hinese Journal of atalysis 35 (2014) 159 167 催化学报 2014 年第 35 卷第 2 期 www.chxb.cn available at www.sciencedirect.com journal homepage: www.elsevier.com/locate/chnjc Article The effect of doping transition metal oxides on copper manganese oxides for the catalytic oxidation of Lina ai a, Zhenhao Hu a, Peter Branton b, Wencui Li a, * a State Key Laboratory of Fine hemicals, Department of hemical Engineering, Dalian University of Technology, Dalian 116024, Liaoning, hina b Group Research and Development, British American Tobacco, Regents Park Road, Millbrook, Southampton S15 8TL, UK A R T I L E I N F Article history: Received 27 July 2013 Accepted 2 September 2013 Published 20 February 2014 Keywords: opper manganese oxide Transition metal oxide Doping arbon monoxide oxidation Diffuse reflectance infrared Fourier transform spectroscopy A B S T R A T A series of copper manganese oxides doped with transition metal oxides were prepared by co precipitation using copper acetate and manganese acetate as precursors, ammonium bicarbonate as precipitant, and metal nitrates as dopants. The catalysts were characterized by N2 adsorption desorption, X ray powder diffraction, temperature programmed reduction, and in situ diffuse reflectance infrared Fourier transform spectroscopy. The results showed that doping transition metal oxides into copper manganese oxides can modify the adsorption ability of the catalyst and thus affect the catalytic oxidation of. 2014, Dalian Institute of hemical Physics, hinese Academy of Sciences. Published by Elsevier B.V. All rights reserved. 1. Introduction atalytic oxidation of is one of the most effective methods of removal at low temperatures and has received considerable attention because of its many applications in industry and environmental fields. These include personal respiratory protective devices, 2 laser gas generation, proton exchange membrane fuel cells, and automobile emission controls [1 6]. ompared with noble metal catalysts, non noble metal oxide catalysts have the advantages of high availability and low cost. Among them, hopcalite based on manganese copper mixed oxide, a well known catalyst for oxidation, has attracted attention because of its low price and relatively high catalytic activity. However, it has poor low temperature catalytic activity and moisture resistance [7 9]. More recently, many attempts have been made to improve the catalytic activity of copper manganese oxides for oxidation, in particular by optimizing the preparation technologies and the improvement of preparation methods [7 20]. Among these methods, doping a transition metal oxide into the copper manganese oxide catalysts could tune the oxygen mobility and the reduction ability of the catalysts, thus improving the catalytic activity. It was found that after doping a small amount of e2 into copper manganese oxide, a highly dispersed e2 phase could prevent sintering and aggregating of the catalyst. In addition, the reducibility was enhanced, the particle size was decreased, and the formation of the active sites for the oxidation of was improved significantly. Therefore, the activity of this rare earth promoted catalyst was enhanced remarkably [7]. It was also reported that copper manganese oxide catalysts, prepared by co precipitation, had improved 2 availability in the lattice and an enhanced surface area. By adding low levels * orresponding author. Tel/Fax: +86 411 84986355; E mail: wencuili@dlut.edu.cn This work was supported by the Fudamental Resarch Funds for the entral Universities (DUT12ZD218) and the Specialized Research Fund for the Doctoral Program of Higher Education (20100041110017). DI: 10.1016/S1872 2067(12)60699 8 http://www.sciencedirect.com/science/journal/18722067 hin. J. atal., Vol. 35, No. 2, February 2014
160 Lina ai et al. / hinese Journal of atalysis 35 (2014) 159 167 (~1.0 wt%) of o, these materials can display much higher activity for oxidation compared with the current commercial copper manganese oxide catalysts at ambient conditions [8]. In our previous work [20], the combination effect of precipitant and precursor in the preparation has been studied. The precipitant shows the greatest influence on the crystalline phases of the catalyst while the precursor shows a greater effect on the number of catalytic active sites, both of which are directly related to the oxidation activity. In the present study, to further improve the catalytic activity of copper manganese oxide, we have prepared copper manganese oxide catalysts via co precipitation by doping with transition metal oxides with (Fex or e2) or without (Zn) oxygen storage capacity. This study focuses on the influences of doping different transition metal oxide on adsorption and the resulting performance of the copper manganese oxide catalysts. 