Silicon carbide supported cobalt for Fischer–Tropsch synthesis: probing into the cause of the intrinsic excellent catalytic performance

Abstract

The thin Si_xO_y layer on a SiC surface is changed to Al₂O₃ to form Al₂O₃@SiC. Co/Al₂O₃@SiC shows distinct different catalytic behaviour with Co/SiC, indicating that the Si_xO_y layer on the surface of SiC plays a great role in the intrinsic excellent catalytic performance of Co/SiC.

---

Fischer–Tropsch synthesis (FTS) is a key technology in the gas-to-liquids (GTL) process, which transforms synthesis gas derived from natural gas, coal or biomass into high quality ultra-clean fuels or valuable chemicals.^1–3 The FTS is a high exothermic reaction, which could cause an adiabatic temperature rise of up to 1750 K.⁴ High temperature gradient might exist within catalyst bed, which could threaten the safety of the plant particularly for large scale commercial production, provided that the generated heat could not be removed effectively.^5,6

Recently, high thermal conductive silicon carbide (β-SiC) has been reported as support for FTS and the prepared catalyst exhibits excellent catalytic performance, that is, high C₅₊ (>90%) and low CH₄ selectivity.^7–11 According to the most accepted viewpoint at present, two factors are responsible for the good catalytic performance of Co/SiC. On one hand, it is assumed that the heat generated during reaction could not be completely evacuated to the entire body of traditional insulated supports (Al₂O₃, SiO₂, TiO₂), leading to the formation of “hot spots” on the surface of catalyst, which favors the formation of light products.⁸ For Co/SiC, the high thermal conductivity of SiC facilitates heat dissipation throughout the catalyst body, avoiding the formation of “hot spots” on catalyst surface, thus enhances the selectivity to long-chain hydrocarbons.¹⁰ However, no research has verified the correctness of the explanation at present. On the other hand, it is reported that the meso–macroporous structure of SiC could significantly enhance intraparticle mass transfer during FTS, which favors the production of heavy hydrocarbons.⁸ However, many research results, both experimental and computational have pointed out that the intraparticle mass transfer limitation could only affect the intrinsic reactive behavior for those catalysts with large pellet sizes.^12,13 Therefore, the pore structure effect on powdered Co/SiC catalysts should be negligible (Fig. S1 and Table S1, ESI†). On the silicon carbide surface, there is a thin amorphous Si_xO_y (3.5 ± 0.5 wt%) layer.^14,15 No research has focused on the effect of the Si_xO_y layer on the FTS performance so far. In brief, the traditional explanations for the remarkable catalytic performance of Co/SiC are not very persuasive.

Actually, it is of great significance to probe into the cause of the intrinsic excellent catalytic performance for Co/SiC, as the obtained research findings could be utilized to guide the preparation of FTS catalysts with remarkable performance similar to Co/SiC. Herein, we have systematically investigated the possible reasons for the good performance of Co/SiC. SiC support was treated with NaOH to remove the Si_xO_y layer, the obtained material was denoted as R-SiC. Then the prepared R-SiC was coated a layer of Al₂O₃ using precipitation method, the obtained sample was denoted as Al₂O₃@SiC. In order to obtain similar surface chemical nature with Al₂O₃@SiC, traditional low thermal conductive Al₂O₃ support was also prepared by precipitation method. Co/Al₂O₃ and Co/Al₂O₃@SiC catalysts are used to investigate the effect of thermal conductivity on FTS performance. Co/SiC and Co/Al₂O₃@SiC catalysts are used to investigate the effect of the Si_xO_y layer on FTS performance. Co/Al catalyst is also used as comparison because of the similar structure with SiC.

