- Email: [email protected]
Effects of cobalt carbide on Fischer–Tropsch synthesis with MnO supported Co-based catalysts
Journal of Energy Chemistry 42 (2020) 227–232
Contents lists available at ScienceDirect
Journal of Energy Chemistry journal homepage: www.elsevier.com/locate/jechem
Effects of cobalt carbide on Fischer–Tropsch synthesis with MnO supported Co-based catalysts Fanfei Sun a,b,c,1, Ruoou Yang b,1, Zhaoming Xia d, Yuqi Yang a,b,c, Ziang Zhao e, Songqi Gu a,b, Dongshuang Wu f,∗, Yunjie Ding e,∗, Zheng Jiang a,b,∗∗ a
Shanghai Institute of Applied Physics, Chinese Academy of Sciences, Shanghai 201204, China Shanghai Synchrotron Radiation Facility, Zhangjiang Laboratory, Shanghai Advanced Research Institute, Chinese Academy of Sciences, Shanghai 201210, China c University of Chinese Academy of Sciences, Beijing 100100, China d Center for Applied Chemical Research, Frontier Institute of Science and Technology, Xi’an Jiaotong University, Xi’an 710049, Shaanxi, China e Dalian National Laboratory for Clean Energy, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, Liaoning, China f Graduate school of science, Kyoto University, Kitashirakawa, Oiwake-cho, Sakyo-ku, Kyoto-shi, Kyoto-fu, 606-8502, Japan b
a r t i c l e
i n f o
Article history: Received 18 April 2019 Revised 27 June 2019 Accepted 11 July 2019 Available online 15 July 2019 Keywords: Cobalt carbide Fischer–Tropsch synthesis X-ray absorption fine spectroscopy nanomaterials Density functional theory
a b s t r a c t Cobalt carbide (Co2 C) was considered as potential catalysts available for large-scale industrialization of transforming syngas (H2 and CO) to clean fuels. Herein, we successfully synthesized Co-based catalysts with MnO supported, to comprehend the effects of Co2 C for Fischer–Tropsch synthesis (FTS) under ambient conditions. The huge variety of product selectivity which was contained by different active sites (Co and Co2 C) has been found. Furthermore, density functional theory (DFT) shows that Co2 C is efficacious of CO adsorption, whereas is weaker for H adsorption than Co. Combining the advantages of Co and Co2 C, the catalyst herein can not only obtain more C5+ products but also suppress methane selectivity. It can be a commendable guide for the design of industrial application products in FTS. © 2019 Science Press and Dalian Institute of Chemical Physics, Chinese Academy of Sciences. Published by Elsevier B.V. and Science Press. All rights reserved.
1. Introduction Fischer–Tropsch synthesis (FTS), a significant step in the transformation of syngas (a mixture of H2 and CO) to clean fuels or value-added chemicals, is of both industrial and fundamental importance [1–4]. The challenge of selective formation of available hydrocarbons through FTS lies in preciously controlling C–C coupling by adjusting hydrogenation and suppressing methane formation under reaction [5]. To this end, catalysts with both high activity and selectivity are highly desirable to obtain products on-demand. Transitional metal carbides (TMC) have been emerged as a promising class of catalytic materials in FTS because their metallic-like properties [6–11]. Among TMC, iron carbide repre-
∗
Corresponding authors. Corresponding author at: Shanghai Institute of Applied Physics, Chinese Academy of Sciences, Shanghai 201204, China. E-mail addresses: [email protected] (D. Wu), [email protected] (Y. Ding), [email protected] (Z. Jiang). 1 These authors contributed equally to this work. ∗∗
sents a leading catalyst to afford long-chain hydrocarbons due to its high stability, per pass conversion and hydrocarbon selectivity [12,13]. In contract, its analog, cobalt carbide, has long been considered as an unwanted byproduct which results in the activity loss of the metallic Co active center during FTS [14]. Quite recently, the role of Co2 C has been turned. For example, Zhong et al. revealed a high relevancy among the particle size, the exposed facets of Co2 C and catalytic performance [15]. The specific Co2 C surface exhibited excellent selectivity towards the formation of lower olefins with low selectivity towards methane production [16]. Ding et al. synthesized stable and efficient Co-Co2 C catalysts [17] and proposed that Co-Co2 C interfaces have synergistic effects and are responsible for the formation of alcohols [18]. Although tremendous efforts have been made to understand the effect of the cobalt carbide, due to the tanglesome reactions in FTS, it is rather difficult to elucidate the formation process of the active site and the characterization of the catalysts [17,19–23]. Herein, we used co-precipitation method to successfully synthesize Co/Co2 C@MnO catalyst which performed different selectivities with Co@MnO and Co2 C@MnO catalysts under ambient reaction conditions (250 °C, 0.1 MPa). TEM and in-situ XRD experiments
https://doi.org/10.1016/j.jechem.2019.07.007 2095-4956/© 2019 Science Press and Dalian Institute of Chemical Physics, Chinese Academy of Sciences. Published by Elsevier B.V. and Science Press. All rights reserved.
228
F. Sun, R. Yang and Z. Xia et al. / Journal of Energy Chemistry 42 (2020) 227–232
confirmed the formation of coexisted Co and Co2 C. The atomic structure of the Co/Co2 C@MnO catalyst was studied by X-ray absorption fine structure (XAFS). Based on the DFT analysis of the Co-based catalysts, we further understood the catalytic mechanism of the cobalt carbide and provided guidance for the molecular design of the catalytic performance of the Co-based catalysts in FTS. 2. Experimental
temperature was further increase into 250 at a heating speed of 0.5 °C/min. At the outlet of the reactor, we use a condenser to separate the gas and liquid products. The catalytic performances were analyzed after stabilization for 12 h. Agilent 6890 N GC gas chromatograph with a HP-5 capillary column and a flame ionization detector (FID) was used to analyze products. The water-bearing products and oil-bearing products were analyzed by butyl with internal standard method and area normalization method, respectively.
2.1. Synthesis of catalyst
2.4. DFT calculations
Co@MnO catalyst was synthesized by co-precipitation method refer to our previous work [14]. In a typical synthesis, an aqueous solution containing cobalt nitrate (Co(NO3 )2 ·6H2 O); Sinopharm) and manganese nitrate (50 wt% Mn(NO3 )2 , Sinopharm) was prepared (2 M, (mol/mol) = 2/1), and Na2 CO3 aqueous solution (2 M) was added. The mixed solution was vigorously stirring at RT. Following, the formed suspension went through a procedure of filtration, repeated washing process and dried overnight at 100 °C. Finally, we calcined the dried powder at 330 °C for 3 h in the air. When the solution precipitant was replaced by the mix of (NH4 )2 CO3 and Na2 CO3 solution (NH/Na (mol/mol) = 1/1), we obtained Co/Co2 C@MnO. If the solution precipitant was replaced by only Na2 CO3 solution, Co2 C@MnO was obtained.