2. Experimental 2.1. atalyst preparation All chemicals used in this study were of analytical grade and used without further purification. The copper manganese oxide catalysts were prepared by co precipitation using NH4H3 as the precipitant, the acetates of the copper and manganese as the precursors, and Fe(N3)3, e(n3)3, and Zn(N3)2 as dopant. The typical procedure to synthesize the catalysts was as follows. A precipitant (30 mmol) was dissolved in deionized water (30 ml) with an initial ph value of ca. 8. The precursors (7.5 mmol) were mixed with deionized water (30 ml) with a 1/2 molar ratio of copper to manganese species. The mixed precursor solution was then added to the precipitant solution at 298 K with vigorous stirring. The resultant suspensions were aged for 30 min with continued vigorous stirring at 25. Finally the precipitate was filtered, washed with deionized water and anhydrous alcohol, dried in air at 50 for 24 h, and then calcination at 300 for 2 h (denoted as umx M, M = Fe, Zn, e) to obtain the final catalysts. The content of the doping transition metal oxide was fixed at 5 wt%. For comparison, the pure copper oxide and manganese oxide catalysts were prepared separately using the acetate as the precursor and NaH as the precipitant, and keeping the other synthesis and after treatment conditions the same as the copper manganese oxides. The corresponding samples were named u and Mnx, respectively. 2.2. atalyst characterization X ray diffraction patterns (XRD) were obtained with a Rigaku D/MAX 2400 diffractometer using u Kα radiation (40 kv, 100 ma, λ = 1.54056 Å). The textural characterizations of the samples were performed by nitrogen sorption at 196 using a Micromeritics Instrument orporation Tristar 3000 device. Approximately 200 mg of the samples were heated to 200 under vacuum for 4 h to remove all adsorbed species. The surface area (SBET) and pore size distribution were calculated using the BET method and BJH method, respectively. The total pore volume (Vtotal) was estimated from the amount adsorbed at a relative pressure of 0.99. The micropore volume was determined using the t plot method. The morphologies of the catalysts were characterized with a FEI Quanta 450 instrument microscope equipped with a cooled energy dispersive X ray (EDX) spectrometer from xford Instruments for point resolved elemental analysis. Hydrogen temperature programmed reduction (H2 TPR) was performed by passing 8% H2/Ar (50 ml/min) over a 20 mg sample (40 60 mesh size) at a heating rate of 10 /min to 900. Before H2 TPR, the samples were pretreated with Ar at 200 for 1 h. The system was then cooled to ambient temperature under Ar. The amount of hydrogen consumed (H2 cons.) by each catalyst was calculated from the peak area of the H2 TPR profile. In situ diffuse reflectance infrared Fourier transform spectra (DRIFTS) were recorded using a Nicolet 6700 FT IR spectrometer at a resolution of 4 cm 1 from 4000 to 640 cm 1. Self supporting disks were prepared from the sample powders and treated directly in the IR cell. The catalysts were connected to a vacuum adsorption apparatus with a residual pressure below 10 3 Pa. Prior to adsorption (5 vol% and N2 in balance), the catalysts were evacuated for 30 min at 200. After flushing with pure He for 10 min, the spectrum was collected again. 2.3. atalytic test The activity of the copper manganese oxide catalysts for oxidation was measured in a quartz tubular fixed bed flow reactor at atmospheric pressure using 200 mg of catalyst (40 60 mesh). The standard composition of the feed gas was 1%, 20% 2, and 79% N2 with a space velocity (SV) of 20000 ml/(h gcat). The temperature was ramped to the final temperature at a rate of 1 /min. The concentrations of were analyzed at the outlet of the reactor by a Techcomp G 7890T gas chromatograph equipped with a thermal conductivity detector. Temperatures for 100% conversion of (T100%) and 50% conversion of (T50%) were used to evaluate the activity of the catalysts. Long term stability test of the umnx Fe sample was conducted under atmosphere 30, and SV = 20000 ml/(h gcat). 3. Results and discussion 3.1. Structure analyses of copper manganese oxides XRD analysis was used to determine the final phase of the copper manganese oxide catalysts doped with different metal oxides after heat treatment at 300 in static air for 2 h (Fig. 1). It can be seen that the main crystal phase composition of the catalysts are Mn23, u, and u0.139mn0.8612 with lower crystallinity. The crystal phase composition of the catalysts did not change significantly by doping with transition metal oxides, indicating that this addition does not significantly alter the bulk composition of the catalyst. Moreover, the characteristic diffraction peaks of the doped transition metal or any derivative did not appear due to low doping contents (below 5 wt%) or
