OH) 和 n(ch 3 OH) 分别为反应前及反应后 CH 3 OH 的摩尔数, n i (CH 3 OH) 为生成产物 i 消耗的 CH 3 OH 摩尔数. 3. 结果与讨论 3.1. 在介质阻挡放电条件下 对甲醇合成乙二醇反应的促进作用用等离子体方法转化甲醇已有许多研究报道, 但目的产物主要是 或合成气 [8 17] [18,19]. 刘昌俊课题组曾在甲醇电晕放电制氢的研究中发现有微量乙二醇生成. 我们也发现, 用甲醇的等离子体直接合成乙二醇需要具备两个关键性的前提条件 : 一是等离子体的放电强度要适中 ; 二是要用 和甲醇共进料 [6]. 本文用一种放电强度适中的介质阻挡放电反应器, 在甲醇进料量不变的情况下, 考察了不同 进料流量对介质阻挡放电转化甲醇的影响. 如表 1 所示, 当不加入 时, 甲醇转化率仅为 9.6%, 放电主产物是 CO 和 CH 4, 其选择性分别为 55.2% 和 16.6%. 此时乙二醇选择性仅为 8.% 左右. 当 流量为 8 ml/min 时, 甲醇转化率可达 3.5%, 乙二醇选择性高达 75.4%, 比没有 时分别增加了近 2.2 和 8.4 倍. 由此可见,

8 Jing Zhang et al. / Chinese Journal of Catalysis 36 (215) 在介质阻挡放电转化甲醇时, 加入 不但能提高甲醇转化率, 而且能选择性促进甲醇 C H 键的解离, 从而提高乙二醇选择性. 该现象非同寻常. 这是因为, 如果 仅仅起载气和稀释气作用的话, 那么, 当乙二醇的选择性随着 流量增加而提高时, 甲醇转化率很可能会因为反应物分子在放电区的停留时间缩短而下降. 能够同时提高甲醇转化率及乙二醇选择性的现象意味着 不但加快了甲醇反应的速率, 而且使反应按照生成乙二醇产物的特定途径进行, 即 表现出了对甲醇直接生成乙二醇反应的催化作用. 需要说明的是, 继续增加 流量, 则甲醇转化率和乙二醇选择性不再增加. 这可能是由于大量 的稀释作用抑制了羟甲基自由基的偶联反应, 而促进了氢原子和羟甲基自由基之间的偶合反应, 即甲醇 C H 键解离反应的逆过程 在介质阻挡放电中的解离及氢原子浓度与乙二醇产率的关系为了研究 在甲醇介质阻挡放电直接合成乙二醇反应过程中的催化作用机制, 我们用发射光谱对不同 流量 ( 1 ml/min) 下的 CH 3 OH/ 介质阻挡放电等离子体进行了原位表征. 如图 2 所示, 在 CH 3 OH/ 介质阻挡放电等离子体中, 能检测到的活性物种都是氢的活性物种, 包括激发态的氢分子 ( 子 (38 55 nm 和 nm 两处连续谱带 ) 和激发态的氢原子 ( 子 (656.3 nm 处谱线 ). 其中, nm 处的连续谱带归属为氢分子的辐射解离带 ( 带 ( a 3 + g b 3 + u ); nm 处的连续谱带归属为氢分子的 Fulcher 跃迁带 ( 带 ( d 3 + u a 3 + g ) [2] ; nm 谱线归属为 H α 谱线 ( 线 (H 3d 2 D H 2p 2 P ) [21]. 一般来说, 激发态氢分子和原子的 OES 信号在加入 的情况下比较强, 而且随着 流量的增加而增强. 当不加入 时, 甲醇等离子体发射光谱的 H α 谱线非常弱, 且几乎看不到氢分子 Fulcher 跃迁带和辐射解离带. 这反映出单纯的甲醇等离子体不能通过 C H 键解离反应产生出大量. 从物理学上讲, H α 谱线的强度与等离子体中基态氢原子浓度有关 [22]. 