Spin-polarized plane-wave basis set density functional theory (DFT) as implemented in the Vienna ab initio Simulation Package (VASP) [29]. Perdew–Burke–Ernzerhof (PBE) [29] generalized gradient approximation (GGA) [30] functional was used to evaluate the exchange-correlation energy. The energy and force convergence were 10−5 eV and 0.02 eV, respectively. The energy cut off was 400 eV. A 16 × 16 × 16 Monkhorst–Pack k-point mesh was employed in the calculations of bulk properties. The lattice volume was fitted with the third-order Birch–Murnaghan isothermal equation of state [31]. The resulted lattice parameters of FCC ˚ b = 4.361 A˚ Co and Co2 C are a = b = c = 3.457 A˚ and a = 2.850 A, ˚ respectively. For the calculation of CO and H2 and c = 4.345 A, molecule, a 20 × 20 × 20 unit cell and 1 × 1 × 1 k-point mesh were employed. A 15 A˚ vacuum region was applied to separate slabs in the calculation of slab model. We only investigate the most stable surface of Co2 C, which is Co2 C (101) as is reported previously [32]. A p(2 × 2) unit cells for Co2 C (101) were utilized in our calculation, which contains 4 layers of Co and 2 layers of C. The bottom 2 Co layers and one C layer were fixed, and other layers were fully relaxed. A p(3 × 3) unit cell for FCC Co (111) was also applied in the calculation. It contained three Co layers with the bottom one layer fixed. (Scheme S1) Monkhorst-Pack mesh k-points of (3 × 3 × 1), (3 × 4 × 1) were used for FCC Co (111), Co2 C (101)-Cterminated and Co2 C (101)-Co-terminated surface calculations, respectively. Dipole corrections have been applied in the calculations to avoid artificial dipole effects. Since CO and H adsorptions on Co (111) were well studied, we only consider the most stable adsorption site. H was on the 3-fold-fcc-hollow site and CO was on the top of Co atom. In the case of adsorption on Co2 C–C-terminated, four sites were taken into consideration, as was shown in Scheme S1. However, some of them were not favorable for CO/H adsorption. There were only three distinguishable configurations for CO adsorption (Scheme S2) and two for H adsorption (Scheme S3). The most stable adsorption site for H and CO on Co2 C (101)-Coterminated was shown in Scheme S4. The adsorption energy was defined as:
2.2. Characterization 2.2.1. In situ X-ray diffraction (XRD) analysis In situ powder X-ray diffraction measurement and analysis for Fischer–Tropsch reaction were referred to our previous work [24]. The phase transformation of the calcined samples was monitored during reduction process at 0.1 MPa in a H2 flux of 20 mL/min with a thermal rate of 1 °C/min from RT to 350 °C and the XRD patterns were collected every 20 °C. Then, the samples were kept at 350 °C about 5 h and the XRD spectra were collected every 30 min at 350 °C (Fig. S1). Afterwards, the cell temperature was cooled down to 250 °C, at which the catalytic behavior was monitored under 0.1 MPa syngas (H2 /CO=2/1) at a flux rate of 20 mL/min (Fig. 1). We collected the XRD spectra every 30 min during the FT process. Each scan was performed at 40 kV and 40 mA with a scanning angle range of 30°–80° at a rate of 2 °/min. All the XRD spectra were carried out by the X’pert HighScore Plus software package with the PDF library. 2.2.2. Transmission electron micrograph (TEM) The morphology of the spent catalysts was observed by Tecnai G2 S-TWIN instrument with an accelerating voltage of 200 kV [25]. 2.2.3. X-ray absorption fine spectroscopy (XAFS) Co K-edge XAFS data were measured in transmission mode at beamline BL14W1 of Shanghai Synchrotron Radiation Facility (SSRF) [26]. The extended X-ray absorption fine structure (EXAFS) data analyses were carried out by using the Demeter software package [27,28]. The Fourier transformation of the k3 -weighted functions was applied to the EXAFS data. The fitting of the EXAFS data was calculated by Artemis program in Ifeffit. 2.3. Catalytic test Catalysts were tested in a fixed-bed stainless steel plug-flow reactor with inner diameter 9 mm refer to our previous work [14]. In a typical test, ca. 2 mL catalyst and 2 mL SiO2 were mixed to obtain isothermal plug flow conditions. The reactor was filled with 0.1 MPa H2 at RT and heated to 175 °C at a rate of 5 °C/min and then the gas atmosphere was changed into syngas (H2 /CO = 2). The
Ead (H ) = Eslab+H − Eslab − EH2 /2 Ead (CO ) = Eslab+CO − Eslab − Eco Ead (H) was the adsorption energy of H. Eslab was the energy of Co (111) or Co2 C(101) slab model. EH2 was the energy of H2 . Eslab+H was the energy of slab model with one H atom. Ead (CO) was the adsorption energy of CO. ECO was the energy of CO. Eslab+CO was the energy of slab model with CO adsorbed on it. 3. Results and discussion 3.1. In situ XRD Despite of different preparing methods, the in-situ XRD results recorded during H2 reduction suggest that all the three catalysts
F. Sun, R. Yang and Z. Xia et al. / Journal of Energy Chemistry 42 (2020) 227–232
229
Fig. 1. In situ XRD pattern for Co@MnO (a), Co/Co2 C@MnO (b) and Co2 C@MnO (c) catalysts in a syngas flow of 6 h (PDF number: MnO (0 0-0 04-0326), Cox Mn1− x O (0 0-0 040326), fcc Co (0 0-0 01-1259) and Co3 O4 (0 0-0 01-1152)).
are composed by a mixture of MnO, metallic Co and Cox Mn1 −x O after the pre-treatment process (Fig. S1). Fig. 1 shows the insitu XRD patterns of the catalysts in syngas. For Co@MnO catalyst, the peaks belonging to face center cubic (FCC) Co and MnO existing in the whole 6 h syngas reaction and Co2 C phase cannot be observed (Fig. 1(a)). For Co/Co2 C@MnO catalyst, in addition to the FCC Co peak, Co2 C peaks were shown in a syngas flow at 250 °C (Fig. 1(b)), which can be assigned as 2θ = 37.0°, 41.3°, 42.6°, 45.8°, and 56.6° corresponding to the crystalline plane of Co2 C (110), (002), (111), (021), and (112) according to PDF 01-0721369, respectively. The formed Co2 C was very stable even after a 10 h syngas gas reaction (Fig. S2). It suggests the co-existence of FCC Co and Co2 C in Co/Co2 C@MnO catalyst. As for Co2 C@MnO catalyst, after the syngas feed was introduced, the cobalt metal gradually turned into Co2 C with time on-stream. Meanwhile, the single phase (Cox Mn1 −x O) segregated into FCC Co and MnO, respectively. After 4 h of reaction, the structure was stable and the metallic cobalt completely transformed into Co2 C. It can be clearly seen that Co2 C formed in Co/Co2 C@MnO and Co2 C@MnO, which suggests the importance role of Na2 CO3 in the formation of Co2 C.