The content of Al₂O₃ on Al₂O₃@SiC is 3.8 wt%, which is close to the content of amorphous Si_xO_y on SiC (Table S2, ESI†). The specific surface area of Al₂O₃@SiC is a little higher than SiC, which could be attributed to the introduction of Al₂O₃ layer on the surface (Table S3, ESI†). No characteristic diffraction patterns of Al₂O₃ are detected for Al₂O₃@SiC and Al powders, which might be attributed to the lower content of alumina or the alumina particles are too small to be detected (Fig. S2, ESI†). From NH₃-TPD result, two peaks at around 530 and 680 K are observed for Al₂O₃@SiC and Al powders, indicating the existence of Al₂O₃ phase. However, no obvious NH₃ desorption peak is observed for SiC, indicating that SiC is a non-acid support (Fig. S3, ESI†). XPS technique is also used to investigate the surface character of the supports. The peak centered at 74.2 eV can be observed for Al₂O₃@SiC and Al powders, which is assigned to the Al 2p of Al₂O₃ (Fig. S4, ESI†).¹⁶ This suggests that Al₂O₃ phase is located on the surface of Al₂O₃@SiC and Al powders. The morphology of the support materials has been studied by TEM. As shown in Fig. 1, there is a thin amorphous layer with a thickness of about 2.21 nm on the surface of SiC, which could be Si_xO_y layer. After treating with NaOH, the amorphous layer could not be seen on R-SiC any more, demonstrating that the amorphous Si_xO_y layer is removed. For Al₂O₃@SiC, a new layer with a thickness of about 1.31 nm can be observed on the surface of Al₂O₃@SiC. According to the XPS result talked above, Al₂O₃ phase is located on the surface of Al₂O₃@SiC, thus the new formed layer must be Al₂O₃. In addition, from the low resolution TEM image (Fig. S5, ESI†) we can observe that the new Al₂O₃ layer is homogeneously coated on Al₂O₃@SiC surface. Likewise, a thin layer of Al₂O₃ is also observed on the Al powders from high resolution TEM image. It should be mentioned that the structure of Al powders is similar with SiC and Al₂O₃@SiC, that is, a thin Al₂O₃ layer is formed on the outer surface and underneath is high thermal conductive metallic Al.

Fig. 1 Typical HRTEM images of the materials: (a) SiC; (b) R-SiC; (c) Al₂O₃@SiC; (d) Al powders.

The spinel Co₃O₄ is the only crystalline cobalt species in the fresh catalysts according to the XRD result (Fig. S6†).¹⁷ Table 1 shows the textural and chemical properties of the prepared catalysts. The Co₃O₄ particle sizes of the catalysts are in the range of 15.4–24.2 nm, which are calculated from Scherrer equation.¹⁸ The corresponding Co⁰ crystalline sizes are in the range of 11.6–18.2 nm. Co⁰ crystalline sizes are also calculated from H₂ chemisorption result, which is in line with the XRD result. It is reported that the cobalt particle sizes smaller than 6 nm can significantly affect the intrinsic selectivity and activity of FTS.¹⁹ As can be seen, the Co⁰ crystalline sizes of the prepared catalysts are all larger than 10 nm, indicating that the cobalt sizes effect should be negligible.

Table 1 Textural and chemical properties of the prepared catalysts

Catalyst	B.E.T. (m^2 g^−1)	PD (nm)	PV (nm)	XRD (nm)		H2 ads.^c	d(Co^0)^d (nm)	R^e (%)
				d(Co3O4)^a	d(Co^0)^b
a The average particle size of Co3O4 in the calcined catalysts was calculated from the most intense Co3O4 line (2θ = 36.8°).b d(Co^0) = 0.75d(Co3O4).c H2 ads. in μmolH2 gcat^−1.d Calculated from H2 chemisorption.e Reducibility calculated by TPR from 400 to 673 K.
10Co/SiC	28.6	14.2	0.12	24.2	18.2	29.6	19.6	73
10Co/Al	16.3	17.9	0.08	19.5	14.6	26.5	15.1	59
10Co/Al2O3@SiC	33.5	13.5	0.12	21.2	15.9	31.9	14.3	56
15Co/Al2O3	160.1	4.1	0.27	15.4	11.6	41.2	10.4	35

