[17a-S2_204-1~13] 23.1 Joint Session N "Informatics"

10:30 AM - 10:45 AM JST(1:30 AM - 1:45 AM UTC)Presentation by Applicant for JSAP Young Scientists Presentation Award

[17a-S2_204-7] Wide-Range Exploration of Ammonia Synthesis Catalysts Based on Open-Data Machine Learning Models

〇(M2)Takuya Horita¹, Ryoji Asahi¹ (1.Nagoya Univ.)

Keywords:

catalysts for ammonia synthesis,machine learning,genetic algorithm

This study aims to propose an algorithm for exploring novel catalysts that enable ammonia synthesis under mild conditions. Using open databases and CrabNet models trained on these data, we analyzed the elemental contributions to adsorption energies. High-activity regions were identified based on a two-dimensional volcano plot employing E(N*) and E(NH*) as descriptors. Furthermore, compositional optimization was performed using a genetic algorithm, demonstrating the effectiveness of the proposed efficient catalyst exploration strategy.