[17a-PB02-1~9] 23.1 Joint Session N "Informatics"

[17a-PB02-3] Comprehensive search for functional contributing factors of Co₂MnGa_1-xGe_x using multimodal and multiscale analysis

〇(B)Naoaki Saruyama¹, Alexandre Lira Foggiatto¹, Takahiro Yamazaki¹, Tomohiro Matsushita², Yusuke Hashimoto², Sota Kawamura², Hiroto Tomita², Zexu Sun², Takuya Moriki², Varun K. Kushwaha³, Yuya Sakuraba³, Yuma Iwasaki³, Masato Kotsugi¹ (1.Tokyo Univ. of Sci., 2.NAIST, 3.NIMS)

Keywords:machine learning, materials informatics, Heusler alloys

Heusler alloys are attracting attention as one of the effective magnetic materials for next-generation spintronics devices. However, the elements that make up Heusler alloys have a huge number of combinations, and the relationship between the structure and electronic state behind the spin polarization, which is important for applications, is complicated, making it extremely difficult to understand the mechanism of functional expression. Therefore, we analyzed the structure and electronic state by multi-modal and multi-scale, and tried to extract the contributing factors of function by exhaustive machine learning.