Machine learning driven rational design of dual atom catalysts on graphene for carbon dioxide electroreduction
Dongxu Jiao, Xinyi Li, Mingzi Sun, Lin Liu, Jinchang Fan, Jingxiang Zhao, Bolong Huang, Xiaoqiang Cui
Abstract
The development of high-performance atomic catalysts for the carbon dioxide reduction reaction (CO<sub>2</sub>RR) is a time-consuming process due to the complexity of the reaction mechanism and the uncertainty of the active site. Herein, we have proposed combining density functional theory (DFT) and machine learning (ML) to investigate the potential of topological graphene-based dual-atom catalysts (DACs) as CO<sub>2</sub>RR electrocatalysts. By analyzing the ML results, we identify the number of d-orbital electrons in the active site as a key factor influencing the CO<sub>2</sub>RR catalytic activity. Additionally, we propose a simple descriptor to measure the CO<sub>2</sub>RR activity of these DACs. Our findings provide plausible explanations for the synergistic interactions between bimetallic atoms in CO<sub>2</sub>RR and allow us to screen the homogeneous Ni-Ni pair as the most promising dual-atom catalysts. This work offers a fast ML approach based on limited DFT calculations to predict the most electroactive and stable DACs on carbon support for CO<sub>2</sub>RR, facilitating rapid screening of high-performance dual-atom catalysts.