3.2. TEM TEM and HRTEM images of the three spent catalysts are shown in Fig. 2. All are near spherical aggregated nanoparticles with quite a similar average particle size approximating 10 nm (Fig. 2(a)–(c)). The plane distance with ca. 1.53 A˚ can be assigned as {220} plane of MnO, which is observed in all three samples. A plane distance of ca. 2.01 A˚ and ca. 2.40 A˚ corresponds to {111} plane of Co and {101} plane of Co2 C, respectively. As shown in Fig. 2(d) and (f) only Co was observed in Co@MnO and only Co2 C existed in Co2 C@MnO. While, in the Co/Co2 C@MnO catalyst, the coexistence of the FCC cobalt metal and the cobalt carbide phase was revealed (Fig. 2(e)). The TEM measurements correspond to the result of in-situ XRD analysis and suggest that the active center in three catalysts might be different, that is, Co, Co/Co2 C, and Co2 C, respectively. 3.3. Catalytic performance In order to understand catalytic mechanism, the catalytic activity of Co-based catalysts is presented. The hydrocarbons and
230
F. Sun, R. Yang and Z. Xia et al. / Journal of Energy Chemistry 42 (2020) 227–232
Fig. 2. TEM images of the Co@MnO (a) and (d), Co/Co2 C@MnO (b) and (e) and Co2 C@MnO (c) and (f) catalysts.
Table 1. Catalytic performance of the Co-based catalyst samples. Product selectivity (C%, CO2 -free) Catalyst
CO content (%)
CO2 selectivity (C%)
CH4
C2–4=
C2–4o
C5+
Oxy
Co@MnO Co/Co2 C@MnO Co2 C@MnO
54.4 55.5 52.2
10.3 32.7 44.2
37.5 18.3 18.1
18.6 22.9 39.2
12.7 15.2 18.8
25.9 42.5 22.3
5.3 1.1 1.6
Reaction conditions: 250 °C, 30 0 0 mL h−1 gcat −1 (where gcat denotes grams of catalyst). Oxy: oxygenates.
alcohols are all possible products through FTS. The catalytic performances of the prepared catalysts during FT reaction were performed at 250 °C, 0.1 MPa, and GHSV = 30 0 0 mLsyngas /(gcat − h) with H2 /CO = 2 in a fixed-bed reactor. The variations in CO conversion and product selectivity for the spent catalysts are listed in Table 1. The CO conversion is similar for all the catalysts. However, we observed diverse product selectivities over these catalysts. Specifically, the selectivities of methane for Co/Co2 C@MnO and Co2 C@MnO are only 18.1% and 18.3%, respectively, which are much lower than that for Co@MnO (37.5%). Moreover, the production of lower olefins is 39.2% on Co2 C@MnO which is higher than those of Co@MnO (18.6%) and Co/Co2 C@MO (22.9%). These results suggest that the Co2 C also can activate CO like metallic cobalt, however, it suppresses methane formation by suppressing the over-hydrogenation. Nevertheless, the highest C5+ selectivity is observed on Co/Co2 C@MnO (42.5%), while, the other two catalysts show only around half of the value that is 25.9% and 22.3% for Co@MnO and Co2 C@MnO, respectively. This result suggests that a higher C–C coupling is favored in the presence of both Co and cobalt carbide. In addition, the effect of MnO is removed by catalytic performance (Table S3) and water gas shift reaction (Table S4) in this FTS condition.
3.4. XAFS measurement X-ray Absorption Near Edge Spectroscopy (XANES) analysis is used to distinguish the structure and valence state of the Co-based catalysts in Fig. 3. Fig. 3(a) shows the Co K-edge normalized XANES spectra of the calcined Co-based catalysts and a comparison of Co, Co3 O4 and Co2 C [14] reference compounds. As the ion charge corresponds to the transition energy, the edge position shifts toward lower energies with oxidation states: 7709.8 eV and 7729.0 eV for Co and Co3 O4 , respectively. All the spectrums of the calcined Co-based catalysts exhibit oxidation states similar to the Co3 O4 reference. Fig. 3(c) exhibits the Co K-edge XANES profile in the spent catalysts and reference compounds for comparison. A peak can be found at the pre-edge at 7712 eV for Co/Co2 C@MnO and Co2 C@MnO, due to the consist of the non-centrosymmetric Co2 C in catalysts. The absorption edge around in the spent Co-based catalysts sample showed a medium signal between the Co reference and CoO, which revealed that the valence state is lower than +2 in the spent catalysts. Moreover, the ratio of Co2 C to Co of Co/Co2 C@MnO catalyst is 1.68 (calculated by linear combination fitting (LCF) listed in Table S2), which impacts methane and C5+ selectivity on the Co-based catalysts.
F. Sun, R. Yang and Z. Xia et al. / Journal of Energy Chemistry 42 (2020) 227–232
231
Fig. 3. XAFS characterization of Co-based catalyst samples. (a) Cobalt K-edge (7709 eV) XANES spectra of calcined Co-based catalyst samples; (b) Fourier transform spectra of the calcined Co-based catalyst samples; (c) Cobalt K-edge (7709 eV) XANES spectra of spent Co-based catalyst samples; (d) Fourier transform spectra of the spent Co-based catalyst samples.
To evaluate the local geometry around Co atoms, both the calcined and the spent catalysts were collected by extended X-ray Absorption Fine Structure Spectroscopy (EXAFS). The k3 -weighted EXAFS Fourier transform magnitude spectra of the calcined and spent Co-based catalysts are provided in Fig. 3(b) and (d). In Fig. 3(b), there are two peaks at 1.60 A˚ and 2.42 A˚ (without phase-corrected), which respectively represents the first Co-O coordination shell and the second Co-Co coordination shell for cobalt oxide. It is consistent with the XRD patterns. The R space spectra of the spent Co-based catalysts are shown in Fig. 3(d). The EXAFS fitting results and the fitted Fourier transforms spectra are given in Table S1 and Figs. S3–S5. For the Co@MnO spent catalyst, the first peak with the coordination number of ∼3.0 at 1.92 A˚ corresponded to the Co-O contribution. The coordination number of the peak at 2.48 A˚ which should be assigned to the Co-Co coordination shell for metallic Co is 5.9, indicating that Co nanoparticles dominate in the Co@MnO spent catalyst. The peak at 1.90 A˚ is the Co-Co coordination shell of Co2 C. It can be observed that the peak at 1.90 A˚ of Co/Co2 C@MnO is around 0.03 A˚ bigger than that of Co2 C@MnO and the coordination number increases from 1.0 to 4.0, which further confirms that Co and Co2 C coexist for Co/Co2 C@MnO-spent catalyst.
These data were also consistent with the in-situ XRD, TEM and XANES measurements.
3.5. DFT calculations Density functional theory (DFT) was used to investigate the adsorption of H and CO on both Co and Co2 C surface. The most stable configurations were shown in Fig. 4. H prefers to be adsorbed on the 3-fold-fcc-hollow site on the Co (111) surface with an adsorption energy of −2.723 eV which is similar previous research [33]. The most stable adsorption site for H on Co2 C (101) was Cobridge site (Fig. 4, Schemes S1 and S3) with an adsorption energy of −2.091 eV which is much weaker than the adsorption on Co. Furthermore, we invested the most stable adsorption site for H and CO on Co-terminated Co2 C (101) with the adsorption energy of −0.669 eV and −2.130 eV (Scheme S4), respectively. Thus, the activation of hydrogen would be much easier on Co compared to Co2 C. However, the CO-adsorption energies were −2.106 eV and −1.629 eV on Co2 C and Co, respectively, that is, CO preferred to be adsorbed on Co2 C rather than Co.
