Direct Prediction of Phonon Density of States With Euclidean Neural Networks
© 2021 The Authors. Advanced Science published by Wiley-VCH GmbH Machine learning has demonstrated great power in materials design, discovery, and property prediction. However, despite the success of machine learning in predicting discrete properties, challenges remain for continuous property predic...
| Prif Awduron: | , , , , , , , , , |
|---|---|
| Awduron Eraill: | |
| Fformat: | Erthygl |
| Iaith: | English |
| Cyhoeddwyd: |
Wiley
2021
|
| Mynediad Ar-lein: | https://hdl.handle.net/1721.1/133320 |
| _version_ | 1826188929291780096 |
|---|---|
| author | Chen, Zhantao Andrejevic, Nina Smidt, Tess Ding, Zhiwei Xu, Qian Chi, Yen‐Ting Nguyen, Quynh T Alatas, Ahmet Kong, Jing Li, Mingda |
| author2 | Massachusetts Institute of Technology. Department of Mechanical Engineering |
| author_facet | Massachusetts Institute of Technology. Department of Mechanical Engineering Chen, Zhantao Andrejevic, Nina Smidt, Tess Ding, Zhiwei Xu, Qian Chi, Yen‐Ting Nguyen, Quynh T Alatas, Ahmet Kong, Jing Li, Mingda |
| author_sort | Chen, Zhantao |
| collection | MIT |
| description | © 2021 The Authors. Advanced Science published by Wiley-VCH GmbH Machine learning has demonstrated great power in materials design, discovery, and property prediction. However, despite the success of machine learning in predicting discrete properties, challenges remain for continuous property prediction. The challenge is aggravated in crystalline solids due to crystallographic symmetry considerations and data scarcity. Here, the direct prediction of phonon density-of-states (DOS) is demonstrated using only atomic species and positions as input. Euclidean neural networks are applied, which by construction are equivariant to 3D rotations, translations, and inversion and thereby capture full crystal symmetry, and achieve high-quality prediction using a small training set of (Formula presented.) examples with over 64 atom types. The predictive model reproduces key features of experimental data and even generalizes to materials with unseen elements, and is naturally suited to efficiently predict alloy systems without additional computational cost. The potential of the network is demonstrated by predicting a broad number of high phononic specific heat capacity materials. The work indicates an efficient approach to explore materials' phonon structure, and can further enable rapid screening for high-performance thermal storage materials and phonon-mediated superconductors. |
| first_indexed | 2024-09-23T08:06:51Z |
| format | Article |
| id | mit-1721.1/133320 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T08:06:51Z |
| publishDate | 2021 |
| publisher | Wiley |
| record_format | dspace |
| spelling | mit-1721.1/1333202023-02-22T21:30:40Z Direct Prediction of Phonon Density of States With Euclidean Neural Networks Chen, Zhantao Andrejevic, Nina Smidt, Tess Ding, Zhiwei Xu, Qian Chi, Yen‐Ting Nguyen, Quynh T Alatas, Ahmet Kong, Jing Li, Mingda Massachusetts Institute of Technology. Department of Mechanical Engineering Massachusetts Institute of Technology. Department of Materials Science and Engineering Massachusetts Institute of Technology. Department of Physics Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Massachusetts Institute of Technology. Department of Nuclear Science and Engineering © 2021 The Authors. Advanced Science published by Wiley-VCH GmbH Machine learning has demonstrated great power in materials design, discovery, and property prediction. However, despite the success of machine learning in predicting discrete properties, challenges remain for continuous property prediction. The challenge is aggravated in crystalline solids due to crystallographic symmetry considerations and data scarcity. Here, the direct prediction of phonon density-of-states (DOS) is demonstrated using only atomic species and positions as input. Euclidean neural networks are applied, which by construction are equivariant to 3D rotations, translations, and inversion and thereby capture full crystal symmetry, and achieve high-quality prediction using a small training set of (Formula presented.) examples with over 64 atom types. The predictive model reproduces key features of experimental data and even generalizes to materials with unseen elements, and is naturally suited to efficiently predict alloy systems without additional computational cost. The potential of the network is demonstrated by predicting a broad number of high phononic specific heat capacity materials. The work indicates an efficient approach to explore materials' phonon structure, and can further enable rapid screening for high-performance thermal storage materials and phonon-mediated superconductors. 2021-10-27T19:52:07Z 2021-10-27T19:52:07Z 2021 2021-08-11T17:06:54Z Article http://purl.org/eprint/type/JournalArticle https://hdl.handle.net/1721.1/133320 en 10.1002/ADVS.202004214 Advanced Science Creative Commons Attribution 4.0 International license https://creativecommons.org/licenses/by/4.0/ application/pdf Wiley Wiley |
| spellingShingle | Chen, Zhantao Andrejevic, Nina Smidt, Tess Ding, Zhiwei Xu, Qian Chi, Yen‐Ting Nguyen, Quynh T Alatas, Ahmet Kong, Jing Li, Mingda Direct Prediction of Phonon Density of States With Euclidean Neural Networks |
| title | Direct Prediction of Phonon Density of States With Euclidean Neural Networks |
| title_full | Direct Prediction of Phonon Density of States With Euclidean Neural Networks |
| title_fullStr | Direct Prediction of Phonon Density of States With Euclidean Neural Networks |
| title_full_unstemmed | Direct Prediction of Phonon Density of States With Euclidean Neural Networks |
| title_short | Direct Prediction of Phonon Density of States With Euclidean Neural Networks |
| title_sort | direct prediction of phonon density of states with euclidean neural networks |
| url | https://hdl.handle.net/1721.1/133320 |
| work_keys_str_mv | AT chenzhantao directpredictionofphonondensityofstateswitheuclideanneuralnetworks AT andrejevicnina directpredictionofphonondensityofstateswitheuclideanneuralnetworks AT smidttess directpredictionofphonondensityofstateswitheuclideanneuralnetworks AT dingzhiwei directpredictionofphonondensityofstateswitheuclideanneuralnetworks AT xuqian directpredictionofphonondensityofstateswitheuclideanneuralnetworks AT chiyenting directpredictionofphonondensityofstateswitheuclideanneuralnetworks AT nguyenquynht directpredictionofphonondensityofstateswitheuclideanneuralnetworks AT alatasahmet directpredictionofphonondensityofstateswitheuclideanneuralnetworks AT kongjing directpredictionofphonondensityofstateswitheuclideanneuralnetworks AT limingda directpredictionofphonondensityofstateswitheuclideanneuralnetworks |