Universal machine learning for the response of atomistic systems to external fields
Abstract Machine learned interatomic interaction potentials have enabled efficient and accurate molecular simulations of closed systems. However, external fields, which can greatly change the chemical structure and/or reactivity, have been seldom included in current machine learning models. This wor...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-10-01
|
| Series: | Nature Communications |
| Online Access: | https://doi.org/10.1038/s41467-023-42148-y |
| _version_ | 1797558468957175808 |
|---|---|
| author | Yaolong Zhang Bin Jiang |
| author_facet | Yaolong Zhang Bin Jiang |
| author_sort | Yaolong Zhang |
| collection | DOAJ |
| description | Abstract Machine learned interatomic interaction potentials have enabled efficient and accurate molecular simulations of closed systems. However, external fields, which can greatly change the chemical structure and/or reactivity, have been seldom included in current machine learning models. This work proposes a universal field-induced recursively embedded atom neural network (FIREANN) model, which integrates a pseudo field vector-dependent feature into atomic descriptors to represent system-field interactions with rigorous rotational equivariance. This “all-in-one” approach correlates various response properties like dipole moment and polarizability with the field-dependent potential energy in a single model, very suitable for spectroscopic and dynamics simulations in molecular and periodic systems in the presence of electric fields. Especially for periodic systems, we find that FIREANN can overcome the intrinsic multiple-value issue of the polarization by training atomic forces only. These results validate the universality and capability of the FIREANN method for efficient first-principles modeling of complicated systems in strong external fields. |
| first_indexed | 2024-03-10T17:30:53Z |
| format | Article |
| id | doaj.art-a5b311be41444e91b33bb6e8cb90894a |
| institution | Directory Open Access Journal |
| issn | 2041-1723 |
| language | English |
| last_indexed | 2024-03-10T17:30:53Z |
| publishDate | 2023-10-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Nature Communications |
| spelling | doaj.art-a5b311be41444e91b33bb6e8cb90894a2023-11-20T10:01:12ZengNature PortfolioNature Communications2041-17232023-10-0114111110.1038/s41467-023-42148-yUniversal machine learning for the response of atomistic systems to external fieldsYaolong Zhang0Bin Jiang1Key Laboratory of Precision and Intelligent Chemistry, Department of Chemical Physics, Key Laboratory of Surface and Interface Chemistry and Energy Catalysis of Anhui Higher Education Institutes, University of Science and Technology of ChinaKey Laboratory of Precision and Intelligent Chemistry, Department of Chemical Physics, Key Laboratory of Surface and Interface Chemistry and Energy Catalysis of Anhui Higher Education Institutes, University of Science and Technology of ChinaAbstract Machine learned interatomic interaction potentials have enabled efficient and accurate molecular simulations of closed systems. However, external fields, which can greatly change the chemical structure and/or reactivity, have been seldom included in current machine learning models. This work proposes a universal field-induced recursively embedded atom neural network (FIREANN) model, which integrates a pseudo field vector-dependent feature into atomic descriptors to represent system-field interactions with rigorous rotational equivariance. This “all-in-one” approach correlates various response properties like dipole moment and polarizability with the field-dependent potential energy in a single model, very suitable for spectroscopic and dynamics simulations in molecular and periodic systems in the presence of electric fields. Especially for periodic systems, we find that FIREANN can overcome the intrinsic multiple-value issue of the polarization by training atomic forces only. These results validate the universality and capability of the FIREANN method for efficient first-principles modeling of complicated systems in strong external fields.https://doi.org/10.1038/s41467-023-42148-y |
| spellingShingle | Yaolong Zhang Bin Jiang Universal machine learning for the response of atomistic systems to external fields Nature Communications |
| title | Universal machine learning for the response of atomistic systems to external fields |
| title_full | Universal machine learning for the response of atomistic systems to external fields |
| title_fullStr | Universal machine learning for the response of atomistic systems to external fields |
| title_full_unstemmed | Universal machine learning for the response of atomistic systems to external fields |
| title_short | Universal machine learning for the response of atomistic systems to external fields |
| title_sort | universal machine learning for the response of atomistic systems to external fields |
| url | https://doi.org/10.1038/s41467-023-42148-y |
| work_keys_str_mv | AT yaolongzhang universalmachinelearningfortheresponseofatomisticsystemstoexternalfields AT binjiang universalmachinelearningfortheresponseofatomisticsystemstoexternalfields |