催化学报  2015, Vol. 36 Issue (2): 252-259   PDF (604 #KB)    
##ZuoZheMingXi_Start## ##ZuoZheMingXi_End##
扩展功能
加入收藏夹
复制引文信息
加入引用管理器
Email Alert
RSS
本文作者相关文章
裴彦鹏a,c, 丁云杰a,b, 臧娟a,c, 宋宪根a,c, 董文达a,c, 朱何俊a, 王涛a, 陈维苗a
Fischer-Tropsch synthesis: Characterizing and reaction testing of Co2C/SiO2 and Co2C/Al2O3 catalysts
Yanpeng Peia,c, Yunjie Dinga,b , Juan Zanga,c, Xiangen Songa,c, Wenda Donga,c, Hejun Zhua, Tao Wanga, Weimiao Chena    
a Dalian National Laboratory for Clean Energy, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, Liaoning, China;
b State Key Laboratory for Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, Liaoning, China;
c University of Chinese Academy of Sciences, Beijing 100049, China
Received: 2014-10-24. Accepted: 2014-11-16. Published: 2015-02-20.
* Corresponding author.Tel/Fax:+86-411-84379143;E-mail:dyj@dicp.ac.cn
Abstract: SiO2- and Al2O3-supported Co2C catalysts were prepared by carburizing supported Co precursors with CO. The catalysts were characterized by N2 physisorption, X-ray diffraction and H2 temperature-programmed reduction techniques, and evaluated by the Fischer-Tropsch (F-T) reaction. The results showed that SiO2- and Al2O3-supported Co2C catalysts could be successfully obtained but sufficient carburization time was required. All of the as-prepared supported Co2C catalysts exhibited activity and selectivity towards alcohols. It is considered that surface metallic Co species contributed to the activity, surface Co2C species were responsible for the formation of alcohols, and bulk Co2C species were inert during the F-T reaction.
© 2014, Dalian Institute of Chemical Physics, Chinese Academy of Sciences.
Published by Elsevier B.V. All rights reserved.
Key words: Supported cobalt carbide     Fischer-Tropsch synthesis     Alcohols     Surface sites    

1. Introduction

Because of the depletion of petroleum resources and the deterioration of the environment, Fischer-Tropsch (F-T) synthesis has attracted great attention because it can be used to produce clean liquid fuels from coal, natural gas, and biomass by catalytic conversion of a synthesis gas intermediate. Many metals, such as Fe, Co, Ni, and Ru, can be used as active catalysts for F-T synthesis, and Fe and Co are the most commonly used [1, 2, 3, 4]. One interesting phenomenon that occurs on the supported Fe- and Co-based catalysts is the formation of iron and cobalt carbide during the reaction. The formation of iron carbides is easy, because the apparent activation energy of carbon diffusion in iron (43.9-69.0 kJ/mol) is substantially lower than for the F-T reaction (89.1 ± 3.8 kJ/mol), which makes it difficult to identify whether iron carbides or iron metal are the real active sites for F-T synthesis [3, 4]. In contrast, the consensus is that the real active site for Co catalysts is metallic Co, and the formation of the Co2C phase would result in deactivation of the Co catalysts [1, 5, 6, 7, 8]. However, Volkova et al. [9] suggested that the appearance of cobalt carbide on Cu-Co catalysts was responsible for the formation of alcohols during the F-T reaction and it followed a CO insertion mechanism. Liu et al. [10] attributed the synthesis of higher alcohols from syngas over LaFeO3-supported Co-Cu catalysts to nano-sized Co2C. We have reported that the synthesis of C1-C18 linear alcohols could be achieved over activated carbon-supported Co catalyst in the presence of Co and Co2C species [11, 12, 13]. Recently, we showed that linear alcohols can be directly synthesized from syngas over unsupported Co2C catalysts, and found that the reactivity and selectivity of the catalysts were closely related to surface Co and Co2C active sites [14, 15].

Some density functional theory (DFT) studies have given a better understanding of the relationship between the structure and catalytic properties of Co2C in F-T synthesis and alcohol formation [16, 17, 18]. However, understanding of the Co2C phase is still inadequate and the following question remains to be answered in the field: why is the appearance of Co2C responsible for the deactivation of the supported Co catalysts? To the best of our knowledge, supported Co2C as catalysts for F-T synthesis has not been reported. In this paper, we report the synthesis and performance of SiO2- and Al2O3-supported Co2C catalysts. We also discuss the role that Co2C species play in supported Co catalysts.

