Speaker: Professor Zhipan LIU

Institution: Department of Chemistry, Fudan University

Hosted by: Professor Zhenyang LIN

Abstract

While the underlying potential energy surface (PES) determines the structure and other properties of material, it has been frustrated to predict new materials from theory even with the advent of supercomputing facilities. The accuracy of PES and the efficiency of PES sampling are two major bottlenecks, not least because of the great complexity of material PES. This lecture introduces a “Global-to-Global” approach for material discovery by combining for the first time the global optimization method with neural network (NN) techniques. The novel global optimization method, the stochastic surface walking (SSW) method is carried out massively in parallel for generating a global training data set, the fitting of which by the atom-centered NN produces a multi-dimensional global PES; the subsequent SSW exploration of large systems with the analytic NN PES can provide key information on the thermodynamics and kinetics stability of unknown phases identified from global PES. We describe in detail the current implementation of the SSW-NN method with particular focuses on the size of the global data set and the simultaneous energy/force/stress NN training procedure. All these methods have been implemented in LASP software (www.lasphub.com). A number of important functional materials, in particular those for catalysis e.g. ZnCrO oxides, are utilized as the examples to demonstrate the automated global data set generation, the improved NN training procedure and the application in material discovery and catalysis. As a general tool for material simulation, the SSW-NN method provides an efficient and predictive platform for large-scale computational material screening.