Machine learning surrogate models for prediction of point defect vibrational entropy
| Autor: | Clovis Lapointe, Thomas D. Swinburne, Stéphane Mallat, Laurent Proville, Charlotte Becquart, Louis Thiry, Mihai-Cosmin Marinica |
|---|---|
| Přispěvatelé: | Service de recherches de métallurgie physique (SRMP), Département des Matériaux pour le Nucléaire (DMN), CEA-Direction des Energies (ex-Direction de l'Energie Nucléaire) (CEA-DES (ex-DEN)), Commissariat à l'énergie atomique et aux énergies alternatives (CEA)-Commissariat à l'énergie atomique et aux énergies alternatives (CEA)-Université Paris-Saclay-CEA-Direction des Energies (ex-Direction de l'Energie Nucléaire) (CEA-DES (ex-DEN)), Commissariat à l'énergie atomique et aux énergies alternatives (CEA)-Commissariat à l'énergie atomique et aux énergies alternatives (CEA)-Université Paris-Saclay, Centre Interdisciplinaire de Nanoscience de Marseille (CINaM), Aix Marseille Université (AMU)-Centre National de la Recherche Scientifique (CNRS), Département d'informatique - ENS Paris (DI-ENS), École normale supérieure - Paris (ENS-PSL), Université Paris sciences et lettres (PSL)-Université Paris sciences et lettres (PSL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS), Flatiron Institute, Simons Foundation, Collège de France - Chaire Sciences des données, Collège de France (CdF (institution)), Unité Matériaux et Transformations - UMR 8207 (UMET), Centrale Lille-Institut de Chimie du CNRS (INC)-Université de Lille-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche pour l’Agriculture, l’Alimentation et l’Environnement (INRAE), GENCI (CINES/CCRT) computer center, Grant No. A0070906973, ANR-19-P3IA-0001,PRAIRIE,PaRis Artificial Intelligence Research InstitutE(2019), European Project: 633053,H2020,EURATOM-Adhoc-2014-20,EUROfusion(2014), Centre National de la Recherche Scientifique (CNRS), Centre de Mathématiques Appliquées (CMAP), Université de Versailles Saint-Quentin-en-Yvelines (UVSQ)-École polytechnique (X)-Centre National de la Recherche Scientifique (CNRS), ANR-19-CE46-0006,MeMoPAS,Mesoscale models from massively parallel atomistic simulations: uncertainty driven, self-optimizing strategies for hard materials(2019), Centre National de la Recherche Scientifique (CNRS)-Aix Marseille Université (AMU), Chaire Sciences des données, Institut de Chimie du CNRS (INC)-Université de Lille-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche pour l’Agriculture, l’Alimentation et l’Environnement (INRAE)-Centrale Lille Institut (CLIL), Université de Lille, CNRS, INRA, ENSCL, Département des Matériaux pour le Nucléaire [DMN], Centre Interdisciplinaire de Nanoscience de Marseille [CINaM], Centre de Mathématiques Appliquées [CMAP], Unité Matériaux et Transformations - UMR 8207 [UMET] |
| Jazyk: | angličtina |
| Rok vydání: | 2020 |
| Předmět: |
Hessian matrix
Materials science Physics and Astronomy (miscellaneous) 02 engineering and technology Machine learning computer.software_genre Molecular Dynamics 01 natural sciences [SPI.MAT]Engineering Sciences [physics]/Materials Machine Learning [PHYS.PHYS.PHYS-COMP-PH]Physics [physics]/Physics [physics]/Computational Physics [physics.comp-ph] Molecular dynamics symbols.namesake Harmonic approximation Dimension (vector space) [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG] [STAT.ML]Statistics [stat]/Machine Learning [stat.ML] 0103 physical sciences General Materials Science Point (geometry) 010306 general physics business.industry Vibrational Entropy [CHIM.MATE]Chemical Sciences/Material chemistry 021001 nanoscience & nanotechnology Crystallographic defect Range (mathematics) Phase space Supercell (crystal) symbols Defects Empirical Potentials Artificial intelligence 0210 nano-technology business computer |
| Zdroj: | Physical Review Materials Physical Review Materials, 2020, 4 (6), pp.063802. ⟨10.1103/PhysRevMaterials.4.063802⟩ Physical Review Materials, 2020, 4 (6), ⟨10.1103/PhysRevMaterials.4.063802⟩ Physical Review Materials, American Physical Society, 2020, 4 (6), ⟨10.1103/PhysRevMaterials.4.063802⟩ |
| ISSN: | 2475-9953 |
| DOI: | 10.1103/PhysRevMaterials.4.063802⟩ |
| Popis: | International audience; The temperature variation of the defect densities in a crystal depends on vibrational entropy. This contribution to the system thermodynamics remains computationally challenging as it requires a diagonalisation of the system's Hessian which scales as $O(N^3)$ for a crystal made of N atoms. Here, to circumvent such an heavy computational task and make it feasible even for systems containing millions of atoms the harmonic vibrational entropy of point defects is estimated directly from the relaxed atomic positions through a linear-in-descriptor machine learning approach of order O(N). With a size-independent descriptor dimension and fixed model parameters, an excellent predictive power is demonstrated on a wide range of defect configurations, supercell sizes and external deformations well outside of the training database. In particular, formation entropies in a range of 250 $k_{B}$ are predicted with less than 1.6 $k_B$ error from a training database whose formation entropies span only 25 $k_B$ (train error less than 1.0 kB. This exceptional transferability is found to hold even when the training is limited to a low energy superbasin in the phase space while the tests are performed for a different liquid-like superbasin at higher energies. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|