Accessing the Nature of Active Sites and Particle Size Effect for Reduction of Carbon Dioxide over Copper-Based Catalysts

Although electrocatalytic conversion of CO2 (CO2ER) over bimetallic Cu-based catalysts has been regarded as a promising and compelling route for the sustainable synthesis of fuels and feedstock when combined by carbon-free electricity, questions still remain concerning the fundamental understanding of the reaction mechanism and the nature of active sites, hampering the rational design of catalyst with great activity and selectivity a priori. We report a global optimization in large scale to obtain serious realistic nanoparticles (NPs) models of different particle sizes and the identification of atom-level structures of active sites for CO products on CuZn and CuAu NPs catalysts during CO2ER, using machine learning and density functional theory calculations. After the analysis of 300 surface sites (600 computational data points) through neural network (NN) potential based high-throughput testing, we demonstrate that the bimetallic Cu-based NPs have superior CO2ER because there are many bimetallic synergistic effect sites that significantly stabilize the carboxyl intermediate during CO2 reduction to CO, breaking the inherent linear relationship. This work shed light on the structure- performance relationship over more realistic large NPs, facilitating the rational design of Cu-based catalysts in CO2ER.