First Principles Neural Network Potentials for Reactive Simulations of Large Molecular and Condensed Systems. Issue 42 (18th August 2017)

• Record Type: Journal Article
• Authors: Behler, Jörg
• Is Part Of: Angewandte Chemie international edition. Volume 56:Issue 42(2017)
• Page Start: 12828
• Page End: 12840
• Publication Date: 2017-08-18
• Subjects: computational chemistry -- density functional calculations -- molecular dynamics -- neural networks -- potential energy surfaces; Chemistry -- Periodicals; 540
• DOI: 10.1002/anie.201703114
• Languages: English
• ISSNs: 1433-7851

Abstract

Abstract: Modern simulation techniques have reached a level of maturity which allows a wide range of problems in chemistry and materials science to be addressed. Unfortunately, the application of first principles methods with predictive power is still limited to rather small systems, and despite the rapid evolution of computer hardware no fundamental change in this situation can be expected. Consequently, the development of more efficient but equally reliable atomistic potentials to reach an atomic level understanding of complex systems has received considerable attention in recent years. A promising new development has been the introduction of machine learning (ML) methods to describe the atomic interactions. Once trained with electronic structure data, ML potentials can accelerate computer simulations by several orders of magnitude, while preserving quantum mechanical accuracy. This Review considers the methodology of an important class of ML potentials that employs artificial neural networks. Abstract : Potential energy surfaces : A new class of interatomic potentials has emerged in recent years employing machine learning (ML) techniques. These potentials combine the accuracy of first principles methods with the efficiency of simple classical force fields. This Review discusses the methodology of an important class of ML potentials that makes use of artificial neural networks and is applicable to complex systems.