Lina ai et al. / hinese Journal of atalysis 35 (2014) 159 167 161 u 0.139Mn 0.861 2 e 2 Mn 2 3 u Intensity 10 20 30 40 50 60 70 80 2 /( o ) Fig. 1. XRD patterns of the copper manganese oxide catalysts. umnx; umnx Fe; umnx Zn; umnx e. highly dispersed or into the copper manganese oxide crystals lattice forming a solid solution [21]. Figure 2 shows the N2 adsorption desorption isotherms and the corresponding pore size distributions of the copper manganese oxide catalysts doped with different transition metal oxides. All the samples showed the typical mesoporous structure, which illustrated IV type isotherms with a hysteresis loop at relative pressures (p/p0) of 0.4 0.8. The pore sizes of umnx, umnx Zn, and umnx e catalysts were mainly concentrated at 4.5 nm, and umnx Fe at 2.7 nm. The specific surface area and the total pore volume of umnx, umnx Fe, and umnx Zn were similar. But these of umnx e were decreased slightly. These characterizations demonstrated that the specific surface area, pore structure, and phase structure of copper and manganese oxide catalysts were only slightly affected by doping with transition metal oxides. Figure 3 shows the SEM images and corresponding EDX elemental mapping images of the umnx Fe and umnx e Fig. 3. SEM images of the catalysts. (a) umnx Fe; (b) umnx e; (c) Mapped results of umnx Fe; (d) Mapped results of umnx e. catalysts. The obtained catalysts showed a sphere morphology with diameters of 0.5 1.5 μm. The elemental mapping images showed that the doped catalysts contained the expected elements (Fe, e). The Fe and e species were evenly distributed in the obtained catalysts, revealing that the transition metal oxide could be doped into copper manganese oxide during the co precipitation process. 3.2. atalytic performance of the copper manganese oxides for oxidation Figure 4 displays the catalytic performance of the obtained catalysts for oxidation, and the corresponding activity is listed in Table 1. As shown in Fig. 4(a), the catalytic activity of all the catalysts increased as the reaction temperature and the activity towards oxidation strongly depends on the doped transition metal oxide, in the order umnx Fe > umnx e > 360 (a) (b) 300 0.16 Vads (cm 3 /g, STP) 240 180 120 dv/dd 0.08 60 0 0.0 0.2 0.4 0.6 0.8 1.0 Relative pressure (p/p 0) 1 10 100 Pore diameter (nm) Fig. 2. N2 adsorption desorption isotherms (a) and pore size distributions (b) of the copper manganese oxide catalysts. umnx; umnx Fe; umnx Zn; umnx e. urves,, and in (a) offset vertically by 80, 160, and 240 cm 3 /g, STP, respectively.
162 Lina ai et al. / hinese Journal of atalysis 35 (2014) 159 167 conversion (%) 100 80 60 40 20 (a) conversion (%) 100 80 60 40 20 (b) 0 30 50 70 90 110 130 150 Temperature ( o ) 0 0 20 40 60 80 100 120 Time (min) Fig. 4. (a) atalytic activity of copper manganese oxides umnx, umnx Fe, umnx Zn, and umnx e for oxidation. (b) Longterm stability of umnx Fe at 30. umnx Zn > umnx. The umnx catalyst without doping showed 44% of conversion at 30 and achieved complete conversion (T100%) at 140. Although there was no obvious improvement on the catalytic activity of umnx Zn catalyst at 30, the complete conversion temperature of shifted to 100, exhibiting higher catalytic activity than that of umnx catalyst. The improvement in the catalytic activity of umnx e sample was due to the ability of the included e to enhance the oxygen storage capacity and oxygen mobility in the catalysts [5]. The conversion over umnx Fe catalyst at 30 was raised by 35% compared with the umnx catalyst, and this catalyst was able to achieve a complete conversion at 60. Table 1 Textural parameters and catalytic activity of the copper manganese oxide catalysts. Sample SBET a (m 2 /g) Vtotal b (cm 3 /g) 30 c (%) T100% d ( ) umnx 196 0.23 44 140 umnx Fe 199 0.24 79 60 umnx Zn 185 0.22 37 100 umnx e 145 0.16 54 80 a Specific surface area calculated by the BET method. b Total pore volume calculated by the amount adsorbed at a relative pressure of 0.99. c conversion at 30. d Temperature at complete conversion of to 2. This remarkable improvement in the catalytic activity could be attributed to the increase in defects, formed by doping with ferric oxide, which improves the adsorption of the reactants, and 2 [8]. Furthermore, a long term stability test (Fig. 4(b)) of a umnx Fe sample was conducted under reaction atmosphere at 30 with a space rate of 20000 ml/(h gcat). The conversion gradually decreased with time and reached about 65% after 120 min. 3.3. Reduction properties of copper manganese oxides catalysts and doped catalysts Figure 5 gives H2 TPR profiles of fresh calcined samples. The obtained catalysts showed three asymmetric reduction peaks except for the umnx Fe catalyst (four reduction peaks). The corresponding fitted H2 TPR profiles were also shown in Fig. 5. The fitted data and the H2 consumption data calculated from the integration of the corresponding peak areas are listed in Table 2. For umnx, the reduction peaks at 141.9 and 164.9 could be assigned to the step reductions of u u2 u, which were significantly lower than the reported in the literature [22]. This illustrated that the reducibility of u species could be improved by the interaction between u and Mn species. Another two fitted peaks at higher temperature could be attributed to the step reduction of Mn23 Mn34 Mn [15]. (a) (b) (c) (d) Intensity 50 150 250 350 50 150 250 350 50 150 250 350 50 150 250 350 Temperature ( o ) Temperature ( o ) Temperature ( o ) Temperature ( o ) Fig. 5. H2 TPR profiles of the copper manganese oxide catalysts. (a) umnx; (b) umnx Fe; (c) umnx Zn; (d) umnx e.
