Design of BC 3 N 2 -Based Single Atom Catalyst for CO 2 Electroreduction: Assisted by a Multiobjective Optimization Algorithm

The massive emission of CO2 has caused serious environmental and energy problems. Electroreduction of CO2 is an effective method to solve this issue. However, the process of screening favorable catalysts has dramatically hindered catalyst design. We have therefore attempted to combine the multiobjective optimization algorithm with catalyst design to search for the optimal catalyst, TM@BC3N2 (BC3N2 substrate with transition metals loaded on it), for CRR. We have used properties, including binding energies, adsorption energies, and C–O bond lengths, as three objectives and eventually determined Ti@BC3N2 and V@BC3N2 as promising candidates due to their excellent capture and activation capabilities. The catalyst Ti@BC3N2 can produce HCOOH and CO as CRR products, with limiting potentials of −0.42 V and −0.52 V, respectively. For V@BC3N2, HCOOH and CO should be produced at limiting potentials of −0.65 V and −0.6 V. The reactivity of *CO2 protonation to *COOH is significantly higher on Ti@BC3N2 than that on V@BC3N2, with reaction energies of 0.29 and 0.6 eV, respectively. Because Ti@BC3N2 changes from a magnetic state (pure Ti@BC3N2) to a nonmagnetic state after CO2 adsorption, the decrease of parallel spins destabilizes *CO2 relative to *COOH, eventually decreasing the reaction energy of *CO2 → *COOH step. Therefore, Ti@BC3N2 is ultimately selected as the optimal catalyst among the TM@BC3N2 candidates investigated. This work can be treated as a starting point for combining the multiobjective optimization algorithm with catalyst screening. It has provided a novel design idea for the rapid screening of the optimal catalyst.