2. Experimental
2.1. Sample preparation

Two types of SiO2 supports were purchased from Qingdao Haiyang Chemical Co., Ltd. (Qingdao, China), and are denoted as SiO2(1) and SiO2(2). The Al2O3 support was obtained from Qingdao Meigao Chemical Co., Ltd. (Qingdao, China). The supports were impregnated using an incipient wetness impregnation method with an appropriate Co(NO3)2·6H2O solution. Co loading was kept constant at approximately 15 wt.% (as metal Co). Following drying at 100 °C for 6 h, the catalyst precursors were calcined in flowing Ar at 350 °C for 4 h. All of the samples were then reduced at 250 °C in flowing H2 (60 mL/min) for 24 h and carburized with CO (1 atm, 60 mL/min) at 220 °C. The SiO2(1), SiO2(2), and Al2O3 supported catalyst precursors were carburized for 140, 145 and 120 h, and are denoted as Co2C/SiO2(1), Co2C/SiO2(2), and Co2C/Al2O3, respectively. For comparison, a SiO2(1) supported Co2C sample carburized for 72 h was prepared and is denoted as Co2C*/SiO2(1).

2.2. Catalyst characterization

The surface area, average pore diameter, and pore volume of the supports and catalysts were measured with nitrogen physisorption at −196 °C using a Micromeritics instrument (Autosorb-1). Prior to the measurements, the samples were degassed in flowing He at 200 °C for 2 h. The average pore diameter and pore volume were determined with the BJH method, and the specific surface area was estimated with the BET method.

The crystalline phases of the catalysts were examined by X-ray diffraction (XRD) with Cu Kα1 radiation on a PANalytical X’Pert PRO diffractometer at 40 kV and 40 mA. The spectra were recorded from 30° to 60° at a scanning rate of 10°/min. For the measurements, the catalysts were ground into a fine powder and placed inside a dish.

H2 temperature-programmed reduction (H2-TPR) experiments were conducted on an Altamira Instruments AMI-300 unit. The sample (150 mg) was treated with a 10% H2/Ar gas mixture at a flow rate of 50 mL/min, and the reduction temperature was increased from room temperature to 900 °C at a heating rate of 10 °C/min. The desorbed exit-gases were continuously monitored using an online mass spectrometer (MS) (Oministar, Pfeiffer Vacuum).

2.3. F-T synthesis

The F-T reaction was performed in a stainless-steel fixed-bed microreactor (9 mm i.d. and 450 mm length). The catalyst (4 mL) was diluted with quartz sand to a volume ratio of 1:1 in the reactor. After loading the catalysts into the reactor, the gas was switched from air to syngas (H2/CO = 1, GHSV = 500 h−1) at 3.0 MPa and the temperature was increased at a 0.3 °C/min heating rate to 220 °C. The reaction effluent passed through a condenser to collect the liquid products. The moment when the reactor reached the designed temperature was taken as the starting time. After 24 h stabilization, the liquid organic products and the aqueous products for another 24 h were off-line analyzed on an HP-6890 gas chromatograph (GC) with 5% PH ME capillary columns and a flame ionization detector after being carefully separated. The outlet gas was on-line analyzed using a HP-6890 GC with a Parapack-Q column and thermal conductivity detector.

3. Results and discussion
3.1. Performance of the F-T reaction

The activity and selectivity results of F-T synthesis over the supported Co2C catalysts are listed in Table 1. Normal hydrocarbons, CO2, and linear α-alcohols in the range of C1-C18 were produced over all the catalysts, which is in agreement with our previous results [11, 12]. The Co2C/SiO2(1) catalyst showed the highest CO conversion (43.9%) and alcohol selectivity (65.6%). The Co2C/SiO2(2) catalyst had only 8.8% CO conversion, but 61.0% alcohol selectivity. The Co2C/Al2O3 catalyst exhibited considerable CO conversion (16.4%) but negligible alcohol selectivity (8.2%).

The alcohols obtained on these catalysts were all linear, and their distribution is in agreement with an Anderson-Schulz -Flory (ASF)-type distribution (not shown here). This agrees with our previous results [11, 12]. Additionally, from Table 1, the proportion of C6+ alcohols in the alcohol fraction for the Co2C/Al2O3 catalyst was 19.7%, while for the Co2C/SiO2(1) and Co2CSiO2(2) catalysts it was only 4.2% and 2.0%, respectively. It is worth mentioning that the maximum alcohol carbon numbers for the Co2C/SiO2(1) and Co2CSiO2(2) catalysts were eight and seven, respectively, while for the Co2C/Al2O3 catalyst the number reached 18. We will further discuss the performance of all the catalysts in the later sections.