Table 2 H2 TPR fitted results of the copper manganese oxide catalysts. Lina ai et al. / hinese Journal of atalysis 35 (2014) 159 167 163 Sample u Mnx thers T/ H2 consumption (cm 3 /g) T/ H2 consumption (cm 3 /g) T/ H2 consumption (cm 3 /g) umnx 141.9, 164.9 103.8 197.1, 247.6, 67.8 umnx Fe 148.4, 166.8 100.4 197.8, 260.8 56.2 336.1 12.5 umnx Zn 133.6, 150.0 59.5 199.0, 264.8 68.3 umnx e 146.6, 169.9 86.0 202.4, 290.5 68.5 The TPR patterns of umnx Zn and umnx e were similar to umnx. However, there are four peaks over umnx Fe. The fourth reduction peak could be assigned to the reduction of Fex species [23,24]. Based on the integrated analysis of Figure 4 and Table 2, the reduction peak of the umnx Zn catalyst shifted to lower temperature, demonstrating its enhanced reducibility. Moreover, the reduction peaks of umnx Fe and umnx e shifted to higher temperature and the hydrogen consumption amount decreased, confirming the decrease in the catalyst s reducibility. 3.4. adsorption ability of the copper manganese oxide catalysts To further demonstrate the impact of doping transition metal oxide on the copper manganese oxide catalysts, in situ DRIFTS of the obtained samples was performed to study the adsorption behavior on the catalysts surface. Pure u catalyst and Mnx catalyst were also studied for comparison. As shown in Fig. 6, Mnx showed three weak peaks, which can be assigned to weakly adsorbed on the Mn species, while u exhibited strong adsorption of centered at 2171, 2121, and 2055 cm 1. The strong at 2121 cm 1 band was attributed to the linear adsorption of on, which is the typical absorption mode on u catalysts [25,26]. The weaker bands at 2055 and 2171 cm 1 were assigned to the linearly adsorbed on u 0 [27] and u 2+ [28], respectively. The liner adsorption peak on copper manganese oxide catalysts is red shifted to 2110 cm 1. This can be attributed to the strong interaction between copper oxide and manganese oxide, which in turn weakens the bond. After doping with a transition metal oxide, the adsorption peak of the catalysts does not change. This demonstrates that doping with transition metal oxide does not weaken the interaction between copper and manganese oxides. However, the intensity of the adsorption peak is enhanced, indicating that the number adsorption sites increases. The intensity of the vibrational absorption peak at 2110 cm 1 was ordered umnx Fe > umnx e > umnx Zn umnx. ombined with the H2 TPR results, doping with Fe or e oxides could promote the interaction between the copper and manganese oxides, and this enhanced the u 2+ reduction. By increasing the content, the adsorption capacity of the catalyst is increased significantly. The adsorption capacity of the zinc oxide catalyst was the same as the copper manganese catalyst, which is consistent with the catalytic activity data. Figure 7 shows the in situ DRIFTS spectra of catalysts adsorbed at 25 at different time intervals. The umnx Fe Absorbance 0.05 (6) (5) 2172 2110 2121 2055 2300 2250 2200 2150 2100 2050 2000 Wavenumber (cm 1 ) Fig. 6. DRIFTS spectra of adsorbed on the catalysts after 20 min at 25. Mnx; u; umnx; umnx Zn; (5) umnx e; (6) umnx Fe. Absorbance 0.04 0.02 10 min 8 min 5 min 2 min (a) 0.15 0.10 0.05 10 min 8 min 5 min 2 min 0.5 min (b) 0.04 0.02 10 min 8 min 5 min 2 min (c) 0.09 0.06 0.03 10 min 8 min 5 min 2 min 0.5 min (d) 2300 2200 2100 2000 2300 2200 2100 2000 2300 2200 2100 2000 2300 2200 2100 2000 Wavenumber (cm 1 ) Wavenumber (cm 1 ) Wavenumber (cm 1 ) Wavenumber (cm 1 ) Fig. 7. DRIFTS spectra of the catalysts adsorbed at 25 at different time intervals. (a) umnx; (b) umnx Fe; (c) umnx Zn; (d) umnx e.
