Magnetic materials exploration using data-driven AI that can explain structure-property mechanisms

• Project/Area Number: 20K05301
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 29010:Applied physical properties-related
• Research Institution: Japan Advanced Institute of Science and Technology
• Principal Investigator: DAM Hieu Chi 北陸先端科学技術大学院大学, 先端科学技術研究科, 教授 (70397230)
• Co-Investigator(Kenkyū-buntansha): 水上 卓 北陸先端科学技術大学院大学, 先端科学技術研究科, 助教 (50270955)
• Project Period (FY): 2020-04-01 – 2023-03-31
• Project Status: Completed (Fiscal Year 2022)
• Budget Amount: ¥4,290,000 (Direct Cost: ¥3,300,000、Indirect Cost: ¥990,000); Fiscal Year 2022: ¥130,000 (Direct Cost: ¥100,000、Indirect Cost: ¥30,000); Fiscal Year 2021: ¥2,080,000 (Direct Cost: ¥1,600,000、Indirect Cost: ¥480,000); Fiscal Year 2020: ¥2,080,000 (Direct Cost: ¥1,600,000、Indirect Cost: ¥480,000)
• Keywords: 磁性材料 / データ駆動型AI / 第一原理計算 / データ駆動型材料探索 / 結晶性磁性材料 / 説明可能なAI / マテリアルインフォマティクス / 磁性材料探索

Outline of Research at the Start

本研究は，強い磁力と高い構造安定性をもつ結晶の発見を目的とする．そのため応用と学理を両立する物性現象の理解を重視した説明可能なAIと材料科学との融合を図り，磁石材料の開発に寄与する新しい基盤を構築する．希土類磁性材料の探索問題に適用し，系統的に安定な結晶構造を発掘し，構造・安定性・磁性の相関関係を抽出し，高性能磁石の候補材料を提案する.

Outline of Final Research Achievements

We developed an integrated framework of AI and materials science to gain a better understanding of the mechanisms underlying structural properties. Specifically, we systematically explored the candidate space of rare earth-metal (RE)-transition metal (T)-light element (X) crystalline magnetic materials and discovered alternative crystal candidates with high magnetic ability and structural stability. We selected SmFe12, a compound with a ThMn12 structure, as a host lattice due to its excellent magnetic properties. To generate a large number of candidate spaces, we used an automatic generator to substitute elements into each atomic site of the crystalline structure. We established a material search method that integrates first-principles calculations and active learning methods, resulting in the identification of promising candidates for the SmFe12 substitution structures. Our results demonstrate the effectiveness of this approach in discovering new materials with desirable properties.

Academic Significance and Societal Importance of the Research Achievements

本研究は構造物性機構の理解を重視したAIと材料科学の首尾一貫的な枠組みを構築し，RE（希土類金属）ーT（遷移金属・構造安定化元素）ーX（軽元素）結晶性磁性材料の候補空間を系統的に探索し，高磁気能で構造安定性をもつ代替結晶候補を発見することと構造物性科学に対して，理解・応用の両面で貢献するした。特に、ThMn12構造を持つ磁石材料SmFe12化合物に対して、網羅的に結晶構造の各原子サイトに元素を置換を行い、第一原理計算と能動学習手法を統合した材料探索法を確立し、SmFe12置換構造の有望な候補を見つけた。この結果、新規磁石材料の開発に貴重な構造、物性、安定性に関する知見を得ることができた。

Research Products

• [Journal Article] Evidence-based data mining method to reveal similarities between materials based on physical mechanisms (2023)
  • Author(s): Ha Minh-Quyet、Nguyen Duong-Nguyen、Nguyen Viet-Cuong、Kino Hiori、Ando Yasunobu、Miyake Takashi、Deneux Thierry、Huynh Van-Nam、Dam Hieu-Chi
  • Journal Title: Journal of Applied Physics Volume: 133 Issue: 5 Pages: 053904-053904
  • DOI: 10.1063/5.0134999
  • Peer Reviewed / Open Access / Int'l Joint Research
• [Journal Article] Machine learning-aided Genetic algorithm in investigating the structure property relationship of SmFe12-based structures (2023)
  • Author(s): Nguyen Duong-Nguyen、Dam Hieu-Chi
  • Journal Title: Journal of Applied Physics Volume: 133 Issue: 6 Pages: 063902-063902
  • DOI: 10.1063/5.0134821
  • Peer Reviewed / Open Access
• [Journal Article] Explainable active learning in investigating structure stability of SmFe12-α-βXαYβ structures X, Y {Mo, Zn, Co, Cu, Ti, Al, Ga} (2022)
  • Author(s): Nguyen Duong-Nguyen、Kino Hiori、Miyake Takashi、Dam Hieu-Chi
  • Journal Title: MRS Bulletin Volume: 48 Issue: 1 Pages: 31-44
  • DOI: 10.1557/s43577-022-00372-9
  • Peer Reviewed / Open Access
• [Journal Article] Evidence-based recommender system for high-entropy alloys (2021)
  • Author(s): Minh-Quyet Ha, Duong-Nguyen Nguyen, Viet-Cuong Nguyen, Takahiro Nagata, Toyohiro Chikyow, Hiori Kino, Takashi Miyake, Thierry Denoeux, Van-Nam Huynh, Hieu-Chi Dam
  • Journal Title: Nature Computational Science Volume: 1 Issue: 7 Pages: 470-478
  • DOI: 10.1038/s43588-021-00097-w
  • Peer Reviewed / Open Access / Int'l Joint Research
• [Journal Article] Boron cage effects on Nd-Fe-B crystal structure’s stability (2020)
  • Author(s): Nguyen Duong-Nguyen、Dao Duc-Anh、Miyake Takashi、Dam Hieu-Chi
  • Journal Title: The Journal of Chemical Physics Volume: 153 Issue: 11 Pages: 114111-114111
  • DOI: 10.1063/5.0015977
  • Peer Reviewed / Open Access
• [Journal Article] Solvent screening for efficient chemical exfoliation of graphite (2020)
  • Author(s): Nu Thanh Ton Nhan、Ha Minh-Quyet、Ikenaga Takuma、Thakur Ashutosh、Dam Hieu-Chi、Taniike Toshiaki
  • Journal Title: 2D Materials Volume: 8 Issue: 1 Pages: 015019-015019
  • DOI: 10.1088/2053-1583/abc08a
  • Peer Reviewed
• [Journal Article] Explainable machine learning for materials discovery: predicting the potentially formable Nd-Fe-B crystal structures and extracting the structure-stability relationship (2020)
  • Author(s): Pham Tien-Lam、Nguyen Duong-Nguyen、Ha Minh-Quyet、Kino Hiori、Miyake Takashi、Dam Hieu-Chi
  • Journal Title: IUCrJ Volume: 7 Issue: 6 Pages: 1036-1047
  • DOI: 10.1107/s2052252520010088
  • Peer Reviewed
• [Presentation] Learning similarities between materials from data (2021)
  • Author(s): DAM Hieu Chi
  • Organizer: Materials Research Meeting 2021
  • Int'l Joint Research / Invited