Table 1
Performance of the F-T reaction over the supported Co2C catalysts a.
3.2. Nitrogen physisorption measurements

The BET surface area, average pore diameter, and pore volume data of the supports and catalysts are listed in Table 2. The BET surface areas, average pore diameters, and pore volumes of the supported Co2C catalysts were considerably smaller than those of the corresponding supports, which is mainly because of pore blockage [19, 20].

Table 2
Textural properties of the supports and the supported Co2C catalysts.
3.3. XRD results

Fig. 1 shows the XRD patterns of the Co2C/SiO2(1), Co2C/SiO2(2), Co2C/Al2O3, and Co2C/SiO2(1) samples after preparation. The peaks at 2θ = 37.0° and 42.5° for all the samples can be attributed to the (110) and (111) planes of hexagonal Co2C, and the peak at 2θ = 45.7° for Co2C/Al2O3 can be assigned to the Co2C (021) plane (PDF 01-072-1369). No peaks corresponding to cobalt were observed in the patterns of the Co2C/SiO2(1), Co2C/SiO2(2), and Co2C/Al2O3 samples. However, for the Co2C/SiO2(1) sample, a peak at 2θ = 47.5° was observed, which can be ascribed to the (101) plane of hcp cobalt (PDF 00-001-1278). It is known that unsupported Co2C can be obtained by carburizing metallic Co with CO at 220 °C, but more than 500 h carburization time is required [21]. In the present work, the XRD results revealed that supported Co2C catalysts can be successfully prepared by carburizing supported Co precursor after reduction and significantly less carburization time was spent than for unsupported Co2C, but considerable carburization time was still necessary.

Fig. 1. XRD patterns of supported Co2C samples after preparation.

Fig. 2 shows the XRD patterns of the Co2C/SiO2(1), Co2C/SiO2(2), and Co2C/Al2O3 catalysts after reaction. The used Co2C/SiO2(1) and Co2C/SiO2(2) catalysts showed XRD peaks that can only be attributed to Co2C. However, for the used Co2C/Al2O3 catalyst, the characteristic Co2C peaks at 37.0° and 45.7° were observed, but the characteristic Co2C peak at 42.5° almost disappeared and new peaks at 44.3° and 47.5° appeared, which can be attributed to the (111) plane of fcc cobalt (PDF 00-00-1259) and the (101) plane of hcp cobalt (PDF 00-001-1278), respectively. It has been reported that Co2C species readily react with H2 to form CH4 and Co (Co2C + 2H2 ® 2Co + CH4) even at a low temperature of 150 °C [22]. Under real F-T reaction conditions, Co2C would be partially decomposed by H2 into cobalt, although the CO partial pressure would ensure the existence of Co2C species, as we previously reported [15]. Therefore, the metallic cobalt species that were present on the used Co2C/Al2O3 catalyst were derived from the partial decomposition of Co2C. Moreover, the disparate behavior of Co2C species on the SiO2 and Al2O3 supports indicated that, as expected, there may be a support effect that can influence the Co2C phase transition during the reaction. We will discuss this topic more extensively with the findings of the H2-TPR analyses in the next section.

Fig. 2. XRD patterns of supported Co2C catalysts after reaction.
3.4. H2-TPR-MS results

As mentioned in Section 3.3, Co2C species can react with H2 to form CH4 and Co. Based on this, we attempted to use CH4 formed in situ during the catalysts’ H2-TPR process as a probe to investigate the state of Co2C species in contact with the supports. Fig. 3 shows the CH4 formation signals during the H2-TPR process of the supported Co2C catalysts after preparation. Three CH4 peaks appeared at 343, 407 and 556 °C for Co2C/SiO2(1), two CH4 peaks appeared at 343 and 814 °C for Co2C/SiO2(2), and two CH4 peaks appeared at 442 and 662 °C for Co2C/Al2O3. We suggest that the peak at 343 °C for Co2C/SiO2(1) and Co2C/SiO2(2) is because of hydrogenation of surface carbon, the peaks at 407 °C for Co2C/SiO2(2) and 442 °C for Co2C/Al2O3 are because of hydrogenation of subsurface carbon, and the peaks above 509 °C for all the catalysts are because of the hydrogenation of lattice carbon of Co2C strongly interacting with the supports. Based on this characterization and XRD results, we believe that the supported Co2C catalysts can be successfully prepared not only in bulk but also on the surface.

