Recently, Professor Wang Bin and Distinguished Researcher Liu Xinyan, both from the Institute of Fundamental and Frontier Sciences at the University of Electronic Science and Technology of China, in collaboration with Professor Cheng Jianli from the School of Optoelectronic Science and Engineering, introduced a machine learning-aided screening approach. This method is designed to swiftly pinpoint effective catalysts for Li-CO₂ and Li-Air batteries among transition metal compounds (TMCs). Their research was published in the renowned Journal of the American Chemical Society (JACS).

Employing an iterative machine learning process, the research team sifted through 15,012 transition metal compounds and successfully synthesized three representative catalysts. Experimental tests confirmed that the predictive model boasted an average absolute error of a mere 0.106V. Among the synthesized catalysts, Co₀.₁Mo₀.₉N stood out, demonstrating remarkably low overpotentials of 0.55V in Li-CO₂ batteries and 0.65V in Li-Air batteries, all under a current density of 50mA g⁻¹.

A deeper mechanistic analysis uncovered that Co doping played a pivotal role in modifying the electronic structure of MoN. This modification facilitated enhanced electron transfer and significantly boosted the catalytic activity. This groundbreaking study not only offers a fresh perspective on expediting the screening and design of innovative battery catalysts but also sets the stage for the development of efficient and eco-friendly electrochemical energy systems.