---

The H₂-TPR profiles of the calcined catalysts are shown in Fig. 2A. Three main reduction peaks centered at approximately 620, 661 and 723 K are observed for Co/SiC. The first two peaks are typically assigned to the two-step reduction of Co₃O₄ to CoO and CoO to Co, the third peak could be attributed to the reduction of bulk cobalt species.^20,21 The H₂-TPR profile shapes of Co/Al₂O₃@SiC and Co/Al are similar but different with Co/SiC, indicating that the surface species layer on SiC could affect the reduction behavior of Co₃O₄. The second reduction peak of Co/Al₂O₃@SiC shifts to higher temperature than Co/SiC. This suggests that the interaction between Co₃O₄ and support becomes stronger after changing the surface Si_xO_y species to Al₂O₃. The degrees of Co reduction estimated by H₂-TPR experiments are also shown in Table 1. Co/SiC catalyst possesses the highest Co reduction degree, but it decreases to 56% for Co/Al₂O₃@SiC catalyst. This suggests that the Co₃O₄ is difficult to reduce after altering the surface Si_xO_y species on SiC. The cobalt reduction degree for Co/Al₂O₃@SiC and Co/Al is at the same level, which could be attributed to the similar structure of the two supports, that is, a thin layer of Al₂O₃ is existed on the surface of the two materials. Fig. 2B shows the XPS results of the catalysts. Two peaks centered at about 780 and 795 eV are observed, which can be ascribed to the Co 2p_3/2 and Co 2p_1/2 peaks of Co₃O₄.¹² For the Co/Al₂O₃@SiC, the peak of Co 2p_3/2 shift to a higher binding energy than Co/SiC, indicating that the interaction between Co₃O₄ and support becomes stronger after changing the Si_xO_y to Al₂O₃ on the surface of SiC.

Fig. 2 (A) H₂-TPR profiles of the catalysts: (a) Co/SiC, (b) Co/Al₂O₃@SiC, (c) Co/Al, (d) Co/Al₂O₃; (B) Co 2p XPS spectra of the catalysts: (a) Co/SiC, (b) Co/Al₂O₃@SiC, (c) Co/Al, (d) Co/Al₂O₃.

The catalytic results of the catalysts are listed in Table 2. Reaction data were collected after 48 h time-on stream in order to obtain the steady state. The Co/Al₂O₃@SiC and Co/Al₂O₃ are compared to investigate the effect of support thermal conductivity on the intrinsic reaction behavior of FTS. The Al₂O₃ and Al₂O₃@SiC supports are all prepared by precipitation method, thus the two supports could possess similar surface chemical nature. In addition, 15 wt% cobalt is impregnated into Al₂O₃ in order to rule out cobalt particle sizes effect. Hence the major variable between them is the thermal conductivity property, as Al₂O₃ is insulator but Al₂O₃@SiC is high thermal conductive material. As can be seen, Co/Al₂O₃@SiC and Co/Al₂O₃ exhibit similar catalytic performance, namely the CH₄ selectivity is 11.8% and 12.3% while C₅₊ selectivity is 78.3% and 77.6%. This suggests that the thermal conductivity of catalyst support makes no effect on the intrinsic reaction behavior of FTS. Actually, inert solid heat disperser was often used in FTS and the temperature gradient in the catalyst bed could be negligible, especially for our laboratory fixed-bed reactor.^12,22 At the reaction steady state, the generated and transferred heat could also be in equilibrium on each catalyst particle, regardless of the thermal conductive property for catalysts. Therefore, the traditional speculation that “hot spots” exist on the surface of insulated supports but do not form on high thermal conductive SiC surface thus can enhance heavy hydrocarbon selectivity may be inaccurate.

Table 2 FTS performance of the catalysts^a

Catalyst	XCO (%)	Products selectivity (wt%)				CTY^b	TOF (10^−2 s^−1)
		CH4	C2–C4	C5+	CO2
a Reaction conditions: n(H2)/n(CO) = 2, GHSV = 1.0 Lsyngas gcat h^−1, T = 503 K, P = 2.0 MPa, TOS = 48 h.b Cobalt time yield (10^−5 molCO gCo^−1 s^−1, molar CO conversion rate per gram of cobalt per hour).
10Co/SiC	39.3	6.5	4.9	88.4	0.2	1.6	2.7
10Co/Al	45.9	11.2	9.1	78.1	1.6	1.9	2.9
10Co/Al2O3@SiC	53.8	11.8	9.0	78.3	0.9	2.2	2.6
15Co/Al2O3	49.0	12.3	7.9	77.6	2.2	1.4	2.9

