Due to the strong surge of natural gas production, the feedstock for the chemical industry for the near future is expected to shift towards lighter hydrocarbons, in particular methane. The success of a Gas-to- Chemicals process via synthesis gas, a mixture of CO and H2, depends on the ability of catalysts that suppress the formation of methane and of carbon dioxide. We designed a Co1Mn3-Na2S catalyst, which is shown to have negligible Water-Gas-Shift activity and a hydrocarbon product spectrum, which deviates from the Anderson–Schulz–Flory distribution. At 240 °C and 1 bar, this catalyst showed a C2-C4 olefins selectivity of 54 %C with a C2-C4 olefin/paraffin ratio of 17. At more industrially relevant conditions of 240 °C, 10 bar, H2/CO = 2, 18 – 30 % CO conversion, the catalytic performance of Co1Mn3-Na2S was compared to other Co-based catalysts (Figure1). At 240 °C and 10 bar, this catalyst displayed still a selectivity towards chemicals of 30 %C and to fuels of 59 %C; the rest being methane and C2-C4 paraffins.
To obtain further mechanistic insights into the various catalytic systems, the detailed C product flow of 1-olefin and n-paraffin for each C number product was studied and Co1Mn3-Na2S produced significantly more primary olefins than linear paraffins for each C containing hydrocarbon product. This suggests that β-H elimination is the dominant termination pathway for Co1Mn3-Na2S and secondary hydrogenation of olefins was suppressed. Besides, the low fraction of 2-butenes in the C4 hydrocarbon product spectrum of Co1Mn3-Na2S implied the suppression of secondary isomerization of olefins. Tentatively the addition of Na2S to Co1Mn3 was proposed to deactivate sites for secondary olefin hydrogenation and isomerization and for methanation, whereas the low “degree of alkalization” as compared with Na2O is sufficiently low to inhibit the WGS reaction.
Detailed characterization of the spent catalyst using XRD and TEM revealed ~10 nm crystallites with hcp Co metal phase. Figure 2 shows the TEM images and particle size distribution of spent Co1Mn3-Na2S after industrially relevant conditions (240 - 280 °C, 10 bar, and H2/CO = 2), and STEM-EDX maps to differentiate Co and Mn. The Co particle size distribution from TEM revealed the average Co particle size to be 9.6 nm with a standard deviation of 4.4 nm, in agreement with the Co crystallite size of 9.2 ± 1.9 nm from XRD. The elemental maps of Co and Mn in Figure 2f confirmed that spent Co1Mn3-Na2S consisted of Co nanoparticles well dispersed on MnO support. In summary, we propose a synergistic effect of Na plus S, which act as electronic promoters on the metallic Co surface, thus improving the selectivities towards lower olefins and fuels while largely reducing formation of CO2 and CH4.
Figure 1. Product selectivities of Co catalysts at 240 °C, 10 bar, H2/CO = 2, 18 – 30 % CO conversion and 100 hours time-on-stream.
Figure 2. (a – b) TEM images, (c) Co particle size distribution and (d - f) STEM-EDX mapping of spent Co1Mn3-Na2S after FT synthesis at 240 - 280 °C, 10 bar, H2/CO = 2 and 100 h time-on-stream.