164 Lina ai et al. / hinese Journal of atalysis 35 (2014) 159 167 Graphical Abstract hin. J. atal., 2014, 35: 159 167 doi: 10.1016/S1872 2067(12)60699 8 The effect of doping transition metal oxides on copper manganese oxides for the catalytic oxidation of Lina ai, Zhenhao Hu, Peter Branton, Wencui Li * Dalian University of Technology, hina; British American Tobacco, UK DRIFTS Absorbance 0.05 umn x -Fe 2172 2110 2055 umn x Doping with specific transition metal oxides can enhance the adsorption on in a copper manganese oxide catalyst and benefits its catalytic oxidation. Wavenumber (cm -1 ) catalysts show a strong adsorption of occurred at 2 min while the adsorption of was complete at 5 min; the other catalyst generally required about 8 10 min before reaching saturation. This implies that umnx Fe can adsorb quickly during the reaction, thereby increasing the catalytic activity. 4. onclusions opper manganese oxide catalysts doped with Fe, e, and Zn oxides were prepared via co precipitation. The addition of iron oxide and cerium oxide increase oxygen storage capacity of the copper manganese oxide catalysts, and can significantly improve adsorption capacity, and thus the performance of oxidation. Zinc oxide with no oxygen storage capacity, can increase the reduction performance of copper manganese oxide catalyst and improve the catalytic activity for oxidation. atalytic tests showed that the umnx Fe exhibited excellent catalytic performance with a 30% 40% enhancement in the conversion at 30 and total oxidation of can be achieved at 60. References [1] Feng Y F, Wang L, Zhang Y H, GuoY, Guo Y L, Lu G Z. hin J atal ( 封雅芬, 王丽, 张艳慧, 郭耘, 郭杨龙, 卢冠忠. 催化学报 ), 2013, 34: 923 [2] Frey K, Iablokov V, Sáfrán G, sán J, Sajó I, Szukiewicz R, henakin S, Kruse N. J atal, 2012, 287: 30 [3] Yu Y B, Takei T, hashi H, He H, Zhang X L, Haruta M. J atal, 2009, 267: 121 [4] Biabani A, Rezaei M, Fattah Z. J Nat Gas hem, 2012, 21: 415 [5] Zhang R R, Ren L H, Lu A H, Li W. atal ommun, 2011, 13: 18 [6] Zhang H P, Liu H. J Energy hem, 2013, 22: 98 [7] Zhang X B, Ma K Y, Zhang L H, Yong G P, Dai Y, Liu S M. hin J hem Phys, 2011, 24: 97 [8] Jones, Taylor S H, Burrows A, rudace M J, Kiely J, Hutchings G J. hem ommun, 2008: 1707 [9] Li X Q, Xu J, Zhou L P, Gao J, Wang F, hen. hin J atal ( 李晓强, 徐杰, 周利鹏, 高进, 王峰, 陈晨. 