Insight into Catalytic Active Sites on TiO2/RuO2 and SnO2/RuO2 Alloys for Electrochemical CO2 Reduction to CO and Formic Acid

Density functional theory calculations are used to analyze and determine the active sites for CO2 reduction reaction (CO2RR) toward CO and formic acid on TiO2/RuO2 and SnO2/RuO2 alloys in their rutile structure with the (110) facet. Ti and Sn atoms in TiO2 and SnO2 catalysts are substituted with Ru atoms with different ratios and compositions in order to determine recently observed experimental trends and gain insights into catalytic active sites. We base our analysis on constructing volcano plots in order to predict the overpotential needed for CO2RR on all the model systems. We observe that catalyst compositions having alternating bridge Ru-Ti as binding sites for the key intermediates of COOH or OCHO result in higher overpotentials than the reference RuO2 surface where only H2 is formed experimentally. If the binding sites are either bridge Ru-Ru or especially bridge Ti-Ti, it significantly lowers the overpotentials for CO formation, which indicates that these are the active sites of the TiO2/RuO2 alloys. For formic acid formation, the bridge Ru-Ru sites result in the lowest overpotentials, whereas the bridge Ti-Ti sites bind the OCHO intermediate too strongly and give rise to large overpotentials. Furthermore, the calculations show clearly that when replacing Cu for one bridge Ru atom in a RuO2 overlayer on TiO2, the overpotential decreases significantly toward formic acid and especially CO formation in agreement with experimental observations. Finally, for the SnO2/RuO2 alloys, replacing Sn with Ru in the coordinatively unsaturated sites decreases the overpotential compared with all other model systems of the SnO2/RuO2 alloys, which is due to electronic effects since the key intermediates are catalyzed on the neighboring bridge sites. The knowledge gained from these synergistic effects when manufacturing these alloys may be used to engineer the active sites for in order to the and decrease the required overpotential.