232
F. Sun, R. Yang and Z. Xia et al. / Journal of Energy Chemistry 42 (2020) 227–232
Fig. 4. The most stable adsorption energies and their corresponding configurations.
4. Conclusions In summary, Co@MnO, Co/Co2 C@MnO and Co2 C@MnO catalysts were synthesized by co-precipitation method and characterized by in situ XRD, TEM, XANES, and EXAFS. The formation of Co2 C is a new active site of activating CO and essential for suppressing overhydrogenation and methane formation as revealed by both experimental data and DFT calculations. On the other hand, the Co metal is preferable for C–C coupling. A higher C–C coupling (C5+ ) is favored in the presence of both Co and cobalt carbide. We hypothesized that the different ratios between Co and Co2 C have a direct influence on the light olefins and C5+ selectivity during FT synthesis. This finding opens the door for further investigation of the industrial products selectivity of the Co-based catalysts in FTS. Acknowledgments This work has been supported from the National Natural Science Foundation of China, Grant/Award Number: U1732267, 21503218. Supplementary materials Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jechem.2019.07.007. References [1] [2] [3] [4] [5]
A.Y. Khodakov, W. Chu, P. Fongarland, Chem. Rev. 107 (2007) 1692–1744. Q.H. Zhang, J.C. Kang, Y. Wang, 2 (2010) 1030–1058. H. Schulz, Appl. Catal. A – Gen. 186 (1999) 3–12. D. Leckel, Energy Fuel 23 (2009) 2342–2358. F. Jiao, J.J. Li, X.L. Pan, J.P. Xiao, H.B. Li, H. Ma, M.M. Wei, Y. Pan, Z.Y. Zhou, M.R. Li, S. Miao, J. Li, Y.F. Zhu, D. Xiao, T. He, J.H. Yang, F. Qi, Q. Fu, X.H. Bao, Science 351 (2016) 1065–1068.
[6] H. Jiang, L.S. Wang, W. Cui, Y.D. Xu, Catal. Lett. 57 (1999) 95–102. [7] D. Ma, D.Z. Wang, L.L. Su, Y.Y. Shu, Y. Xu, X.H. Bao, J. Catal. 208 (2002) 260–269. [8] N. Ji, T. Zhang, M.Y. Zheng, A.Q. Wang, H. Wang, X.D. Wang, J.G.G. Chen, Angew. Chem. Int. Ed. 47 (2008) 8510–8513. [9] C. Yang, H.B. Zhao, Y.L. Hou, D. Ma, J. Am. Chem. Soc. 134 (2012) 15814–15821. [10] J. Li, L.T. Liu, Y. Liu, M.Z. Li, Y.H. Zhu, H.C. Liu, Y. Kou, J.Z. Zhang, Y. Han, D. Ma, Energy Environ. Sci. 7 (2014) 393–398. [11] S.T. Hunt, M. Milina, A.C. Alba-Rubio, C.H. Hendon, J.A. Dumesic, Y. Roman-Leshkov, Science 352 (2016) 974–978. [12] B. Jager, Nat. Gas Convers. V 119 (1998) 25–34. [13] P. Zhai, C. Xu, R. Gao, X. Liu, M.Z. Li, W.Z. Li, X.P. Fu, C.J. Jia, J.L. Xie, M. Zhao, X.P. Wang, Y.W. Li, Q.W. Zhang, X.D. Wen, D. Ma, Angew. Chem. Int. Ed. 55 (2016) 9902–9907. [14] R.O. Yang, Z.M. Xia, Z. Zhao, F.F. Sun, X.L. Du, H.S. Yu, S.Q. Gu, L.S. Zhong, J.T. Zhao, Y.J. Ding, Z. Jiang, J. Energy Chem. 32 (2019) 118–123. [15] Y.Y. Dai, T.J. Lin, S.G. Li, F. Yu, T.L. An, X.X. Wang, K. Xiao, F.F. Sun, Z. Jiang, Y.W. Lu, H. Wang, L.S. Zhong, Y.H. Sun, ACS Catal. 9 (2019) 798–809. [16] L. Zhong, F. Yu, Y. An, Y. Zhao, Y. Sun, Z. Li, T. Lin, Y. Lin, X. Qi, Y. Dai, L. Gu, J. Hu, S. Jin, Q. Shen, H. Wang, Nature 538 (2016) 84–91. [17] Y.P. Pei, J.X. Liu, Y.H. Zhao, Y.J. Ding, T. Liu, W.D. Dong, H.J. Zhu, H.Y. Su, L. Yan, J.L. Li, W.X. Li, ACS Catal. 5 (2015) 3620–3624. [18] Z. Zhao, W. Lu, R.O. Yang, H.J. Zhu, W.D. Dong, F.F. Sun, Z. Jiang, Y. Lyu, T. Liu, H. Du, Y.J. Ding, ACS Catal. 8 (2018) 228–241. [19] J. Anton, J. Nebel, H.Q. Song, C. Froese, P. Weide, H. Ruland, M. Muhler, S. Kaluza, J. Catal. 335 (2016) 175–186. [20] Y.Z. Xiang, N. Kruse, Nat. Commun. 7 (2016) 13058–13063. [21] V.M. Lebarbier, D.H. Mei, D.H. Kim, A. Andersen, J.L. Male, J.E. Holladay, R. Rousseau, Y. Wang, J. Phys. Chem. C 115 (2011) 17440–17451. [22] Z. Wang, J.J. Spivey, Appl. Catal. A – Gen. 507 (2015) 75–81. [23] M.K. Gnanamani, G. Jacobs, U.M. Graham, M.C. Ribeiro, F.B. Noronha, W.D. Shafer, B.H. Davis, Catal. Today 261 (2016) 40–47. [24] Z. Zhao, W. Lu, C.H. Feng, X.K. chen, H.J. Zhu, R.O. Yang, W.D. Dong, M. Zhao, Y. Lyu, T. Liu, Z. Jiang, Y.J. Ding, J. Catal. 370 (2019) 251–264. [25] W.D. Dong, J.X. Liu, H.J. Zhu, Y.J. Ding, Y.P. Pei, J. Liu, H. Du, M. Jiang, T. Liu, H.Y. Su, W.X. Li, J. Phys. Chem. C 118 (2014) 19114–19122. [26] H. Yu, J. Li, S. Gu, S. Zhang, Nucl. Sci. Technol. 26 (2015) 4–10. [27] J.C. Mohandas, M.K. Gnanamani, G. Jacobs, W.P. Ma, Y.Y. Ji, S. Khalid, B.H. Davis, ACS Catal. 1 (2011) 1581–1588. [28] M. Newville, J. Synchrotron. Radiat. 8 (2001) 322–324. [29] J. Klimes, D.R. Bowler, A. Michaelides, J. Phys. Condens. Matter 22 (2010) 022201–022205. [30] Y.H. Zhao, H.Y. Su, K.J. Sun, J.X. Liu, W.X. Li, Surf. Sci. 606 (2012) 598–604. [31] Z.J. Zuo, W. Huang, P.D. Han, Z.H. Li, Appl. Surf. Sci. 256 (2010) 5929–5934.