催化学报 ), 2006, 5: 369 [10] Morales M R, Barbero B P, adús L E. Appl atal B, 2006, 67: 229 [11] Mirzaei A A, Shaterian H R, Joyner R W, Stockenhuber M, Taylor S H, Hutchings G J. atal ommun, 2003, 4: 17 [12] Mirzaei A A, Shaterian H R, Habibi M, Hutchings G J, Taylor S H. Appl atal A, 2003, 253: 499 [13] Hutchings G J, Mirzaei A A, Joyner R W, Siddiqui M R H, Taylor S H. Appl atal A, 1998, 166: 143 [14] Li J, Zhu P F, Zhou R X. J Power Sources, 2011, 196: 9590 [15] Njagi E, hen H, Genuino H, Galindo H, Huang H, Suib S L. Appl atal B, 2010, 99: 103 [16] Tang Z R, Jones D, Aldridge J K W, Davies T E, Bartley J K, arley A F, Taylor S H, Allix M, Dickinson, Rosseinsky M J, laridge J B, Xu Z L, rudace M J, Hutchings G J. hemathem, 2009, 1: 247 [17] Hasegawa Y I, Maki R U, Sano M, Miyake T. Appl atal A, 2009, 371: 67 [18] Liu Q, Wang L, hen M, Liu Y M, ao Y, He H Y, Fan K N. atal Lett, 2008, 121: 144 [19] Fortunato G, swald H R, Reller A. J Mater hem, 2001, 11: 905 [20] ai L N, Guo Y, Lu A H, Branton P, Li W. J Mol atal A, 2012, 360: 35 [21] Li J, Zhu P F, Zuo S F, Huang Q Q, Zhou R X. Appl atal A, 2010, 381: 261 [22] Solsona B, Garcia T, Agouram S, Hutchings G J, Taylor S H. Appl atal B, 2011, 101: 388 [23] Khoudiakov M, Gupta M, Deevi S. Appl atal A, 2005, 291: 151 [24] Zieliński J, Zglinicka I, Znak L, Kaszkur Z. Appl atal A, 2010, 381: 191 [25] Neophytides S G, Marchi A J, Froment G F. Appl atal A, 1992, 86: 45 [26] Gott T, yama S T. J atal, 2009, 263: 359 [27] Manzoli M, Monte R D, Boccuzzi F, oluccia S, Kašpar J. Appl atal B, 2005, 61: 192 [28] Gamarra D, Martínez Arias A. J atal, 2009, 263: 189 过渡金属氧化物掺杂对铜锰氧化物催化 氧化性能的影响 蔡丽娜 a, 胡臻皓 a, Peter Branton b a,*, 李文翠 a 大连理工大学化工学院精细化工国家重点实验室, 辽宁大连 116024 b 英美联合烟草公司研发中心, 英国南安普顿 S15 8TL
Lina ai et al. / hinese Journal of atalysis 35 (2014) 159 167 165 摘要 : 以乙酸铜和乙酸锰为铜锰前驱体, 以 NH 4 H 3 为沉淀剂, 相应金属硝酸盐为掺杂剂, 采用共沉淀法制备了不同过渡金属氧化物掺杂的铜锰氧化物催化剂. 采用 N 2 物理吸附 X 射线衍射, 氢气 - 程序升温还原和原位红外漫反射光谱等方法对催化剂进行了表征, 考察了系列催化剂上 反应性能. 结果表明, 掺杂过渡金属氧化物可以调变催化剂对 的吸附能力, 进而影响催化剂性能. 关键词 : 铜锰氧化物 ; 过渡金属氧化物 ; 掺杂 ; 一氧化碳氧化 ; 原位红外漫反射光谱 收稿日期 : 2013-07-27. 接受日期 : 2013-09-02. 出版日期 : 2014-02-20. * 通讯联系人. 电话 / 传真 : (0411)84986355; 电子信箱 : wencuili@dlut.edu.cn 基金来源 : 中央高校基本科研业务费专项资金 (DUT12ZD218); 高等学校博士学科点专项科研基金 (20100041110017). 本文的英文电子版由 Elsevier 出版社在 ScienceDirect 上出版 (http://www.sciencedirect.com/science/journal/18722067). 1. 前言作为一种低温 消除的主要方法之一, 催化氧化法广泛应用于呼吸防护装置的气体净化 潜艇以及航空航天器材等封闭体系中 消除 2 激光器中气体的纯化 质子膜燃料电池气体纯化和汽车尾气控制等多个领域 [1 6]. 与贵金属催化剂相比, 非贵金属催化剂具有储量丰富 廉价易得的优势. 