期刊
ACTA MATERIALIA
卷 261, 期 -, 页码 -出版社
PERGAMON-ELSEVIER SCIENCE LTD
DOI: 10.1016/j.actamat.2023.119348
关键词
Permanent magnets; Rare-earth-free; High-throughput; DFT; Neural networks; Structure prediction
With advancements in crystal structure prediction, databases of new materials are being developed. Through these databases, stable and metastable materials can be discovered and studied for their properties. Using a high-throughput and data-mining approach, four promising candidates for rare-earth-free permanent magnets have been found.
With the recent developments in crystal structure prediction, databases of new (not previously synthesized) materials are being created. One of these databases contains more than a million entries with the distance to the Convex Hull predicted by crystal-graph attention networks. Hence, stable and metastable materials can be extracted and then investigated for any desired properties. A high-throughput and data-mining approach we previously developed to search for rare-earth-free permanent magnets was applied to these compounds. As a result, four promising candidates for novel rare-earth-free permanent magnets were discovered with high magnetization, high uniaxial magnetocrystalline anisotropy, and high Curie temperature - Ta3ZnFe8, AlFe2, Co3Ni2, and Fe3Ge. The materials were investigated in more detail and all were verified to be dynamically stable.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据