Contents lists available at ScienceDirect
Journal of Energy Chemistry journal homepage: www.elsevier.com/locate/jechem
Effects of cobalt carbide on Fischer–Tropsch synthesis with MnO supported Co-based catalysts Fanfei Sun a,b,c,1, Ruoou Yang b,1, Zhaoming Xia d, Yuqi Yang a,b,c, Ziang Zhao e, Songqi Gu a,b, Dongshuang Wu f,∗, Yunjie Ding e,∗, Zheng Jiang a,b,∗∗ a
Shanghai Institute of Applied Physics, Chinese Academy of Sciences, Shanghai 201204, China Shanghai Synchrotron Radiation Facility, Zhangjiang Laboratory, Shanghai Advanced Research Institute, Chinese Academy of Sciences, Shanghai 201210, China c University of Chinese Academy of Sciences, Beijing 100100, China d Center for Applied Chemical Research, Frontier Institute of Science and Technology, Xi’an Jiaotong University, Xi’an 710049, Shaanxi, China e Dalian National Laboratory for Clean Energy, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, Liaoning, China f Graduate school of science, Kyoto University, Kitashirakawa, Oiwake-cho, Sakyo-ku, Kyoto-shi, Kyoto-fu, 606-8502, Japan b
a r t i c l e
i n f o
Article history: Received 18 April 2019 Revised 27 June 2019 Accepted 11 July 2019 Available online 15 July 2019 Keywords: Cobalt carbide Fischer–Tropsch synthesis X-ray absorption fine spectroscopy nanomaterials Density functional theory
a b s t r a c t Cobalt carbide (Co2 C) was considered as potential catalysts available for large-scale industrialization of transforming syngas (H2 and CO) to clean fuels. Herein, we successfully synthesized Co-based catalysts with MnO supported, to comprehend the effects of Co2 C for Fischer–Tropsch synthesis (FTS) under ambient conditions. The huge variety of product selectivity which was contained by different active sites (Co and Co2 C) has been found. Furthermore, density functional theory (DFT) shows that Co2 C is efficacious of CO adsorption, whereas is weaker for H adsorption than Co. Combining the advantages of Co and Co2 C, the catalyst herein can not only obtain more C5+ products but also suppress methane selectivity. It can be a commendable guide for the design of industrial application products in FTS. © 2019 Science Press and Dalian Institute of Chemical Physics, Chinese Academy of Sciences. Published by Elsevier B.V. and Science Press. All rights reserved.
1. Introduction Fischer–Tropsch synthesis (FTS), a significant step in the transformation of syngas (a mixture of H2 and CO) to clean fuels or value-added chemicals, is of both industrial and fundamental importance [1–4]. The challenge of selective formation of available hydrocarbons through FTS lies in preciously controlling C–C coupling by adjusting hydrogenation and suppressing methane formation under reaction [5]. To this end, catalysts with both high activity and selectivity are highly desirable to obtain products on-demand. Transitional metal carbides (TMC) have been emerged as a promising class of catalytic materials in FTS because their metallic-like properties [6–11]. Among TMC, iron carbide repre-
∗
Corresponding authors. Corresponding author at: Shanghai Institute of Applied Physics, Chinese Academy of Sciences, Shanghai 201204, China. E-mail addresses: [email protected] (D. Wu), [email protected] (Y. Ding), [email protected] (Z. Jiang). 1 These authors contributed equally to this work. ∗∗
sents a leading catalyst to afford long-chain hydrocarbons due to its high stability, per pass conversion and hydrocarbon selectivity [12,13]. In contract, its analog, cobalt carbide, has long been considered as an unwanted byproduct which results in the activity loss of the metallic Co active center during FTS [14]. Quite recently, the role of Co2 C has been turned. For example, Zhong et al. revealed a high relevancy among the particle size, the exposed facets of Co2 C and catalytic performance [15]. The specific Co2 C surface exhibited excellent selectivity towards the formation of lower olefins with low selectivity towards methane production [16]. Ding et al. synthesized stable and efficient Co-Co2 C catalysts [17] and proposed that Co-Co2 C interfaces have synergistic effects and are responsible for the formation of alcohols [18]. Although tremendous efforts have been made to understand the effect of the cobalt carbide, due to the tanglesome reactions in FTS, it is rather difficult to elucidate the formation process of the active site and the characterization of the catalysts [17,19–23]. Herein, we used co-precipitation method to successfully synthesize Co/Co2 C@MnO catalyst which performed different selectivities with Co@MnO and Co2 C@MnO catalysts under ambient reaction conditions (250 °C, 0.1 MPa). TEM and in-situ XRD experiments
https://doi.org/10.1016/j.jechem.2019.07.007 2095-4956/© 2019 Science Press and Dalian Institute of Chemical Physics, Chinese Academy of Sciences. Published by Elsevier B.V. and Science Press. All rights reserved.
228
F. Sun, R. Yang and Z. Xia et al. / Journal of Energy Chemistry 42 (2020) 227–232
confirmed the formation of coexisted Co and Co2 C. The atomic structure of the Co/Co2 C@MnO catalyst was studied by X-ray absorption fine structure (XAFS). Based on the DFT analysis of the Co-based catalysts, we further understood the catalytic mechanism of the cobalt carbide and provided guidance for the molecular design of the catalytic performance of the Co-based catalysts in FTS. 2. Experimental
temperature was further increase into 250 at a heating speed of 0.5 °C/min. At the outlet of the reactor, we use a condenser to separate the gas and liquid products. The catalytic performances were analyzed after stabilization for 12 h. Agilent 6890 N GC gas chromatograph with a HP-5 capillary column and a flame ionization detector (FID) was used to analyze products. The water-bearing products and oil-bearing products were analyzed by butyl with internal standard method and area normalization method, respectively.
2.1. Synthesis of catalyst
2.4. DFT calculations
Co@MnO catalyst was synthesized by co-precipitation method refer to our previous work [14]. In a typical synthesis, an aqueous solution containing cobalt nitrate (Co(NO3 )2 ·6H2 O); Sinopharm) and manganese nitrate (50 wt% Mn(NO3 )2 , Sinopharm) was prepared (2 M, (mol/mol) = 2/1), and Na2 CO3 aqueous solution (2 M) was added. The mixed solution was vigorously stirring at RT. Following, the formed suspension went through a procedure of filtration, repeated washing process and dried overnight at 100 °C. Finally, we calcined the dried powder at 330 °C for 3 h in the air. When the solution precipitant was replaced by the mix of (NH4 )2 CO3 and Na2 CO3 solution (NH/Na (mol/mol) = 1/1), we obtained Co/Co2 C@MnO. If the solution precipitant was replaced by only Na2 CO3 solution, Co2 C@MnO was obtained.
