Understanding of the Fate of α-Fe2O3 in CO2 Hydrogenation through Combined Time-Resolved In Situ Characterization and Microkinetic Analysis

CO2 hydrogenation over Fe-based catalysts is a promising pathway to mitigate emissions of this greenhouse gas and provides a possibility for crude-oil-free production of chemicals and fuels. Understanding of (i) the role of crystalline phases or/and surface species in the working catalysts, (ii) the factors affecting their formation under reaction conditions, and (iii) the kind and reactivity of surface precursors of gas-phase products is vital for controlling the efficiency of CO2 hydrogenation. In this study, we applied time-resolved in situ characterization techniques for monitoring phase transformations during Fe2O3 reduction, starting of CO2 hydrogenation, steady-state operation, and finally catalyst deactivation. The obtained structural information after different times on stream was related to kinetic data obtained from temporal analysis of H2 and CO2 activation as well as from steady-state isotopic transient kinetic analysis (SSITKA). Fe2O3 is easily reduced to Fe3O4 and Fe in H2 above 300 °C. Fe5C2 and Fe3C, which are quickly formed from metallic Fe/Fe3O4 under CO2 hydrogenation conditions, do not undergo oxidation with rising time on the reaction stream under ambient-pressure conditions. Nevertheless, the catalyst loses its initial activity and, particularly, the selectivity to hydrocarbons in favor of CO. Thus, we do not confirm the well-recognized deactivation mechanism of CO/CO2 hydrogenation through oxidation of iron carbides. Instead, surface carbonaceous species identified by in situ Raman and pseudo in situ XPS measurements were concluded to cause catalyst deactivation and deselectivation due to hindering the catalyst ability to generate surface species from H2 and CO2. Specifically, the strength of CO2 adsorption and the catalyst activity to dissociate adsorbed CO2 to adsorbed CO decrease in the presence of carbon deposits. Kinetic evaluation of SSITKA tests revealed the presence of (i) only one kind of surface intermediate yielding gas-phase CO after 1.5 h on reaction stream but (ii) at least two kinds (short-lived and long-lived) of surface intermediates participating in CH4 formation in parallel. Carbon deposits seem to block the sites responsible for the formation of short-lived intermediates.