其中, 铜锰氧化物 (hopcalite), 作为一类价格低廉的过渡金属氧化物催化剂引起了广泛的关注, 如何提高铜锰氧化物的低温活性和抗水汽性能是拓宽其应用范围的关键 [7 9]. 目前, 提高铜锰氧化物催化活性的方法主要是合成工艺的优化和制备方法的改进 [7 20]. 其中, 掺杂过渡金属氧化物可调变催化剂的供氧能力和氧化还原能力, 进而调变催化活性. 如将 e 2 高度分散于铜锰氧化物的表面, 可以减小催化剂的颗粒尺寸, 有效阻止铜锰氧化物的烧结, 提高催化剂的储放氧能力和晶格氧移动能力, 进而提高其催化活性 [7]. 在铜锰氧化物中掺杂约 1.0 wt% 的 o, 可以提高催化剂的比表面积和活性氧物种的数量, 进而提高它在常温常压条件催化 氧化活性 [8]. 我们前期研究了共沉淀过程中沉淀剂和前驱体的选择与匹配, 发现前驱体可以微弱改变催化剂的活性位数量, 而沉淀剂主要改变催化剂的晶相组成, 沉淀剂和前驱体匹配合适, 可获得高活性的铜锰氧化物催化剂 [20]. 在此基础上, 本文针对性选取具有储放氧能力的氧化铁 氧化铈和不具备储放氧能力的氧化锌对铜锰氧化物进行掺杂, 重点研究了不同过渡金属氧化物掺杂对铜锰氧化物催化剂 吸附性能和催化性能的影响. 2. 实验部分 2.1. 催化剂的制备本文中使用的药品均为分析纯, 使用前未经纯化. 以乙酸铜和乙酸锰为铜锰前驱体, 相应过渡金属 (M) 硝酸盐为掺杂剂, NH 4 H 3 为沉淀剂, 采用共沉淀法合成 铜锰氧化物催化剂. 其中 u/mn 比为 1/2, 过渡金属氧化物的掺杂量为 5 wt%. 按照相应的比例称取 2.5 mmol 乙酸铜 5 mmol 乙酸锰和一定量金属硝酸盐 (M = Fe, e, Zn), 加入到 30 ml 去离子水中, 磁力搅拌均匀. 将一定量的 NH 4 H 3 (M/NH 4 H 3 = 1/4) 溶解到 30 ml 去离子水中, 初始 ph 约为 8, 磁力搅拌均匀. 将上述盐溶液快速倒入 NH 4 H 3 溶液中, 磁力搅拌反应 0.5 h 后, 沉淀产物经离心分离, 去离子水于 50 洗涤后干燥 24 h, 300 焙烧 2 h 后, 得到铜锰氧化物催化剂, 命名为 umn x -M. 作为对比, 以醋酸盐为前驱体, 氢氧化钠为沉淀剂, 采用沉淀法分别合成 u 催化剂和 Mn x 催化剂, 并其它合成条件和后处理过程与铜锰氧化物催化剂保持一致. 2.2. 催化剂的表征样品的比表面积 孔体积和孔径分布在麦克公司 Tristar3000 型物理吸附仪上测得, 测试前样品在 200 真空脱气 4 h. X 射线衍射 (XRD) 在日本理学公司 D/Max 2400 型 X 射线衍射仪上测定, 扫描范围 10 80, 扫描速率为 10 /min, 步长为 0.02. 催化剂的形貌和元素分析由 Hitachi S-4800 型扫描电子显微镜 (SEM) 和其搭载的能量色散 X 射线光谱仪 (EDX) 完成. 程序升温还原 (H 2 -TPR) 测试在麦克公司 Auto hem II 2920 型化学吸附仪上进行, 采用热导检测器. 样品在测试之前, 先在 200 Ar 气氛下吹扫 2 h, 降至室温后在 8% H 2 /Ar 气氛下程序升温到 600, 升温速率 10 /min, 由峰面积计算得到耗氢量. 原位红外漫反射光谱测试 (DRIFTS) 在 Nicolet 6700 型红外光谱仪 (MT 检测器 ) 上进行, 采用 KBr 窗片, 样品测试前先在 200 抽真空 (< 10 3 Pa) 处理 0.5 h, 降至 25 后通入 He 吹扫 2 min, 然后通入 5% /N 2, 采集数据. 2.3. 催化剂的评价催化剂的评价在常压固定床流动反应器中进行, 催化剂用量为 200 mg (40 60 目 ). 反应气氛为 1% 和 20% 2, 79% N 2 为平衡气, 反应空速为 20000 ml/(h g cat ). 测试时保持程序升温速率为 1 /min. 出口气体用天美气相色谱 (G7890T) 分析, 检测器是热导池检测器
166 Lina ai et al. / hinese Journal of atalysis 35 (2014) 159 167 (TD). 采用完全转化温度 T 100% 和半转化温度 T 50% 来衡量催化剂的活性. umn x -Fe 催化剂的稳定性测试温度为 30, 空速为 20000 ml/(h g cat ). 3. 结果与讨论 3.1. 铜锰氧化物的结构分析图 1 为不同金属氧化物掺杂的铜锰氧化物催化剂的 XRD 谱, 由图可知, 共沉淀法制备的催化剂是由低结晶度的 Mn 2 3, u 和 u 0.139 Mn 0.861 2 组成 ; 掺杂过渡金属氧化物后, 催化剂的晶相结构没有发生明显变化, 说明掺杂金属氧化物不会显著改变催化剂的体相组成, 且未出现掺杂过渡金属及其衍生物的特征衍射峰, 这可能是由于掺杂的过渡金属氧化物含量较低 ( 约 5 wt%), 且在铜锰氧化物中高度分散或进入到铜锰氧化物的晶格中形成了固溶体 [21]. 图 2 为各掺杂铜锰氧化物催化剂的 N 2 物理吸附等温线和由吸附分支得到的相应孔径分布图. 由图可知, 在相对压力 p/p 0 为 0.4 0.8 处, 各样品吸附分支和脱附分支形成滞后环, 是 IV 型等温线的典型特征, 表明具有介孔特征. 另外, umn x 的孔径分布主要集中在 4.5 nm, 掺杂氧化铁后减小至 2.7 nm, 而 umn x -Zn 和 umn x -e 孔径分布基本不变. 此外, 由表 1 可见, 与 umn x 相比, umn x -Fe 和 umn x -Zn 的比表面积和总孔体积基本不变, 而 umn x -e 的略有降低. 