Fig. 3. H2-TPR profiles of the supported Co2C catalysts after preparation.

Fig. 4 shows the H2-TPR results of the three catalysts after reaction. CH4 peaks at 343 and 423 °C (probably because of hydrogenation of surface or subsurface carbon) were observed for the used Co2C/SiO2(1) and Co2C/SiO2(2) catalysts, respectively, but no such peaks were observed for the used Co2C/Al2O3 catalyst. This indicated that there were more surface or subsurface Co2C sites on the SiO2 supported catalysts than on the Al2O3-supported catalysts. Additionally, the CH4 formation peaks because of the hydrogenation of the lattice carbon of Co2C strongly interacting with the supports were observed on all the samples, but the temperature greatly differed compared with those on the fresh samples. This indicated that reconstruction of the bulk Co2C phase of all the catalysts may be induced during the reaction. Correlating the H2-TPR characterization with the performance of the supported Co2C catalysts, the alcohol selectivity was closely related to surface Co2C species. On one hand, for the used Co2C/SiO2(1) and Co2C/SiO2(2) catalysts, considerable Co2C species existed at the surface or subsurface and resulted in high alcohol selectivities, whereas nearly no Co2C species existed at the surface of the used Co2C/Al2O3 catalyst, leading to negligible alcohol yield. On the other hand, the areas of the CH4 formation peaks for the used Co2C/SiO2(1) and Co2C/SiO2(2) catalysts were similar, which indicated that there were almost the same amounts of surface or subsurface Co2C species on the catalyst surface, resulting in similar alcohol selectivity for both catalysts. It is well known that CO conversion of cobalt-based c atalysts is mainly related to the number of Co0 active sites. We previously reported that the activity of unsupported Co2C catalysts originated from the partial transition of the Co2C phase to the Co0 phase [15]. By comparing Figs. 3 and 4, for the Co2C/SiO2(1) catalyst, the CH4 peak at 343 °C decreased and the peak at 407 °C completely disappeared after the reaction. In contrast, for the Co2C/SiO2(2) and Co2C/Al2O3 catalysts, the CH4 peak attributed to surface or subsurface carbon did not significantly change. This indicated that a high number of Co0 active sites appeared on the Co2C/SiO2(1) catalyst, which was responsible for the high activity, whereas the low activity of the Co2C/SiO2(2) and Co2C/Al2O3 catalysts could be because of a low number of Co0 active sites, even though metallic Co phase was present on the used Co2C/Al2O3 catalyst. In the present paper, we only obtained a qualitative relationship between the catalyst activity and the number of Co0 active sites. Thus, more studies have to be carried out to quantitatively determine the number of Co0 active sites of Co2C catalysts under real F-T conditions. Note that, for the used Co2C/SiO2(2) and Co2C/Al2O3 catalysts, the bulk Co2C species that were in contact with the supports did not contribute to the catalyst activity, because both catalysts did not have high CO conversion. Therefore, we suggest that bulk Co2C species are inert. In addition, from Table 1, the production of long-chain alcohols (i.e., C6+ alcohols) was inhibited when the Co2C/SiO2(1) and Co2C/SiO2(2) catalysts exposed more Co2C sites on the su r face compared with the Co2C/Al2O3 catalyst. According to the literature [9, 11], alcohol formation on Co-based catalysts follows a CO insertion mechanism, Co0 is the active species for the C chain growth, and Co2C species can activate CO without rupture and insert CO into C chain intermediates, which leads to alcohol synthesis. Our results indicate that alcohol chain termination was favored when too many surface Co2C species were exposed on the catalyst surface. Of course, not only the role of Co2C but also the reasons for alcohol chain propagation need further study.

Fig. 4. H2-TPR profiles of the supported Co2C catalysts after reaction.
4. Conclusions

SiO2- and Al2O3-supported Co2C catalysts for F-T synthesis were successfully prepared by carburizing the corresponding supported Co precursors with CO, although more than 100 h carburization time was necessary. All of the as-prepared supported Co2C catalysts exhibited activity and selectivity for alcohol formation. It is considered that surface metallic Co originating from the decomposition of surface Co2C under reaction conditions contributed to the catalyst activity, surface or subsurface Co2C species were responsible for the formation of alcohols, and bulk Co2C species that interacted with the supports were inert during the F-T reaction.

