Theoretical study of impurity-doped Mg2Si thermoelectric materials using first-principles calculations and machine learning technique

2022 Fiscal Year Final Research Report

• Project/Area Number: 20K05681
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 36020:Energy-related chemistry
• Research Institution: Shimane University
• Principal Investigator: Hirayama Naomi 島根大学, 学術研究院理工学系, 准教授 (70581750)
• Project Period (FY): 2020-04-01 – 2023-03-31
• Keywords: 熱電材料 / 第一原理計算 / 機械学習ポテンシャル

Outline of Final Research Achievements

Our first-principles calculations for Mg2Si systems doped with impurities have revealed that substituting Mg atoms with iso-electronic impurities (such as Ca) results in favorable electronic states for improving thermoelectric performance. Calculations for impurity-doped SrSi2 systems reproduced the narrow band gap and revealed the influence of point defects on their carrier conduction.
Furthermore, the effects of phonons on the electrical conductivity of Sb-doped Mg2Si systems were elucidated through electronic state calculations using the KKR-CPA method and thermoelectric calculations based on the Kubo-Greenwood formula. Moreover, the observed behavior in Ag-doped systems in experiments was successfully reproduced.
Moreover, molecular dynamics calculations using a machine learning potential successfully reproduced stable structures obtained from first-principles calculations. However, there remain challenges in terms of computational cost for replicating polycrystalline structures.

Free Research Field

物性物理学

Academic Significance and Societal Importance of the Research Achievements

近年，エネルギー資源の枯渇と環境汚染の問題が加速するなか，熱電変換技術の本格的な普及が望まれている．しかし，既存の主要な熱電材料はBiやTe, Pb等の重金属を含むものが多く，毒性や資源希少性の問題があった．近年厳しさを増す環境規制と資源問題に対応し，かつ高出力な新規熱電材料の開発が喫緊の課題である．
本研究で作成した高精度な機械学習ポテンシャルや，熱電半導体Mg2Siの有限温度における物性値をよく再現することが示されたKKR-CPA法に基づく計算手法は，材料の微視的構造や熱電輸送特性を調査する上で強力なツールになると期待される．