Rational Design of Graphene-Supported Single Atom Catalysts for High Performance Lithium-Oxygen Batteries

Aprotic lithium-oxygen batteries (LOBs) have been considered as one of the next-generation battery technologies, due to its ultrahigh theoretical energy density (~3600 Wh kg -1 ), 5-10 times higher than the state-of-the-art lithium-ion batteries (LIBs), supporting the development of electric vehicles (EVs). However, the battery system suffers from poor cyclability, low round-trip efficiency and inferior rate capability, mainly originated from the inertness of oxygen gas and the poor electrical conductivity of lithium peroxide (Li 2 O 2 ), leading to the sluggish kinetics of oxygen reduction reaction (ORR)/ oxygen evolution reaction (OER). In response to the challenges, developing a highly efficient catalyst at the cathode is vital to boost the reaction rate and reduce the reaction overpotentials, eventually address the problem. Heterogeneous single-atom catalysts (SACs) on solid support are emerging as a new frontier in this research field. Here, we use the density of state (DFT) calculation to determine the catalytic activities of a wide range of transition metal single atoms distributing on the nitrogen doped graphene support and found that Zn-SAC exhibits the highest ORR/OER activities. We discovered that Zn-N 4 moieties, functioning as catalytic centers, bind with LiO 2 not very strongly, reduce the reaction overpotentials, facilitates the reaction rate and enhance the stability of the catalyst. The catalytic activity of SACs is highly correlated to the Gibbs free energy of the adsorbed LiO 2 . A descriptor, F of the catalysts determining the catalytic activity was developed from the metal properties of electronegativity (EN), enthalpy of vaporization (EV) and number of electrons in d orbital, by using supervised machine learning (ML) method, providing guidance for designing useful SACs in ORR/OER process. This work systematically provides guidance to design highly efficient SAC for LOBs and provides fundamental insights in choosing the proper metal for the ORR/ OER application.