High-dimensional neural-network potentials for multicomponent systems: Applications to zinc oxide

Abstract

Artificial neural networks represent an accurate and efficient tool to construct high-dimensional potentialenergy surfaces based on first-principles data. However, so far the main drawback of this method has been the limitation to a single atomic species. We present a generalization to compounds of arbitrary chemical composition, which now enables simulations of a wide range of systems containing large numbers of atoms. The required incorporation of long-range interactions is achieved by combining the numerical accuracy of neural networks with an electrostatic term based on environment-dependent charges. Using zinc oxide as a benchmark system we show that the neural network potential-energy surface is in excellent agreement with density-functional theory reference calculations, while the evaluation is many orders of magnitude faster.

Neural network interatomic potential for the phase change material GeTe

“…11,13,15 This scheme has been recently extended to binary systems by including long-range Coulomb interaction between environment-dependent ionic charges. 16,36 In the case of GeTe, we are faced with the problem of developing a potential suitable to describe both the semiconducting crystalline and amorphous phases as well as the metallic liquid. As a first step toward the development of a NN potential for GeTe, we neglect long-range Coulomb interactions for atoms being separated by a larger distance than the cutoff radius of the symmetry functions.…”

High-dimensional neural network potentials for metal surfaces: A prototype study for copper

“…Therefore, the present method will be most useful for systems containing up to three or four different elements, while a large number of atoms of each element can be present. First results for the binary system zinc oxide have been published recently, 56 and work on further systems is in progress.…”

“…of bulk materials like silicon, 51,52 sodium, 53 carbon, 54,55 and, augmented by an electrostatic energy term, to zinc oxide. 56 The goal of the present work is to explore the applicability to solid surfaces employing copper as a benchmark system.…”

From DFT to machine learning: recent approaches to materials science–a review

“…Gaussian approximation potentials (GAPs) have been extensively used to study different systems, such as elemental boron [422], amorphous carbon [423,424], silicon [425], thermal properties of amorphous GeTe and carbon [426], thermomechanics and defects of iron [427], prediction structures of inorganic crystals by combing ML with random search [428], λ-SOAP method for tensorial properties of atomistic systems [247], and a unified framework to predict the properties of materials and molecules such as silicon, organic molecules and proteins ligands [429]. A recent review of applications of high-dimensional neural neural network potentials [430] summarized the notable number of molecular and materials systems studied, which ranges from simple semiconductors such as silicon [233,431,432] and ZnO [433], to more complex systems such as water and metallic clusters [434], molecules [435][436][437], surfaces [438,439], and liquid/solid interfaces [414,440]. Force fields for nanoclusters have been developed with 2-, 3-, and many-body descriptors [441], and the hydrogen adsorption on nanoclusters was described with structural descriptors such as SOAP [442].…”