结合 XRD 结果表明掺杂过渡金属氧化物对铜锰氧化物催化剂的比表面积 孔道结构和晶相结构影响较小. 图 3 是 umn x -Fe 和 umn x -e 催化剂的 SEM 照片和相应的面扫元素分布图. 由图可知, 所得催化剂为球形, 尺寸在 0.5 1.5 μm. 由面扫元素分布图可以看出, 掺杂氧化铁和氧化铈的催化剂包含了预期的 Fe 和 e, 且分布均匀, 说明共沉淀法可使 Fe e 等元素均匀掺杂到铜锰氧化物催化剂中. 3.2. 铜锰氧化物催化剂上 氧化反应性能图 4 为铜锰氧化物催化剂上 氧化反应活性图, 表 1 还列出了相应的活性数据. 由图可见, 催化剂活性均随着反应温度上升而提高, 各样品活性顺序为 : umn x - Fe > umn x -e > umn x -Zn > umn x. umn x 催化剂的初始转化率只有 44%, 完全转化温度 (T 100% ) 为 140. 掺杂 Zn 后, 催化剂的低温活性没有明显提高, 但 T 100% 降低至 100. 掺杂 e 2 后, 催化剂活性明显提高, 这是由于 e 2 可以增加催化剂的储放氧能力和氧移动能力 [5]. 其中, umn x -Fe 催化剂表现出优异的催化 氧化活性, 初始 转化率比未掺杂的提高了 35%, 60 即可实现 完全转化, 可能与 Fe 进入铜锰氧化物晶体后形成更多缺陷有关, 因为缺陷有利于反应物种 和 2 的吸附 [6]. 图 4(b) 为 30 时 umn x -Fe 催化剂的稳定性测试. 可以看出, umn x -Fe 催化剂活性随时间延长而下降 ; 反应 2 h 后, 转化率为 65%. 3.3. 铜锰氧化物催化剂的还原性能采用 H 2 -TPR 表征了各催化剂的还原性能, 除 umn x -Fe 的 TPR 图出现 4 个峰外, 其余催化剂都出现 3 个不对称的还原峰, 对该谱图进行拟合结果见图 5, 相应的拟合数据及其耗氢量列于表 2. umn x 催化剂的 H 2 -TPR 谱可拟合为 4 个还原峰, 还原温度最低的两个峰可归属为 u u 2, u 2 u 的还原, 峰值分别为 141.9 和 164.9, 比文献值大幅度降低, 说明铜锰复合后, 两者之间的相互作用增加了铜物种的还原能力 [22]. 还原温度较高的两个峰对应于 Mn 2 3 Mn 3 4, Mn 3 4 Mn 的还原 [15]. umn x -Zn 和 umn x -e 的 TPR 谱与 umn x 相似, 而 umn x -Fe 出现第四个还原峰可归属为 Fe x 物种的还原 [23,24]. 综合分析图 4 和表 2 可以看出, umn x -Zn 催化剂的还原峰向低温方向移动, 表明掺杂 Zn 后, 催化剂的还原能力提高. 而 umn x -Fe 和 umn x -e 中氧化铜和氧化锰物种的还原峰向高温方向移动, 同时各物种的耗氢量也有所降低, 表明掺杂 Fe 和 e 等金属氧化物后催化剂的还原性能下降. 3.4. 铜锰氧化物催化剂的 吸附性能为了进一步研究掺杂对铜锰氧化物催化剂的影响, 采用原位 DRIFTS 考察了 在催化剂表面的吸附行为, 同时与 u 催化剂和 Mn x 催化剂的 吸附性能进行对比. 如图 6 所示, Mn x 只出现 在 Mn 物种上的微弱吸附. u 对 的吸附作用较强, 分别在 2171, 2121 和 2055 cm 1 出现吸附峰. 其中 2121 cm 1 处较强的吸附峰可归属为 的线式吸附 [25,26], 这是 u 催化剂上 的主要吸附形式 ; 2055 和 2171 cm 1 处强度较小的吸附峰可分别归属为 u 0 [27] 的线式吸附和 u 2+ 的线式吸附 [28]. 与 u 催化剂相比, 未掺杂和掺杂后的 umn x 催化剂上 的线式吸附峰红移至 2110 cm 1, 这是由于氧化锰和氧化铜之间的强相互作用使 u 键键能增大, 键能减小所致. 掺杂过渡金属氧化物后, 的线式吸附峰位置没有变化, 说明氧化铜和氧化锰间的相互作用并未减弱, 但吸收峰强度有所增加, 说明金属氧化物的掺杂使催化剂上 吸附位数
Lina ai et al. / hinese Journal of atalysis 35 (2014) 159 167 167 量增加. 掺杂后各催化剂在 2110 cm 1 处吸收峰强弱顺序为 umn x -Fe > umn x -e > umn x -Zn umn x. 结合 H 2 -TPR 结果表明, 掺杂 Fe 和 e 两种金属氧化物有利于增强铜锰间相互作用, 有利于 u 2+ 的还原, 增加了 的比例, 使催化剂的 吸附能力显著增加, 而掺杂 Zn 的催化剂吸附 能力变化不大, 这与催化活性数据基本一致. 图 7 给出了各催化剂在不同吸附时间时原位 DRIFTS 谱. 由图可见, 通入 约 0.5 min 时, umn x -Fe 催化剂上即可出现较强的 吸附, 2 min 时峰强度增加明显, 5 min 后基本上实现 的饱和吸附 ; 而其它催化剂一般需在 8 10 min 才可以达到饱和吸附. 可见, umn x -Fe 催化剂达到吸附饱和所需时间最短, 在反应过程中可以实现 的快速捕获, 进而提高反应活性. 4. 结论采用共沉淀法合成了氧化铁 氧化铈和氧化锌掺杂的铜锰氧化物催化剂. 结果表明, 掺杂不具备储放氧能力的 Zn 可以提高铜锰氧化物的还原性能, 进而提高催化剂的 催化活性, 而掺杂具有储氧能力的氧化铁和氧化铈可以显著提高铜锰氧化物的 吸附活性中心数量和 吸附能力, 进而提高其催化 氧化活性, 可使 30 时 转化率提高 30% 40%, 完全转化温度降至 60.