Machine Learning Interatomic Potential Calculations for Designing Layered Oxides in Sodium-ion Battery Systems with Low Environmental Impacts

Layered P2-type structures of MnNi oxides are attracting attention as cathode materials for sodium-ion batteries. However, life-cycle assessment analysis has shown that Ni usage has substantial environmental impact, making its substitution necessary. This paper presents a benchmarking study on machine learning interatomic potentials for P2-type layered oxides. Using the selected potential, we comprehensively performed element substitution and evaluated candidates for Ni replacement. Experimental results revealed that substitution with Fe and the introduction of Ni vacancies can effectively reduce environmental impact.