[15p-P06-1~15] 23 Joint Session N "Informatics" (Poster)

[15p-P06-4] Analysis of Proton Transport in Polymer Electrolyte Membranes Using Machine-Learning Molecular Dynamics Simulations

〇Yuta Yoshimoto¹, Naoki Matsumura¹, Yuto Iwasaki¹, Hiroshi Nakao¹, Yasufumi Sakai¹ (1.Fujitsu)

Keywords:

molecular dynamics,neural network potential,polymer electrolyte membrane

Molecular dynamics (MD) simulations employing machine learning interatomic potentials have attracted significant attention as a method capable of performing dynamical simulations of larger systems while maintaining accuracy comparable to density functional theory calculations. In this study, we construct a neural network potential for Nafion, a polymer electrolyte membrane. We perform large-scale and long-time MD simulations of hydrated Nafion membranes and successfully reproduce self-diffusion coefficients of protons with higher accuracy than previous studies.