A Deep Neural Network Interface Potential for Li‐Cu Systems

Abstract

AbstractCopper foil is one of the most commonly used current collector materials in Li metal batteries. However, many problems on the Li–Cu interface have not been effectively solved due to the lack of a fundamental understanding of Li–Cu interaction at the atomic scale. In this work, a deep neural network interface potential for Li‐Cu systems using neural networks combined with active learning strategies is developed. The potential shows excellent performances on the energy and force calculations, physical properties predictions, and structure explorations. Moreover, the study of the Li adsorption behaviors on the Cu surface demonstrates the accuracy of this potential in the investigation of the Li–Cu interface. This potential for the Li‐Cu systems provides an important opportunity to advance the understanding of interface problems in Li metal batteries.