尽管等离子体中的电子温度, 或者说电子能量, 对 H α 谱线的强度有影响, 但它同时对基态氢原子的浓度也有影响. 基于 H α 谱线强度与等离子体中基态氢原子浓度的关系, 本文认为, 在 CH 3 OH/ 介质阻挡放电等离子体中氢分子被大量解离为氢原子, 在实验范围内等离子体中氢原子浓度随着 流量增加而增加. OES 诊断揭示出以下两种氢分子解离为氢原子的途径 [23]. 一是基态氢分子通过与电子碰撞获得能量变成 a 3 + g 态, 接着 a 3 + g 态氢分子跃迁到第一激发态 (b 3 + u ), 同时在 nm 范围产生连续的辐射解离带, 然后再由 b 3 + u 态氢分子 ( 排斥态 ) 自发解离为基态氢原子. 这条途径可以表示为 : (a 3 + g ) (b 3 + u ) H (1s) + H (1s). 二是基态氢分子通过与电子碰撞获得能量变成能量更高的 d 3 + u 激发态, 接着 d 3 + u 态氢分子先跃迁到 a 3 + g 态, 同时在 nm 范围内产生连续的 Fulcher 跃迁带, 然后 a 3 + g 态氢分子跃迁到 b 3 + u 态, 伴随产生辐射解离带, 最后由 b 3 + u 态氢分子自发解离为基态氢原子. 这条路径可以表示为 : (d 3 + u ) (a 3 + g ) (b 3 + u ) H (1s) + H (1s)]. 在各种活泼氢物种中, 激发态氢分子 (d 3 + u, a 3 + g 及 b 3 + u ) 和激发态氢原子 (H α ) 寿命都很短, 在发生化学反应之前就退激发了. 然而, 基态氢原子寿命较长, 且具有较高的化学反应活性, 能够参与化学反应. 从图 2 可以看出, 在 CH 3 OH/ 介质阻挡放电等离子体中, H α 谱线强度 ( 对应于氢原子的浓度 ) 与乙二醇的收率都随着 流量的增加而增大, 且变化趋势非常相似. 这说明等离子体中的氢原子与甲醇选择性解离 C H 键, 进而偶联生成乙二醇的反应关系密切. 因此本文认为, H 原子应该是甲醇介质阻挡放电直接合成乙二醇反应的催化活性物种 放电反应条件对 催化性能的影响除了 流量之外, 本文还考察了放电频率 甲醇进料流量和反应压力的影响, 原位监测了不同条件下 CH 3 OH/ 等离子体的发射光谱, 并将得到的 H α 谱线强度 ( 对应于氢原子浓度 ) 与乙二醇收率进行了关联. 如图 3 5 所示, 随着放电频率的提高, 激发态氢原子的 H α 谱线强度先增加后减小, 在放电频率为 12. khz 附近达到最大值 ; 另一方面, 随着甲醇进料流量从 11.1 增至 55.4 ml/min, 以及反应压力从.7 增至.275 MPa, H α 谱线强度都呈单调降低变化. 由此可见, 放电频率 甲醇进料量和反应压力都对 CH 3 OH/ 等离子体有显著影响. 然而, 在上述条件变化过程中, 乙二醇收率的变化始终与激发态氢原子的 H α 谱线强度变化趋势保持一致. 这有力支持了前面的判断, 即氢原子是甲醇介质阻挡放电直接合成乙二醇反应的催化活性物种. 很显然, 放电频率 甲醇进料流量和反应压力对甲醇介质阻挡放电过程中 分子催化作用的影响是通过影响 分子解离来实现的. 放电频率对 解离的影响可归结于两个方面 : 一方面, 增加放电频率缩短了等离子体中的电子加速距离, 从而导致电子温度, 或者说电子动能的降低 ; 另一方面, 随着放电频率的增加, 等离子体中的电子密度将增加, 这是