References
[1] Khodakov A Y, Chu W, Fongarland P. Chem Rev, 2007, 107: 1692
[2] Schulz H. Appl Catal A, 1999, 186: 1
[3] Zhang Q H, Kang J C, Wang Y. ChemCatChem, 2010, 2: 1030
[4] de Smit E, Weckhuysen B M. Chem Soc Rev, 2008, 37: 2758
[5] Xiong J M, Ding Y J, Wang T, Yan L, Chen W M, Zhu H J, Lu Y. Catal Lett, 2005, 102: 265
[6] Tsakoumis N E, Rønning M, Borg Ø, Rytter E, Holmen A. Catal Today, 2010, 154: 162
[7] Karaca H, Hong J P, Fongarland P, Roussel P, Griboval-Constant A, Lacroix M, Hortmann K, Safonova O V, Khodakov A Y. Chem Commun, 2010, 46: 788
[8] Weller S, Hofer L J E, Anderson R B. J Am Chem Soc, 1948, 70: 799
[9] Volkova G G, Yurieva T M, Plyasova L M, Naumova M I, Zaikovskii V I. J Mol Catal A, 2000, 158: 389
[10] Fang Y Z, Liu Y, Zhang L H. Appl Catal A, 2011, 397: 183
[11] Jiao G P, Ding Y J, Zhu H J, Li X M, Li J W, Lin R H, Dong W D, Gong L F, Pei Y P, Lü Y. Appl Catal A, 2009, 364: 137
[12] Pei Y P, Ding Y J, Zang J, Song X G, Dong W D, Zhu H J, Wang T, Chen W M. Chin J Catal (裴彦鹏, 丁云杰, 臧娟, 宋宪根, 董文达, 朱何俊, 王涛, 陈维苗. 催化学报), 2012, 33: 808
[13] Pei Y P, Ding Y J, Zhu H J, Zang J, Song X G, Dong W D, Wang T, Lü Y. Catal Lett, 2014, 144: 1433
[14] Pei Y P, Ding Y J, Zang J, Song X G, Dong W D, Zhu H J, Wang T, Chen W M. Chin J Catal (裴彦鹏, 丁云杰, 臧娟, 宋宪根, 董文达, 朱何俊, 王涛, 陈维苗. 催化学报), 2013, 34: 1570
[15] Pei Y P, Ding Y J, Zhu H J, Zang J, Song X G, Dong W D, Wang T, Yan L, Lu Y. Reac Kinet Mech Catal, 2014, 111: 505
[16] Cheng J, Hu P, Ellis P, French S, Kelly G, Lok C M. J Phys Chem C, 2010, 114: 1085
[17] Lebarbier V M, Mei D H, Kim D H, Andersen A, Male J L, Holladay J E, Rousseau R, Wang Y. J Phys Chem C, 2011, 115: 17440
[18] Zhao Y H, Su H Y, Sun K J, Liu J X, Li W X. Surf Sci, 2012, 606: 598
[19] Bechara R, Balloy D, Dauphin J Y, Grimblot J. Chem Mater, 1999, 11: 1703
[20] Dimitrova P G, Mehandjiev D R. J Catal, 1994, 145: 356
[21] Bahr H A, Jessn V. Ber Chem, 1930, 63B: 2226
[22] Hofer L J E, Peebles W C. J Am Chem Soc, 1947, 69: 2497
费托合成:SiO2和Al2O3负载的Co2C催化剂的表征和评价
裴彦鹏a,c, 丁云杰a,b , 臧娟a,c, 宋宪根a,c, 董文达a,c, 朱何俊a, 王涛a, 陈维苗a    
a 中国科学院大连化学物理研究所洁净能源国家实验室, 辽宁大连116023;
b 中国科学院大连化学物理研究所催化基础国家重点实验室, 辽宁大连116023;
c 中国科学院研究生院, 北京100049
收稿日期: 2014-10-24. 接受日期: 2014-11-16. 出版日期: 2015-02-20.
*通讯联系人.丁云杰,电话/传真: (0411)84379143; 电子信箱: dyj@dicp.ac.cn
摘要：采用CO碳化SiO2和Al2O3负载的Co(NO3)2的方法制备了SiO2和Al2O3负载的Co2C催化剂, 采用N2物理吸附、X射线衍射和H2-程序升温还原技术对催化剂进行了表征, 并用于催化费托合成反应中. 结果显示, 需要较长碳化时间才可合成负载的Co2C催化剂; 所制催化剂表现出CO加氢生成高碳醇的催化性能, 其原因可能在于催化剂表面存在的金属Co物种使CO解离, 表面Co2C物种有利于CO插入, 从而导致醇的生成, 但体相Co2C则不具有催化活性.
关键词： 负载型碳化钴     费托合成     醇     表面活性位    
1. 前言

