Machine Learning a General-Purpose Interatomic Potential for Silicon

By A. P. Bart'ok, James R. Kermode, Noam Bernstein, Gábor Csányi. Physical Review X, 2018

457 citations recorded by the source corpus.

The success of first principles electronic structure calculation for predictive modeling in chemistry, solid state physics, and materials science is constrained by the limitations on simulated length and time scales due to computational cost and its scaling. Techniques based on machine learning ideas for interpolating the Born-Oppenheimer potential energy surface without explicitly describing electrons have recently shown great promise, but accurately and efficiently fitting the physically relevant space of configurations has remained a challenging goal. Here we present a Gaussian Approximation Potential for silicon that achieves this milestone, accurately reproducing density functional theory reference results for a wide range of observable properties, including crystal, liquid, and amorphous bulk phases, as well as point, line, and plane defects. We demonstrate that this new potential enables calculations that would be extremely expensive with a first principles electronic structure method, such as finite temperature phase boundary lines, self-diffusivity in the liquid, formation of the amorphous by slow quench, and dynamic brittle fracture. We show that the uncertainty quantification inherent to the Gaussian process regression framework gives a qualitative estimate of the potential's accuracy for a given atomic configuration. The success of this model shows that it is indeed possible to create a useful machine-learning-based interatomic potential that comprehensively describes a material, and serves as a template for the development of such models in the future.

Subfields: Computational Theory and Mathematics, Materials Chemistry

View Paper: https://www.semanticscholar.org/paper/2ae0ab5303cb374e79193da4846c193c666acc67

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Theory and Mathematics

In Materials Chemistry

Citations in the corpus

Machine Learning for High-entropy Alloys: Progress, Challenges and Opportunities (2022, methodology): “In GAP [147, 144, 155, 156], the total energy is predicted by the Gaussian process regression (GPR) method that measures the “similarity” to the reference of atomic environments [157].”
Neuroevolution machine learning potentials: Combining high accuracy and low cost in atomistic simulations and application to heat transport (2021, unknown): “Because a good machine learning potential should be able to account for nearly all the phases of a given material, as demonstrated for elementary silicon [47], phosphorus [48], and carbon [49, 50], we will consider ﬁtting a general-purpose potential for silicon.”
Machine-Learning Interatomic Potentials for Materials Science (2021, background): “Nevertheless, some of the recent potentials developed with this approach, such as the GAP potentials for Si [100] and C [139], do reproduce an incredibly broad spectrum of physical properties with high accuracy.”
Deep Potentials for Materials Science (2022, background): “The machine learning potential models cannot be “extrapolated” to the configurations that they are not trained on [32], thus the sampling should generate training datasets that explore the relevant configuration space as completely as possible.”
End-to-End Differentiable , Hardware Accelerated , Molecular Dynamics in Pure Python (2019, methodology): “…and more accurate (Behler & Parrinello, 2007; Behler, 2011; Artrith et al., 2011; Artrith & Behler, 2012; Artrith et al., 2018; Deringer et al., 2018a; Bartók et al., 2018; Yao et al., 2018; Seko et al., 2015; Deringer et al., 2018b; Cubuk et al., 2017a; Faber et al., 2017; Gilmer et al., 2017).”
JAX, M.D.: End-to-End Differentiable, Hardware Accelerated, Molecular Dynamics in Pure Python (2019, background): “…and more accurate (Behler & Parrinello, 2007; Behler, 2011; Artrith et al., 2011; Artrith & Behler, 2012; Artrith et al., 2018; Deringer et al., 2018a; Bartók et al., 2018; Yao et al., 2018; Seko et al., 2015; Deringer et al., 2018b; Cubuk et al., 2017a; Faber et al., 2017; Gilmer et al., 2017).”
Physically informed artificial neural networks for atomistic modeling of materials (2018, background): “This explains why the existing ML potentials are usually (with rare exceptions 39 ) narrowly focused on, and only tested for, a particular type of physical properties.”
Machine-learned potentials for next-generation matter simulations (2021, background): “5a) 101–103 , simulate phase diagrams 104 , find new poly-morphs or crystal structures (Fig.”