2020 Fiscal Year Final Research Report
Designing novel functional cluster-assembled materials via machine learning and first principle calculations
| Project/Area Number |
17K14803
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Physical properties of metals/Metal-base materials
|
|---|
| Research Institution | Hokkaido University (2019-2020)
National Institute for Materials Science (2017-2018) |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2017-04-01 – 2021-03-31
|
| Keywords | マテリアルズインフォマティクス / 原子クラスター / 人工結晶 / 機械学習 / 触媒 / データ科学 / 材料 / ナノ材料 |
|---|
| Outline of Final Research Achievements |
Atomic clusters consist of a few to tens of atoms and are found to have significantly different structures and properties when compared to their bulk and nanoparticle states. Atomic clusters are treated as building blocks for the design of artificial crystals. The difficulty of creating such crystals often lies upon finding suitable binders. High throughput first principle calculations are carried out and determine that K3O cluster acts as a good binder. As a result, the artificial crystal [Cu12FeK3O]6 is designed. In addition, implementation of machine learning unveils the growth of silver clusters where the thresholds between silver clusters, nanoparticles, and bulk are determined. Lastly, techniques used in this research are expanded towards the design of functional two- dimensional materials and towards catalysts development. Hence, major advancement was achieved where one's understanding of clusters is greatly expanded.
|
| Free Research Field |
マテリアルズインフォマティクス
|
| Academic Significance and Societal Importance of the Research Achievements |
これまでの材料設計では原子1つ基本に材料設計がおこなわれてきたが、数個の原子からなる特異な物性を持つ原子クラスターを原子1つとしてとらえることにより、材料設計の可能性が膨大に広がることを示した。さらにハイスループット計算や機械学習など触媒科学や2次元材料の設計につながる基盤技術も確立されたため、今後の展開が期待される。
|