Spin-polarized plane-wave basis set density functional theory (DFT) as implemented in the Vienna ab initio Simulation Package (VASP) [29]. Perdew–Burke–Ernzerhof (PBE) [29] generalized gradient approximation (GGA) [30] functional was used to evaluate the exchange-correlation energy. The energy and force convergence were 10−5 eV and 0.02 eV, respectively. The energy cut off was 400 eV. A 16 × 16 × 16 Monkhorst–Pack k-point mesh was employed in the calculations of bulk properties. The lattice volume was fitted with the third-order Birch–Murnaghan isothermal equation of state [31]. The resulted lattice parameters of FCC ˚ b = 4.361 A˚ Co and Co2 C are a = b = c = 3.457 A˚ and a = 2.850 A, ˚ respectively. For the calculation of CO and H2 and c = 4.345 A, molecule, a 20 × 20 × 20 unit cell and 1 × 1 × 1 k-point mesh were employed. A 15 A˚ vacuum region was applied to separate slabs in the calculation of slab model. We only investigate the most stable surface of Co2 C, which is Co2 C (101) as is reported previously [32]. A p(2 × 2) unit cells for Co2 C (101) were utilized in our calculation, which contains 4 layers of Co and 2 layers of C. The bottom 2 Co layers and one C layer were fixed, and other layers were fully relaxed. A p(3 × 3) unit cell for FCC Co (111) was also applied in the calculation. It contained three Co layers with the bottom one layer fixed. (Scheme S1) Monkhorst-Pack mesh k-points of (3 × 3 × 1), (3 × 4 × 1) were used for FCC Co (111), Co2 C (101)-Cterminated and Co2 C (101)-Co-terminated surface calculations, respectively. Dipole corrections have been applied in the calculations to avoid artificial dipole effects. Since CO and H adsorptions on Co (111) were well studied, we only consider the most stable adsorption site. H was on the 3-fold-fcc-hollow site and CO was on the top of Co atom. In the case of adsorption on Co2 C–C-terminated, four sites were taken into consideration, as was shown in Scheme S1. However, some of them were not favorable for CO/H adsorption. There were only three distinguishable configurations for CO adsorption (Scheme S2) and two for H adsorption (Scheme S3). The most stable adsorption site for H and CO on Co2 C (101)-Coterminated was shown in Scheme S4. The adsorption energy was defined as:
2.2. Characterization 2.2.1. In situ X-ray diffraction (XRD) analysis In situ powder X-ray diffraction measurement and analysis for Fischer–Tropsch reaction were referred to our previous work [24]. The phase transformation of the calcined samples was monitored during reduction process at 0.1 MPa in a H2 flux of 20 mL/min with a thermal rate of 1 °C/min from RT to 350 °C and the XRD patterns were collected every 20 °C. Then, the samples were kept at 350 °C about 5 h and the XRD spectra were collected every 30 min at 350 °C (Fig. S1). Afterwards, the cell temperature was cooled down to 250 °C, at which the catalytic behavior was monitored under 0.1 MPa syngas (H2 /CO=2/1) at a flux rate of 20 mL/min (Fig. 1). We collected the XRD spectra every 30 min during the FT process. Each scan was performed at 40 kV and 40 mA with a scanning angle range of 30°–80° at a rate of 2 °/min. All the XRD spectra were carried out by the X’pert HighScore Plus software package with the PDF library. 2.2.2. Transmission electron micrograph (TEM) The morphology of the spent catalysts was observed by Tecnai G2 S-TWIN instrument with an accelerating voltage of 200 kV [25]. 2.2.3. X-ray absorption fine spectroscopy (XAFS) Co K-edge XAFS data were measured in transmission mode at beamline BL14W1 of Shanghai Synchrotron Radiation Facility (SSRF) [26]. The extended X-ray absorption fine structure (EXAFS) data analyses were carried out by using the Demeter software package [27,28]. The Fourier transformation of the k3 -weighted functions was applied to the EXAFS data. The fitting of the EXAFS data was calculated by Artemis program in Ifeffit. 2.3. Catalytic test Catalysts were tested in a fixed-bed stainless steel plug-flow reactor with inner diameter 9 mm refer to our previous work [14]. In a typical test, ca. 2 mL catalyst and 2 mL SiO2 were mixed to obtain isothermal plug flow conditions. The reactor was filled with 0.1 MPa H2 at RT and heated to 175 °C at a rate of 5 °C/min and then the gas atmosphere was changed into syngas (H2 /CO = 2). The
Ead (H ) = Eslab+H − Eslab − EH2 /2 Ead (CO ) = Eslab+CO − Eslab − Eco Ead (H) was the adsorption energy of H. Eslab was the energy of Co (111) or Co2 C(101) slab model. EH2 was the energy of H2 . Eslab+H was the energy of slab model with one H atom. Ead (CO) was the adsorption energy of CO. ECO was the energy of CO. Eslab+CO was the energy of slab model with CO adsorbed on it. 3. Results and discussion 3.1. In situ XRD Despite of different preparing methods, the in-situ XRD results recorded during H2 reduction suggest that all the three catalysts
F. Sun, R. Yang and Z. Xia et al. / Journal of Energy Chemistry 42 (2020) 227–232
229
Fig. 1. In situ XRD pattern for Co@MnO (a), Co/Co2 C@MnO (b) and Co2 C@MnO (c) catalysts in a syngas flow of 6 h (PDF number: MnO (0 0-0 04-0326), Cox Mn1− x O (0 0-0 040326), fcc Co (0 0-0 01-1259) and Co3 O4 (0 0-0 01-1152)).
are composed by a mixture of MnO, metallic Co and Cox Mn1 −x O after the pre-treatment process (Fig. S1). Fig. 1 shows the insitu XRD patterns of the catalysts in syngas. For Co@MnO catalyst, the peaks belonging to face center cubic (FCC) Co and MnO existing in the whole 6 h syngas reaction and Co2 C phase cannot be observed (Fig. 1(a)). For Co/Co2 C@MnO catalyst, in addition to the FCC Co peak, Co2 C peaks were shown in a syngas flow at 250 °C (Fig. 1(b)), which can be assigned as 2θ = 37.0°, 41.3°, 42.6°, 45.8°, and 56.6° corresponding to the crystalline plane of Co2 C (110), (002), (111), (021), and (112) according to PDF 01-0721369, respectively. The formed Co2 C was very stable even after a 10 h syngas gas reaction (Fig. S2). It suggests the co-existence of FCC Co and Co2 C in Co/Co2 C@MnO catalyst. As for Co2 C@MnO catalyst, after the syngas feed was introduced, the cobalt metal gradually turned into Co2 C with time on-stream. Meanwhile, the single phase (Cox Mn1 −x O) segregated into FCC Co and MnO, respectively. After 4 h of reaction, the structure was stable and the metallic cobalt completely transformed into Co2 C. It can be clearly seen that Co2 C formed in Co/Co2 C@MnO and Co2 C@MnO, which suggests the importance role of Na2 CO3 in the formation of Co2 C.