---

The dilution experiment results of low thermal conductive Co/Al₂O₃ and high thermal conductive Co/SiC are shown in Table S4.† As can be seen, the temperature difference between the reactor wall and the centerline (ΔT) for the undiluted Co/Al₂O₃ is 1.3 K higher than the diluted Co/Al₂O₃. This is because that the heat transfer condition is bad for the undiluted Co/Al₂O₃ and the generated heat is accumulated in the catalyst bed, leading to a higher temperature gradient between reactor wall and the centerline.²³ As a consequence, the undiluted Co/Al₂O₃ displays higher CO conversion than the diluted Co/Al₂O₃, i.e. 78.4% instead of 59.3%. The CO conversion of 59.1% is chosen for diluted Co/SiC, in order to make sure that the generated heat during reaction is similar with diluted Co/Al₂O₃. No obvious temperature difference (ΔT) is observed for the diluted and undiluted Co/SiC. This is because that Co/SiC catalyst possesses well heat transfer ability thus a homogenous temperature gradient inside the reactor could be maintained even without using heat disperser. Therefore, the real effect of the high thermal conductive SiC support is to improve heat transfer efficiency inside the reactor, thus maintaining a homogeneous temperature gradient within catalyst bed, which is beneficial to the safety of fixed-bed reactor particularly for large scale commercial plants.

The only difference between SiC and Al₂O₃@SiC is the species on the surface. Hence Co/SiC and Co/Al₂O₃@SiC are compared to investigate the effect of SiC surface Si_xO_y species on the intrinsic reaction behavior of FTS. As can be seen, Co/SiC shows remarkable catalytic performance, that is, a lower CH₄ selectivity of 6.5% and higher C₅₊ selectivity of 88.4%. However, after the Si_xO_y layer on SiC surface is changed to Al₂O₃, obvious different reaction result is observed for the prepared Co/Al₂O₃@SiC. The CH₄ selectivity increases to 11.8% and C₅₊ selectivity decreases to 78.3%. It is accepted that the metallic Co⁰ is the active phase in FTS.^22,24 For Co/Al₂O₃@SiC, the interaction between Co₃O₄ and Al₂O₃ layer is stronger thus lead to a lower cobalt reduction degree, which enhances CH₄ selectivity.^25,26 After reduction, the stronger interaction could also change the electronic state of the Co⁰ and affect the CO dissociation ability, thus influencing the products selectivity.^12,27 For Co/SiC, the interaction between Co₃O₄ and surface Si_xO_y layer is relative weaker thus could own appropriate cobalt reduction degree. Furthermore, the electronic state of the Co⁰ might not be significantly affected because of the relative weaker interaction. Therefore, we believe that the surface Si_xO_y layer on SiC is the main cause for the intrinsic good catalytic performance of Co/SiC. To further confirm our conclusion, Co/Al catalyst is also tested under the same reaction condition. The structure of Al powders is similar with SiC and Al₂O₃@SiC, that is, a thin layer of Al₂O₃ is formed on the outer surface and underneath is high thermal conductive metallic Al. As can be seen, the CH₄ selectivity is 11.2% and C₅₊ selectivity is 78.1% for Co/Al, which is similar with Co/Al₂O₃@SiC but different with Co/SiC. This could be attributed to the different surface species on these supports. For Al₂O₃@SiC and Al powders, the surface species is Al₂O₃ but for SiC is Si_xO_y. The result confirms our conclusion that the surface Si_xO_y layer on SiC is the main cause for the intrinsic good catalytic performance of Co/SiC.

In summary, we have demonstrated that the traditional explanations for the intrinsic excellent catalytic performance of Co/SiC are not reliable. The high thermal conductive property is not the cause for the intrinsic good FTS performance of Co/SiC. The real effect of the high thermal conductivity is to increase heat transfer efficiency and maintain a homogeneous temperature gradient inside the fixed-bed reactor, thus avoiding the formation of “hot spots”, which could improve the security of plants. The surface Si_xO_y layer on SiC is the main cause for the intrinsic good catalytic performance of Co/SiC, which possesses proper interaction between cobalt phase and Si_xO_y. Therefore, the attempt to utilize other high thermal conductive materials as supports to prepare FTS catalysts with good catalytic performance similar with Co/SiC might not be working. Regulating the interaction between cobalt phase and support might be an effective approach to prepare FTS catalysts with excellent catalytic performance.

Acknowledgements

We gratefully acknowledge the financial support from the National Natural Science Foundation of China (No. 21273265 and 21203232), the Strategic Priority Research Program Demonstration of Key Technologies for Clean and Efficient Utilization of Low-rank Coal (No. XDA07070700).