9 282 Jing Zhang et al. / Chinese Journal of Catalysis 36 (215) 在放电功率一定的情况下等离子体中电子能量降低导致的必然结果. 电子动能降低不利于 解离, 但电子密度增加有利于 解离. 因此不难理解, 当放电频率低于最佳值 12. khz 时, 等离子体中电子密度低可能是制约 解离的主要因素, 而当放电频率高于最佳值 12. khz 时, 等离子体中电子能量低可能是制约 解离的主要因素. 与放电频率的影响不同, 在实验范围内甲醇进料流量和反应压力对 分子解离的影响都是单调变化的. 这是因为增大甲醇进料流量会降低电子与 分子的碰撞机会, 而增加反应压力会降低电子动能 在介质阻挡放电条件下甲醇合成乙二醇中 的催化循环根据上述实验结果, 我们对 的催化作用机制归纳如下. 首先, 分子通过与电子碰撞获得能量, 变成激发态 分子 (d 3 + u, a 3 + g ). 激发态的氢分子不稳定, 向低激发态跃迁 ( 产生 Fulcher 跃迁带和辐射解离带 ), 并通过 b 3 + u 排斥态自发解离为基态氢原子. 然后, 氢原子与甲醇分子发生碰撞反应导致 C H 键选择性解离. 据报道, 当 H 原子与甲醇分子碰撞反应时, 优先夺取醇甲基上的 H 原子而不是醇羟基上的 H 原子 [24 26]. 由氢原子与甲醇分子发生碰撞而导致的甲醇分子 C H 键解离反应 (H + CH 3 OH C OH + ) 活化能很低, 只有 kcal/mol, 远低于甲醇分子直接解离 C H 键的活化能 (94.57 kcal/mol), 也远低于甲醇分子直接解离 C O 键 (81.51 kcal/mol) 和 O H 键 (1.78 kcal/mol) 等反应的活化能. 由此可见, 氢分子催化作用的本质与常规催化剂相同, 都是降低目标反应的活化能. 乙二醇由两个羟甲基自由基偶 联得到. 上述反应的净结果是 : 甲醇在 的催化作用下 ( 氢原子是催化活性物种 ) 发生 C H 解离反应, 生成的羟甲基自由基偶联成为乙二醇, 生成的氢原子则复合成为. 作为催化剂引入反应器的 在等离子体区参与了化学反应, 但最后仍然以 的形式离开等离子体区. 由于甲醇脱氢偶联是一个产氢过程, 因此 作为催化剂在反应过程中不会被消耗. 本文为此设计了 的催化循环, 如图 6 所示. 应该指出的是, 本文所说的 催化作用只存在于高压电场或者说 的等离子体中. 一旦离开放电区, 就会回复到我们所熟悉的稳定状态. 这可能是 分子催化作用的特别之处. 另外, 由于 分子的解离并不容易 ( 键能为 kcal/mol), 那么在 CH 3 OH/ 等离子体中为什么氢分子会如此高效率地加速甲醇生成乙二醇反应呢? 这可能与氢分子的特性有关 : 氢分子可以被能量较低的电子经多次碰撞导致累积激发 [27]. 4. 结论在甲醇介质阻挡放电脱氢偶联一步合成乙二醇过程中, 表现出明显的催化作用. 根据乙二醇收率与激发态氢原子的 H α 谱线强度的密切关系推测, 氢原子应该是起催化作用的活性氢物种. 在介质阻挡放电等离子体中, 基态氢分子通过与电子碰撞获得能量变成激发态, 激发态氢分子通过退激发跃迁到第一激发态后, 自动解离为具有催化作用的基态氢原子. 放电反应条件通过影响氢分子解离即活性氢物种的生成影响 的催化作用. 在非平衡等离子体中显露出来的催化作用有可能为开辟新的化学合成途径提供重要机遇.