由于石油资源的减少和自然环境的恶化,将煤、天然气和生物质经合成气催化转化为洁净燃料的费托合成引起了广泛关注. 许多金属,比如Fe,Co,Ni和Ru都可以用于费托合成反应,其中Fe和Co是最常用的[1, 2, 3, 4]. 有趣是,负载型Fe和Co基催化剂在反应过程中可以形成Fe和Co的碳化物. Fe的碳化物很容易形成,这是由于C在Fe晶格内的扩散活化能(43.9-69.0kJ/mol)显著低于费托反应活化能(89.1±3.8kJ/mol),因而不利于区别Fe和Fe的碳化物在费托反应中的作用[3, 4]. 相比之下,人们普遍认为金属Co是费托合成Co基催化剂的活性中心,而Co2C的形成会导致催化剂失活[1, 5, 6, 7, 8]. 然而,Volkova等[9]报道了Cu-Co催化剂上醇的生成可能与Co2C的出现有关,其且遵循CO插入机理. Liu等[10]将LaFeO3负载的Co-Cu催化剂上CO加氢生成醇的性能归因于纳米形态的Co2C物种. 我们报道了活性炭负载的Co基催化剂在反应过程中可以原位生成Co2C物种,可导致C1-C18线性混合醇的生成[11, 12, 13]. 随后,我们发现线性混合醇可以在非负载的Co2C催化剂上生成,而且催化 剂的活性和选择性与表面Co和Co2C物种有关[14, 15].

密度泛函理论计算结果给出了Co2C结构与其催化性能的关系[16, 17, 18],但是对Co2C的研究仍然不足,特别是如下问题仍未解决:为什么负载型Co基催化剂上出现的Co2C物种会导致催化剂失活. 另一方面,据我们所知,目前没有关于负载型Co2C作为费托合成催化剂的报道. 本文报道了SiO2和Al2O3负载的Co2C催化剂的合成和及其催化费托反应结果,并讨论了Co2C物种在负载型Co基催化剂上的作用.

2. 实验部分
2.1. 样品制备

本文选用两种SiO2作为载体(青岛海洋化工有限公司),分别记为SiO2(1)和SiO2(2),一种Al2O3作为载体(青岛美高化工有限公司). 首先将Co(NO3)2·6H2O浸渍到载体上,使Co含量为15%,然后将样品在100°C干燥6h,在350°C于Ar气氛中焙烧4h. 之后,将样品在250°C于氢气流(60mL/min)中还原24h,然后采用CO(0.1MPa,60mL/min)于220°C碳化. SiO2(1),SiO2(2) 和Al2O3负载的样品分别碳化140,145和120h,记为Co2C/SiO2(1),Co2C/SiO2(2)和Co2C/Al2O3. 为了比较,制备了SiO2(1)负载的、碳化72h的样品,记为Co2C*/SiO2(1).

2.2. 催化剂表征

样品的比表面积、平均孔径和孔体积在MicromeriticsAutosorb-1型物理吸附仪上进行. N2吸附前,样品先在200°C脱气处理2h. 样品比表面积由N2吸附等温线结合BET方程求得,孔体积和平均孔径由BJH模型计算而得.

采用PANalytical公司的X’PertPRO型X射线粉末衍射(XRD)仪测定样品的晶相组成. 选用CuKα1靶,管流40mA,管压40kV,扫描范围30°-60°,扫描速度为10°/min.测试前,样品经粉碎研磨后压入玻璃片内.

H2程序升温还原(H2-TPR)实验在Altamira公司的AMI-300型化学吸附仪器上进行. 样品装填量为150mg,以10%H2/Ar为还原气,流量为50mL/min,以10oC/min的升温速率升温至900°C,同时用瑞士Oministar Pfeiffer Vacuum型质谱仪跟踪测定程序升温过程中的气体产物.