3.2. TEM TEM and HRTEM images of the three spent catalysts are shown in Fig. 2. All are near spherical aggregated nanoparticles with quite a similar average particle size approximating 10 nm (Fig. 2(a)–(c)). The plane distance with ca. 1.53 A˚ can be assigned as {220} plane of MnO, which is observed in all three samples. A plane distance of ca. 2.01 A˚ and ca. 2.40 A˚ corresponds to {111} plane of Co and {101} plane of Co2 C, respectively. As shown in Fig. 2(d) and (f) only Co was observed in Co@MnO and only Co2 C existed in Co2 C@MnO. While, in the Co/Co2 C@MnO catalyst, the coexistence of the FCC cobalt metal and the cobalt carbide phase was revealed (Fig. 2(e)). The TEM measurements correspond to the result of in-situ XRD analysis and suggest that the active center in three catalysts might be different, that is, Co, Co/Co2 C, and Co2 C, respectively. 3.3. Catalytic performance In order to understand catalytic mechanism, the catalytic activity of Co-based catalysts is presented. The hydrocarbons and
230
F. Sun, R. Yang and Z. Xia et al. / Journal of Energy Chemistry 42 (2020) 227–232
Fig. 2. TEM images of the Co@MnO (a) and (d), Co/Co2 C@MnO (b) and (e) and Co2 C@MnO (c) and (f) catalysts.
Table 1. Catalytic performance of the Co-based catalyst samples. Product selectivity (C%, CO2 -free) Catalyst
CO content (%)
CO2 selectivity (C%)
CH4
C2–4=
C2–4o
C5+
Oxy
Co@MnO Co/Co2 C@MnO Co2 C@MnO
54.4 55.5 52.2
10.3 32.7 44.2
37.5 18.3 18.1
18.6 22.9 39.2
12.7 15.2 18.8
25.9 42.5 22.3
5.3 1.1 1.6
Reaction conditions: 250 °C, 30 0 0 mL h−1 gcat −1 (where gcat denotes grams of catalyst). Oxy: oxygenates.
alcohols are all possible products through FTS. The catalytic performances of the prepared catalysts during FT reaction were performed at 250 °C, 0.1 MPa, and GHSV = 30 0 0 mLsyngas /(gcat − h) with H2 /CO = 2 in a fixed-bed reactor. The variations in CO conversion and product selectivity for the spent catalysts are listed in Table 1. The CO conversion is similar for all the catalysts. However, we observed diverse product selectivities over these catalysts. Specifically, the selectivities of methane for Co/Co2 C@MnO and Co2 C@MnO are only 18.1% and 18.3%, respectively, which are much lower than that for Co@MnO (37.5%). Moreover, the production of lower olefins is 39.2% on Co2 C@MnO which is higher than those of Co@MnO (18.6%) and Co/Co2 C@MO (22.9%). These results suggest that the Co2 C also can activate CO like metallic cobalt, however, it suppresses methane formation by suppressing the over-hydrogenation. Nevertheless, the highest C5+ selectivity is observed on Co/Co2 C@MnO (42.5%), while, the other two catalysts show only around half of the value that is 25.9% and 22.3% for Co@MnO and Co2 C@MnO, respectively. This result suggests that a higher C–C coupling is favored in the presence of both Co and cobalt carbide. In addition, the effect of MnO is removed by catalytic performance (Table S3) and water gas shift reaction (Table S4) in this FTS condition.
3.4. XAFS measurement X-ray Absorption Near Edge Spectroscopy (XANES) analysis is used to distinguish the structure and valence state of the Co-based catalysts in Fig. 3. Fig. 3(a) shows the Co K-edge normalized XANES spectra of the calcined Co-based catalysts and a comparison of Co, Co3 O4 and Co2 C [14] reference compounds. As the ion charge corresponds to the transition energy, the edge position shifts toward lower energies with oxidation states: 7709.8 eV and 7729.0 eV for Co and Co3 O4 , respectively. All the spectrums of the calcined Co-based catalysts exhibit oxidation states similar to the Co3 O4 reference. Fig. 3(c) exhibits the Co K-edge XANES profile in the spent catalysts and reference compounds for comparison. A peak can be found at the pre-edge at 7712 eV for Co/Co2 C@MnO and Co2 C@MnO, due to the consist of the non-centrosymmetric Co2 C in catalysts. The absorption edge around in the spent Co-based catalysts sample showed a medium signal between the Co reference and CoO, which revealed that the valence state is lower than +2 in the spent catalysts. Moreover, the ratio of Co2 C to Co of Co/Co2 C@MnO catalyst is 1.68 (calculated by linear combination fitting (LCF) listed in Table S2), which impacts methane and C5+ selectivity on the Co-based catalysts.
F. Sun, R. Yang and Z. Xia et al. / Journal of Energy Chemistry 42 (2020) 227–232
231
Fig. 3. XAFS characterization of Co-based catalyst samples. (a) Cobalt K-edge (7709 eV) XANES spectra of calcined Co-based catalyst samples; (b) Fourier transform spectra of the calcined Co-based catalyst samples; (c) Cobalt K-edge (7709 eV) XANES spectra of spent Co-based catalyst samples; (d) Fourier transform spectra of the spent Co-based catalyst samples.
To evaluate the local geometry around Co atoms, both the calcined and the spent catalysts were collected by extended X-ray Absorption Fine Structure Spectroscopy (EXAFS). The k3 -weighted EXAFS Fourier transform magnitude spectra of the calcined and spent Co-based catalysts are provided in Fig. 3(b) and (d). In Fig. 3(b), there are two peaks at 1.60 A˚ and 2.42 A˚ (without phase-corrected), which respectively represents the first Co-O coordination shell and the second Co-Co coordination shell for cobalt oxide. It is consistent with the XRD patterns. The R space spectra of the spent Co-based catalysts are shown in Fig. 3(d). The EXAFS fitting results and the fitted Fourier transforms spectra are given in Table S1 and Figs. S3–S5. For the Co@MnO spent catalyst, the first peak with the coordination number of ∼3.0 at 1.92 A˚ corresponded to the Co-O contribution. The coordination number of the peak at 2.48 A˚ which should be assigned to the Co-Co coordination shell for metallic Co is 5.9, indicating that Co nanoparticles dominate in the Co@MnO spent catalyst. The peak at 1.90 A˚ is the Co-Co coordination shell of Co2 C. It can be observed that the peak at 1.90 A˚ of Co/Co2 C@MnO is around 0.03 A˚ bigger than that of Co2 C@MnO and the coordination number increases from 1.0 to 4.0, which further confirms that Co and Co2 C coexist for Co/Co2 C@MnO-spent catalyst.
These data were also consistent with the in-situ XRD, TEM and XANES measurements.
3.5. DFT calculations Density functional theory (DFT) was used to investigate the adsorption of H and CO on both Co and Co2 C surface. The most stable configurations were shown in Fig. 4. H prefers to be adsorbed on the 3-fold-fcc-hollow site on the Co (111) surface with an adsorption energy of −2.723 eV which is similar previous research [33]. The most stable adsorption site for H on Co2 C (101) was Cobridge site (Fig. 4, Schemes S1 and S3) with an adsorption energy of −2.091 eV which is much weaker than the adsorption on Co. Furthermore, we invested the most stable adsorption site for H and CO on Co-terminated Co2 C (101) with the adsorption energy of −0.669 eV and −2.130 eV (Scheme S4), respectively. Thus, the activation of hydrogen would be much easier on Co compared to Co2 C. However, the CO-adsorption energies were −2.106 eV and −1.629 eV on Co2 C and Co, respectively, that is, CO preferred to be adsorbed on Co2 C rather than Co.
