[18a-A21-1~11] 23.1 Joint Session N "Informatics"

11:30 AM - 11:45 AM JST(2:30 AM - 2:45 AM UTC)

[18a-A21-10] Crystal Structure Prediction Using Universal Neural Network Potential PFP

〇Takuya Shibayama¹, Kohei Shinohara¹, Hideaki Imamura¹, Katsuhiko Nishimra¹, So Takamoto¹, Chikashi Shinagawa¹ (1.PFN)

Keywords:

Machine Learning,CSP,Material Discovery

Our research group has developed a universal interatomic potential (PFP). In this study, we leveraged the PFP to develop a crystal structure prediction (CSP) system that incorporates novel algorithms and efficient computational environments. Employing this system, we conducted searches on several elemental systems comprising 2 to 10 different elements, leading to the discovery of multiple novel stable crystal candidates. It is hypothesized that CSP utilizing universal interatomic potentials can accelerate the discovery and design of new materials.