2.3. 费托合成反应

费托反应在内径为9mm的不锈钢固定床反应器(长度为450mm)中进行. 催化剂装填量为4mL,并采用体积比为1:1的石英砂稀释. 将催化剂置至于反应器内后,引入合成气(H2/CO=1,GHSV=500h-1),将压力设为3.0MPa,然后以0.3°C/min的速率升温至220°C. 反应器温度达到设定值时开始计时. 经24h稳定后,液相产物每24h放样一次. 液相产物经水浴冷阱收集,静置至油相和水相分层后,用分液漏斗分离,采用HP-6890型气相色谱仪(5%PHME柱和FID检测器)离线分析,&l t;/ span>以仲丁醇为内标物,采用内标法分析水相中含氧化合物含量;尾气采用HP-6890GC型气相色谱仪(Parapack-Q柱和TCD检测器)在线分析.

3. 结果与讨论
3.1. 费托反应结果

表1为所制负载型Co2C催化剂上费托反应结果. 可以看到,所有催化剂上均生成烃类、CO2和C1-C18线性混合醇,与我们以前的结果[11, 12]类似. Co2C/SiO2(1)催化剂具有最高的CO转化率(43.9%)和醇的选择性(65.6%);Co2C/SiO2(2)催化剂的CO转化率只有8.8%,但醇的选择性达到了61.0%;Co2C/Al2O3催化剂具有一定的CO转化率,但醇选择性仅为8.2%.

我们发现,所得的醇类产物都是线性的,且遵循ASF分布,与我们之前的研究结果[11, 12]一致. 还可以看到,Co2C/Al2O3催化剂上C6+醇占总醇的19.7%;而对于Co2C/SiO2(1)和Co2C/SiO2(2)催化剂,C6+醇分别只占4.2%和2.0%,且最大碳数的醇分别为正辛醇和正庚醇;而对于Co2C/Al2O3催化剂,最大碳数的醇是正十八醇.

3.2. N2物理吸附表征结果

载体及各催化剂的比表面积、平均孔径和孔体积结果列于表2中. 负载的Co2C催化剂的比表面积、平均孔径和孔体积明显小于相应的载体,这可能是由于载体的孔堵塞所导致的[yes].

3.3. XRD结果

图1为新鲜Co2C/SiO2(1),Co2C/SiO2(2),Co2C/Al2O3和Co2C/SiO2(1)样品的XRD谱. 所有样品均在2θ=37.0°和42.5°出现对应于Co2C的(110)和(111)面的衍射峰,同样,Co2C/Al2O3样品在2θ=45.7°出现对应于Co2C的(021)面(PDF01-072-1369)的衍射峰. 另外,Co2C/SiO2(1),Co2C/SiO2(2)和Co2C/Al2O3样品均未出现Co的特征峰. 而在Co2C/SiO2(1)样品上,在2θ=47.5°出现对应hcp Co的(101)面(PDF 00-001-1278)的衍射峰. 研究发现,需碳化500h以上才可得到非负载的Co2C[21]. 本文则可以通过碳化负载金属Co(NO3)2的方式得到负载型Co2C,虽然碳化时间大幅减少,但仍需足够的时间才可保证碳化完全.

图2为反应后的Co2C/SiO2(1),Co2C/SiO2(2)和Co2C/Al2O3催化剂的XRD谱. 由图可见,Co2C/SiO2(1)和Co2C/SiO2(2)催化剂上只出现Co2C的特征衍射峰;而在Co2C/Al2O3催化剂上可以观察到位于37.0°和45.7°的Co2C特征峰,但是位于42.5°的Co2C特征峰几乎消失,同时在44.3°和45.7°出现了新的衍射峰,分别对应fccCo的(111)面(PDF00-00-1259)和hcpCo的& lt; /span>(101)面(PDF00-001-1278). 研究表明[22],Co2C很容易与H2反应转化为金属Co,同时生成CH4(反应温度可低至150°C). 我们也发现[15],在真实费托反应条件下,Co2C会与H2作用而部分分解成金属Co,同时由于CO分压的作用,一部分Co2C又得以稳定存在. 因此,反应后Co2C/Al2O3催化剂上出现的金属Co峰源于Co2C的分解.这明显不同于SiO2负载的Co2C性质,说明载体可以影响Co2C在反应条件下的转变.&l t;/ span>

