Harnessing Machine Learning for Quantum-Accurate Predictions of Non-Equilibrium Behavior in 2D Materials
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912325269716992 |
|---|---|
| author | Zhang, Yue Appleton, Robert J. Lin, Kui McCarthy, Megan J. Paci, Jeffrey T. Sankaranarayanan, Subramanian K. R. S. Strachan, Alejandro Espinosa, Horacio D. |
| author_facet | Zhang, Yue Appleton, Robert J. Lin, Kui McCarthy, Megan J. Paci, Jeffrey T. Sankaranarayanan, Subramanian K. R. S. Strachan, Alejandro Espinosa, Horacio D. |
| contents | Accurately predicting the non-equilibrium mechanical properties of two-dimensional (2D) materials is essential for understanding their deformation, thermo-mechanical properties, and failure mechanisms. In this study, we parameterize and evaluate two machine learning (ML) interatomic potentials, SNAP and Allegro, for modeling the non-equilibrium behavior of monolayer MoSe2. Using a density functional theory (DFT) derived dataset, we systematically compare their accuracy and transferability against the physics-based Tersoff force field. Our results show that SNAP and Allegro significantly outperform Tersoff, achieving near-DFT accuracy while maintaining computational efficiency. Allegro surpasses SNAP in both accuracy and efficiency due to its advanced neural network architecture. Both ML potentials demonstrate strong transferability, accurately predicting out-of-sample properties such as surface stability, inversion domain formation, and fracture toughness. Unlike Tersoff, SNAP and Allegro reliably model temperature-dependent edge stabilities and phase transformation pathways, aligning closely with DFT benchmarks. Notably, their fracture toughness predictions closely match experimental measurements, reinforcing their suitability for large-scale simulations of mechanical failure in 2D materials. This study establishes ML-based force fields as a powerful alternative to traditional potentials for modeling non-equilibrium mechanical properties in 2D materials. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_09686 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Harnessing Machine Learning for Quantum-Accurate Predictions of Non-Equilibrium Behavior in 2D Materials Zhang, Yue Appleton, Robert J. Lin, Kui McCarthy, Megan J. Paci, Jeffrey T. Sankaranarayanan, Subramanian K. R. S. Strachan, Alejandro Espinosa, Horacio D. Materials Science Accurately predicting the non-equilibrium mechanical properties of two-dimensional (2D) materials is essential for understanding their deformation, thermo-mechanical properties, and failure mechanisms. In this study, we parameterize and evaluate two machine learning (ML) interatomic potentials, SNAP and Allegro, for modeling the non-equilibrium behavior of monolayer MoSe2. Using a density functional theory (DFT) derived dataset, we systematically compare their accuracy and transferability against the physics-based Tersoff force field. Our results show that SNAP and Allegro significantly outperform Tersoff, achieving near-DFT accuracy while maintaining computational efficiency. Allegro surpasses SNAP in both accuracy and efficiency due to its advanced neural network architecture. Both ML potentials demonstrate strong transferability, accurately predicting out-of-sample properties such as surface stability, inversion domain formation, and fracture toughness. Unlike Tersoff, SNAP and Allegro reliably model temperature-dependent edge stabilities and phase transformation pathways, aligning closely with DFT benchmarks. Notably, their fracture toughness predictions closely match experimental measurements, reinforcing their suitability for large-scale simulations of mechanical failure in 2D materials. This study establishes ML-based force fields as a powerful alternative to traditional potentials for modeling non-equilibrium mechanical properties in 2D materials. |
| title | Harnessing Machine Learning for Quantum-Accurate Predictions of Non-Equilibrium Behavior in 2D Materials |
| topic | Materials Science |
| url | https://arxiv.org/abs/2504.09686 |