232
F. Sun, R. Yang and Z. Xia et al. / Journal of Energy Chemistry 42 (2020) 227–232
Fig. 4. The most stable adsorption energies and their corresponding configurations.
4. Conclusions In summary, Co@MnO, Co/Co2 C@MnO and Co2 C@MnO catalysts were synthesized by co-precipitation method and characterized by in situ XRD, TEM, XANES, and EXAFS. The formation of Co2 C is a new active site of activating CO and essential for suppressing overhydrogenation and methane formation as revealed by both experimental data and DFT calculations. On the other hand, the Co metal is preferable for C–C coupling. A higher C–C coupling (C5+ ) is favored in the presence of both Co and cobalt carbide. We hypothesized that the different ratios between Co and Co2 C have a direct influence on the light olefins and C5+ selectivity during FT synthesis. This finding opens the door for further investigation of the industrial products selectivity of the Co-based catalysts in FTS. Acknowledgments This work has been supported from the National Natural Science Foundation of China, Grant/Award Number: U1732267, 21503218. Supplementary materials Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jechem.2019.07.007. References [1] [2] [3] [4] [5]
A.Y. Khodakov, W. Chu, P. Fongarland, Chem. Rev. 107 (2007) 1692–1744. Q.H. Zhang, J.C. Kang, Y. Wang, 2 (2010) 1030–1058. H. Schulz, Appl. Catal. A – Gen. 186 (1999) 3–12. D. Leckel, Energy Fuel 23 (2009) 2342–2358. F. Jiao, J.J. Li, X.L. Pan, J.P. Xiao, H.B. Li, H. Ma, M.M. Wei, Y. Pan, Z.Y. Zhou, M.R. Li, S. Miao, J. Li, Y.F. Zhu, D. Xiao, T. He, J.H. Yang, F. Qi, Q. Fu, X.H. Bao, Science 351 (2016) 1065–1068.
[6] H. Jiang, L.S. Wang, W. Cui, Y.D. Xu, Catal. Lett. 57 (1999) 95–102. [7] D. Ma, D.Z. Wang, L.L. Su, Y.Y. Shu, Y. Xu, X.H. Bao, J. Catal. 208 (2002) 260–269. [8] N. Ji, T. Zhang, M.Y. Zheng, A.Q. Wang, H. Wang, X.D. Wang, J.G.G. Chen, Angew. Chem. Int. Ed. 47 (2008) 8510–8513. [9] C. Yang, H.B. Zhao, Y.L. Hou, D. Ma, J. Am. Chem. Soc. 134 (2012) 15814–15821. [10] J. Li, L.T. Liu, Y. Liu, M.Z. Li, Y.H. Zhu, H.C. Liu, Y. Kou, J.Z. Zhang, Y. Han, D. Ma, Energy Environ. Sci. 7 (2014) 393–398. [11] S.T. Hunt, M. Milina, A.C. Alba-Rubio, C.H. Hendon, J.A. Dumesic, Y. Roman-Leshkov, Science 352 (2016) 974–978. [12] B. Jager, Nat. Gas Convers. V 119 (1998) 25–34. [13] P. Zhai, C. Xu, R. Gao, X. Liu, M.Z. Li, W.Z. Li, X.P. Fu, C.J. Jia, J.L. Xie, M. Zhao, X.P. Wang, Y.W. Li, Q.W. Zhang, X.D. Wen, D. Ma, Angew. Chem. Int. Ed. 55 (2016) 9902–9907. [14] R.O. Yang, Z.M. Xia, Z. Zhao, F.F. Sun, X.L. Du, H.S. Yu, S.Q. Gu, L.S. Zhong, J.T. Zhao, Y.J. Ding, Z. Jiang, J. Energy Chem. 32 (2019) 118–123. [15] Y.Y. Dai, T.J. Lin, S.G. Li, F. Yu, T.L. An, X.X. Wang, K. Xiao, F.F. Sun, Z. Jiang, Y.W. Lu, H. Wang, L.S. Zhong, Y.H. Sun, ACS Catal. 9 (2019) 798–809. [16] L. Zhong, F. Yu, Y. An, Y. Zhao, Y. Sun, Z. Li, T. Lin, Y. Lin, X. Qi, Y. Dai, L. Gu, J. Hu, S. Jin, Q. Shen, H. Wang, Nature 538 (2016) 84–91. [17] Y.P. Pei, J.X. Liu, Y.H. Zhao, Y.J. Ding, T. Liu, W.D. Dong, H.J. Zhu, H.Y. Su, L. Yan, J.L. Li, W.X. Li, ACS Catal. 5 (2015) 3620–3624. [18] Z. Zhao, W. Lu, R.O. Yang, H.J. Zhu, W.D. Dong, F.F. Sun, Z. Jiang, Y. Lyu, T. Liu, H. Du, Y.J. Ding, ACS Catal. 8 (2018) 228–241. [19] J. Anton, J. Nebel, H.Q. Song, C. Froese, P. Weide, H. Ruland, M. Muhler, S. Kaluza, J. Catal. 335 (2016) 175–186. [20] Y.Z. Xiang, N. Kruse, Nat. Commun. 7 (2016) 13058–13063. [21] V.M. Lebarbier, D.H. Mei, D.H. Kim, A. Andersen, J.L. Male, J.E. Holladay, R. Rousseau, Y. Wang, J. Phys. Chem. C 115 (2011) 17440–17451. [22] Z. Wang, J.J. Spivey, Appl. Catal. A – Gen. 507 (2015) 75–81. [23] M.K. Gnanamani, G. Jacobs, U.M. Graham, M.C. Ribeiro, F.B. Noronha, W.D. Shafer, B.H. Davis, Catal. Today 261 (2016) 40–47. [24] Z. Zhao, W. Lu, C.H. Feng, X.K. chen, H.J. Zhu, R.O. Yang, W.D. Dong, M. Zhao, Y. Lyu, T. Liu, Z. Jiang, Y.J. Ding, J. Catal. 370 (2019) 251–264. [25] W.D. Dong, J.X. Liu, H.J. Zhu, Y.J. Ding, Y.P. Pei, J. Liu, H. Du, M. Jiang, T. Liu, H.Y. Su, W.X. Li, J. Phys. Chem. C 118 (2014) 19114–19122. [26] H. Yu, J. Li, S. Gu, S. Zhang, Nucl. Sci. Technol. 26 (2015) 4–10. [27] J.C. Mohandas, M.K. Gnanamani, G. Jacobs, W.P. Ma, Y.Y. Ji, S. Khalid, B.H. Davis, ACS Catal. 1 (2011) 1581–1588. [28] M. Newville, J. Synchrotron. Radiat. 8 (2001) 322–324. [29] J. Klimes, D.R. Bowler, A. Michaelides, J. Phys. Condens. Matter 22 (2010) 022201–022205. [30] Y.H. Zhao, H.Y. Su, K.J. Sun, J.X. Liu, W.X. Li, Surf. Sci. 606 (2012) 598–604. [31] Z.J. Zuo, W. Huang, P.D. Han, Z.H. Li, Appl. Surf. Sci. 256 (2010) 5929–5934.