3.4. H2-TPR-MS结果

如前所述,Co2C可与H2反应而转化为金属Co和CH4. 因此,我们以Co2C在H2-TPR过程中生成的CH4为探针,考察了Co2C物种与载体的接触状态. 图3给出了各新鲜催化剂在H2-TPR过程中生成CH4的信号图. 可见,Co2C/SiO2(1)在343,407和556°C出现了3个峰;Co2C/SiO2(2)在343和814°C出现了2个峰;Co2C/Al2O3在442和662°C也出现了2个峰. 我们认为,Co2C/SiO2(1)和Co2C/SiO2(2)在343°C出现的CH4峰是由于样品表面Co2C物种加氢而形成的;Co2C/SiO2(2)在407°C和Co2C/Al2O3在442°C出现的CH4峰归因于样品次表面Co2C物种的加氢,而各催化剂上高于509°C出现的CH4峰是由于样品体相Co2C物种(与载体接触)加氢形成的. 综合H2-TPR和XRD结果可认为,负载型Co2C催化剂的合成 是成功的 ,不仅催化剂体相由Co2C构成,而且表面也存在Co2C物种.

图4为反应后各催化剂的H2-TPR谱. 可见,Co2C/SiO2(1)和Co2C/SiO2(2)分别在343和423°C出现CH4峰,可能源于催化剂表面或次表面Co2C物种的加氢. 然而Co2C/Al2O3催化剂上没有明显地观察到此类峰的出现. 这意味着SiO2负载的催化剂比Al2O3负载的催化剂在反应中存在着更多的表面或次表面Co2C物种. 另外,所有反应后的催化剂均出现体相Co2C物种加氢而形成的甲烷峰,但是与反应前的相比,峰温存在明显差别. 这意味着催化剂的体相Co2C结构可能在反应过程中发生重构. 将H2-TPR结果与催化剂性能相关联,可以看出,醇的选择性与表面Co2C物种紧密相关. 一方面,在反应后Co2C/SiO2(1)和Co2C/SiO2(2)催化剂上存在着大量的Co2C物种,从而导致了高的醇的选择性,而Co2C/Al2O3催化剂上几乎没有表面Co2C物种,因而醇的选择性非常低;另一方面,Co2C/SiO2(1)和Co2C/SiO2(2)催化剂上CH4形成峰的面积几乎相同,说明这两种催化剂的表面Co2C物种数量几乎相等,因而醇的选择性相近. 众所周知,Co基催化剂上Co0是催化活性中心. 我们之前报道了非负载的Co2C催化剂的活性源于费托反应中Co2C部分分解而形成的金属Co相[15]. 比较图3和图4可以发现,对于Co2C/SiO2(1)催化剂,343°C出现的CH4峰面积显著减小,而407°C处的峰几乎消失;对于Co2C/SiO2(2)和Co2C/Al2O3催化剂,由表面Co2C加氢形成的CH4峰没有显著的变化. 这可能说明反应中Co2C/SiO2(1)催化剂上Co0物种较多,从而导致高的催化活性,而Co2C/SiO2(2)和Co2 C /Al2O3催化剂上Co0的数量少,因而催化剂活性低. 我们仅定性地分析了催化剂活性与Co0数量的关系,定量研究仍需大量的工作. 我们也注意到,对于Co2C/SiO2(2)和Co2C/Al2O3催化剂,与载体接触的体相Co2C的存在并未导致催化剂有高的活性,因此可推断体相Co2C是惰性的. 另外,由表1可见,Co2C/SiO2(1)和Co2C/SiO2(2)催化剂上长链醇(即C6+醇)的形成被抑制,这可能是由于这两个表面暴露了过多的Co2C活性位. 根据文献[9, 11]报道,Co基催化剂上醇的生成可能遵循CO插入机理,并且Co0是碳链增长的中心,而Co2C是非解离吸附CO使其插入到碳链中间体中生成醇的中心. 我们的结果显示,过多的表面Co2C物种可以起到终止醇类碳链增长的作用. 当然,不仅Co2C的催化作用,而且影响醇的碳链增长因素均值得深入考察.

4. 结论

采用CO碳化SiO2和Al2O3负载Co(NO3)2的方法,成功制备了用于费托合成的相应负载型Co2C催化剂,但是制备过程需要100h以上. 所制催化剂显示出了催化CO加氢生成醇的能力. Co2C部分分解而形成的表面金属Co物种使催化剂具有活性,而表面Co2C物种是醇的生成中心,但与载体接触的体相Co2C没有催化活性.