Bayesian force fields from active learning for simulation of inter-dimensional transformation of stanene
Abstract We present a way to dramatically accelerate Gaussian process models for interatomic force fields based on many-body kernels by mapping both forces and uncertainties onto functions of low-dimensional features. This allows for automated active learning of models combining near-quantum accurac...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2021-03-01
|
| Series: | npj Computational Materials |
| Online Access: | https://doi.org/10.1038/s41524-021-00510-y |
| _version_ | 1818841741335199744 |
|---|---|
| author | Yu Xie Jonathan Vandermause Lixin Sun Andrea Cepellotti Boris Kozinsky |
| author_facet | Yu Xie Jonathan Vandermause Lixin Sun Andrea Cepellotti Boris Kozinsky |
| author_sort | Yu Xie |
| collection | DOAJ |
| description | Abstract We present a way to dramatically accelerate Gaussian process models for interatomic force fields based on many-body kernels by mapping both forces and uncertainties onto functions of low-dimensional features. This allows for automated active learning of models combining near-quantum accuracy, built-in uncertainty, and constant cost of evaluation that is comparable to classical analytical models, capable of simulating millions of atoms. Using this approach, we perform large-scale molecular dynamics simulations of the stability of the stanene monolayer. We discover an unusual phase transformation mechanism of 2D stanene, where ripples lead to nucleation of bilayer defects, densification into a disordered multilayer structure, followed by formation of bulk liquid at high temperature or nucleation and growth of the 3D bcc crystal at low temperature. The presented method opens possibilities for rapid development of fast accurate uncertainty-aware models for simulating long-time large-scale dynamics of complex materials. |
| first_indexed | 2024-12-19T04:30:54Z |
| format | Article |
| id | doaj.art-f04a0cef3d874031836d2287821283f8 |
| institution | Directory Open Access Journal |
| issn | 2057-3960 |
| language | English |
| last_indexed | 2024-12-19T04:30:54Z |
| publishDate | 2021-03-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | npj Computational Materials |
| spelling | doaj.art-f04a0cef3d874031836d2287821283f82022-12-21T20:35:53ZengNature Portfolionpj Computational Materials2057-39602021-03-017111010.1038/s41524-021-00510-yBayesian force fields from active learning for simulation of inter-dimensional transformation of staneneYu Xie0Jonathan Vandermause1Lixin Sun2Andrea Cepellotti3Boris Kozinsky4John A. Paulson School of Engineering and Applied Sciences, Harvard UniversityJohn A. Paulson School of Engineering and Applied Sciences, Harvard UniversityJohn A. Paulson School of Engineering and Applied Sciences, Harvard UniversityJohn A. Paulson School of Engineering and Applied Sciences, Harvard UniversityJohn A. Paulson School of Engineering and Applied Sciences, Harvard UniversityAbstract We present a way to dramatically accelerate Gaussian process models for interatomic force fields based on many-body kernels by mapping both forces and uncertainties onto functions of low-dimensional features. This allows for automated active learning of models combining near-quantum accuracy, built-in uncertainty, and constant cost of evaluation that is comparable to classical analytical models, capable of simulating millions of atoms. Using this approach, we perform large-scale molecular dynamics simulations of the stability of the stanene monolayer. We discover an unusual phase transformation mechanism of 2D stanene, where ripples lead to nucleation of bilayer defects, densification into a disordered multilayer structure, followed by formation of bulk liquid at high temperature or nucleation and growth of the 3D bcc crystal at low temperature. The presented method opens possibilities for rapid development of fast accurate uncertainty-aware models for simulating long-time large-scale dynamics of complex materials.https://doi.org/10.1038/s41524-021-00510-y |
| spellingShingle | Yu Xie Jonathan Vandermause Lixin Sun Andrea Cepellotti Boris Kozinsky Bayesian force fields from active learning for simulation of inter-dimensional transformation of stanene npj Computational Materials |
| title | Bayesian force fields from active learning for simulation of inter-dimensional transformation of stanene |
| title_full | Bayesian force fields from active learning for simulation of inter-dimensional transformation of stanene |
| title_fullStr | Bayesian force fields from active learning for simulation of inter-dimensional transformation of stanene |
| title_full_unstemmed | Bayesian force fields from active learning for simulation of inter-dimensional transformation of stanene |
| title_short | Bayesian force fields from active learning for simulation of inter-dimensional transformation of stanene |
| title_sort | bayesian force fields from active learning for simulation of inter dimensional transformation of stanene |
| url | https://doi.org/10.1038/s41524-021-00510-y |
| work_keys_str_mv | AT yuxie bayesianforcefieldsfromactivelearningforsimulationofinterdimensionaltransformationofstanene AT jonathanvandermause bayesianforcefieldsfromactivelearningforsimulationofinterdimensionaltransformationofstanene AT lixinsun bayesianforcefieldsfromactivelearningforsimulationofinterdimensionaltransformationofstanene AT andreacepellotti bayesianforcefieldsfromactivelearningforsimulationofinterdimensionaltransformationofstanene AT boriskozinsky bayesianforcefieldsfromactivelearningforsimulationofinterdimensionaltransformationofstanene |