Revisiting the unified principle for single-atom electrocatalysts in the sulfur reduction reaction: from liquid to solid-state electrolytes

Abstract

The conversion of lithium-polysulfides (LPSs) through the sulfur reduction reaction (SRR) is a crucial process for improving the electrochemical performance of lithium–sulfur (Li–S) batteries. However, the microscopic mechanism of the SRR remains unclear, affecting catalyst design for Li–S batteries. By applying artificial intelligence (AI), we have developed a unified mechanistic model for the SRR on metal–nitrogen-doped carbon (TMNC, TM = 3d/4d/5d transition metals) catalysts. This model reveals the SRR catalytic activity's physical essence in TMNCs, rooted in wavefunction overlap between transition metals and non-metal atoms. This is supported by physical models and experiments. Using this insight, we have anchored FeNCs (and Fe₃C for comparison) onto carbon fibers for the sulfur cathode/lithium anode, enhancing lithium metal's cyclic life to over 10 000 hours. The solid-state Li–S full cell demonstrates an energy density of ∼400 W h kg⁻¹ with consistent cyclic performance. Our AI-enhanced mechanistic understanding of the SRR guides the development of superior SRR catalysts and high-performance Li–S batteries.