Effects of Promoter Deposition Order and Solvent On Preparation of Cobalt Fischer-Tropsch Catalysts

Previous studies have shown that preparation method and pretreatment conditions affect dispersion, extent of reduction, location, and density of cobalt and promoter species in Fisher-Tropsch (FT) catalysts. While the preparation method should lead to highly active and selective cobalt FT catalysts, it also needs to satisfy the technical and economic feasibility of scale-up. Some non-aqueous preparation methods have been found to produce active and selective cobalt catalysts. However, highly volatile and flammable solvents are not suitable for producing industrial scale quantities of catalyst. Accordingly, aqueous methods that result in catalysts of comparable activity may be preferred. In addition, the high cost of noble metal promoters requires they be used efficiently. In this work, we investigate how the solvent nature and the ruthenium deposition order affect physicochemical and activity/selectivity properties of FT catalysts. This study is expected to provide insight into the influence of solvent nature and promoter deposition order on cobalt catalysts and develop an industrial procedure to prepare more economically viable catalysts for the FT synthesis.

Catalysts with dispersions of 16% at loadings of approximately 25 wt% cobalt and ~0.3 wt% ruthenium on lanthanum-treated alumina were prepared by evaporative deposition using water or ethanol/acetone. The deposition of cobalt included three steps. Ruthenium was deposited either in the third cobalt deposition (co-deposition) or in a fourth deposition step (consecutive deposition). After each deposition, dried catalyst particles were pelletized, crushed, sieved to -28 +65 mesh, and then calcined. Characterization techniques, including Hydrogen Chemisorption, BET, Hydrogen-TPR, TEM, SEM and activity/selectivity tests in a fixed-bed reactor were used to study these catalysts.

Analyses of the surface area, pore volume, and pore diameter suggest that the deposition order of promoter has little effect on the physical properties of the Co/Ru/La-Al2O3 catalysts. On the other hand, the lower surface area of the aqueous-deposited (120 m2/g) catalyst relative to its non-aqueous-deposited counterparts (130-135 m2/g) is most probably due to the high surface tension of water that might have caused the collapse of pores.

Our studies show that the deposition order affects the catalyst physicochemical properties. While each preparation technique uses the same amount in weight of cobalt and ruthenium, the resulting chemical compositions are not the same. The non-aqueous-consecutive-deposited shows significantly lower ruthenium content (0.13 wt%) than both the non-aqueous-co- (0.22 wt%) and aqueous-co-deposited (0.23 wt%) catalysts. In addition, the non-aqueous- and aqueous-consecutive-deposited catalysts present lower hydrogen uptakes and extents of reduction probably due to differences in hydrogen spillover. The solvent nature also affects catalyst reducibility. The aqueous preparation results in a catalyst with lower optimal reduction temperature, higher hydrogen uptake and extent of reduction than the non-aqueous preparations do.

By comparing the catalysts at 230 degree C, 300 psi and CO/hydrogen ratio of 2, the CO conversion and turnover frequencies for the non-aqueous-co-deposited catalyst (24%, 0.032 s-1) is higher than that for the aqueous-consecutive-deposited (22%, 0.025 s-1) and non-aqueous-consecutive-deposited (16%, 0.023 s-1) one. Methane and carbon dioxide selectivities are similar for all catalysts. Our results indicate that the aqueous-based method is an effective alternative to produce an active FT catalyst.