Revealing the intrinsic structure-property relationship in hard carbon during (de)sodiation: Insights from direct atomic simulations with neural network potentials

JH Chen and XX Chen and B Xu and LB Chen and CY Wang and CL Wen and XS Zhao and MS Wu and B Sa and CY Ouyang, ENERGY STORAGE MATERIALS, 81, 104546 (2025).

DOI: 10.1016/j.ensm.2025.104546

Sodium ion batteries (SIBs), an immensely attractive substitute to lithium-ion batteries, have received significant attention in applications of large-scale energy storage and transportation owing to the abundant reserves of sodium and highly sufficient similarity to lithium. Among carbon-based materials, hard carbon (HC) is expected to be a promising host for sodium ions, offering high thermodynamic stability when charge transfer. However, the limited understanding of HC's structural characteristics and sodium binding mechanism presents a significant challenge to the rational design of high-performance HC. Here, by employing an independently developed Neural Network Potential (NNP) model for NaC binary system with certain experimental techniques, fundamental HC structure and the correspondence between each segment of voltage curve and sodium storage sites are reproduced. Additionally, intrinsic atomic scale structure-property relationship underlying high capacity in HC is uncovered. It is highlighting that the competition between the Na-Na/Na-C interaction ratio is identified as a critical factor determining the attribute of voltage curve, leading to the discovery of a novel "adsorptioninsertion/filling" hybrid (de)sodiation mechanism. These findings advance methodologies for analyzing amorphous materials and providing atomic-level insights to guide more precise structural optimization for HC, facilitating the development of superior electrode materials for next-generation sodium-ion batteries.

Return to Publications page