Notes and references

1. E. Iglesia, Appl. Catal., A, 1997, 161, 59.
2. J. Wang, H. Li, D. Li, J. P. den Breejenc and B. Hou, RSC Adv., 2015, 5, 65358.
3. Q. Yan, C. Wan, J. Liu, J. Gao, F. Yu, J. Zhang and Z. Cai, Green Chem., 2013, 15, 1631.
4. X. Zhu, X. Lu, X. Liu, D. Hildebrandt and D. Glasser, Ind. Eng. Chem. Res., 2010, 49, 10682.
5. X. Zhu, X. Lu, X. Liu, D. Hildebrandt and D. Glasser, Chem. Eng. J., 2014, 247, 75.
6. Y. Liu, B. de Tymowski, F. Vigneron, I. Florea, O. Ersen, C. Meny, P. Nguyen, C. Pham, F. Luck and C. Pham-Huu, ACS Catal., 2013, 3, 393.
7. M. Lacroix, L. Dreibine, B. de Tymowski, F. Vigneron, D. Edouard, D. Begin, P. Nguyen, C. Pham, S. Savin-Pocent, F. Luck, M. J. Ledoux and C. Pham-Huu, Appl. Catal., A, 2011, 397, 62.
8. Y. Liu, O. Ersen, C. Meny, F. Luck and C. Pham-Huu, ChemSusChem, 2014, 7, 1.
9. I. G. Solomonik, K. O. Gryaznov, V. F. Skoka and V. Z. Mordkovichab, RSC Adv., 2015, 5, 78586.
10. Y. Liu, I. Florea, O. Ersen, C. Pham-Huu and C. Meny, Chem. Commun., 2015, 51, 145.
11. H. Koo, B. Lee, M. Park, D. J. Moon, H. Roh and J. W. Bae, Catal. Sci. Technol., 2014, 4, 343.
12. A. Y. Khodakov, W. Chu and P. Fongarland, Chem. Rev., 2007, 107, 1692.
13. Y. Wang, Y. Xu, H. Xiang, Y. Li and B. Zhang, Ind. Eng. Chem. Res., 2001, 40, 4324.
14. R. Pampuch, W. Ptak, S. Jonas and J. Stoch, Mater. Sci. Monogr., 1980, 6, 435.
15. R. Moene, M. Makee and J. A. Moulijn, Appl. Catal., A, 1998, 167, 321.
16. G. D. Wagner, W. M. Riggs, L. E. Davis, J. F. Moudler and G. E. Muilenberg, Handbook of X-ray photoelectron spectroscopy, Perkin-Elmer Corporation, 1979.
17. DIFFRACplus EVA Release 2001, Version 7.0 rev.0, Bruker AXS, 2001.
18. D. Schanke, S. Vada, E. A. Blekkan, A. M. Hilmen, A. Hoff and A. Holmen, J. Catal., 1995, 156, 85.
19. G. L. Bezemer, J. H. Bitter, H. P. C. E. Kuipers, H. Oosterbeek, J. E. Holewijn, X. Xu, F. Kaptejin, A. J. van Dillen and K. P. de Jong, J. Am. Chem. Soc., 2006, 128, 3956.
20. J. Cho, C. Ahn, C. Pang and J. Bae, Catal. Sci. Technol., 2015, 5, 3525.
21. R. Oukaci, A. H. Singleton and J. G. Goodwin, Appl. Catal., A, 1999, 186, 129.
22. Q. Zhang, J. Kang and Y. Wang, ChemCatChem, 2010, 2, 1030.
23. M. Sheng, H. Y. Yang, D. R. Cahela and B. J. Tatarchuk, J. Catal., 2011, 281, 254.
24. Z. Yan, Z. Wang, D. B. Bukur and D. W. Goodman, J. Catal., 2009, 268, 196.
25. A. A. Khassin, T. M. Yurieva, V. V. Kaichev, V. I. Bukhtiyarov, A. A. Budneva, E. A. Paukshtis and V. N. Parmon, J. Mol. Catal. A: Chem., 2001, 175, 189.
26. S. W. Ho, M. Horialla and D. M. Hercules, J. Phys. Chem., 1990, 94, 6396.
27. G. Prieto, I. S. Deello, P. Concepció, R. Urciano, S. B. C. Pergher and A. Martínez, ACS Catal., 2015, 5, 3323.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra22170f

---