Abstract
Modern research on heterogeneous catalysis calls for new techniques and methods to resolve the active site structure and reaction intermediates at the atomic scale. Here, we overview our recent progress on large-scale atomistic simulation via potential energy surface (PES) global optimization based on neural network (NN) potential, focusing on methodology details and recent applications on catalysis. The combination of stochastic surface walking (SSW) global optimization and the NN method provides a convenient and automated way to generate the transferable and robust NN potential for global PES, which can be utilized to reveal new chemistry from the unknown region of PES with an affordable computational cost. The predictive power of SSW-NN is demonstrated in several examples, where the method is applied to explore the material crystal phases, to follow the surface structure evolution under high pressure hydrogen and to determine the ternary oxide phase diagram. The limitations and future directions to develop the SSW-NN method are also discussed.
Highlights
Heterogeneous catalysts are renowned for their great complexity of material composition and surface structures
As a direct tool to correlate the atomic structure with its energy, theoretical simulations have been widely practiced in modeling catalyst structures and even predicting the activity, especially with the advent of density functional theory (DFT) calculations
The key to improve the quality of the neural network (NN) potential relies much on the representativity of the potential energy surface (PES) dataset. To overcome this deficiency of the NN potential, we propose in 2017 a global-to-global scheme to generate the global NN (G-NN) potential for material simulation
Summary
Heterogeneous catalysts are renowned for their great complexity of material composition and surface structures. The NN potential method developed in the past decade demonstrates its power in treating complex PES problems, from the gas phase reactions to material dynamics.. The NN potential method developed in the past decade demonstrates its power in treating complex PES problems, from the gas phase reactions to material dynamics.12–15 It was utilized for solving the structures of heterogeneous catalysts, for example, Pd(O), Pt(H), CuAu, CuCeO, and CuZnO.. The common global optimization methods include simulated annealing, genetic algorithm, basin hopping, and stochastic surface walking (SSW).25,26 This perspective serves to outline our recent contributions in methodology development toward catalyst PES exploration and active site identification. The SSW-NN method together with other common atomic simulation techniques is implemented in LASP software, Large-scale Atomic Simulation with neural network Potential, which allows one, within a user-friendly platform, to perform first principles calculations, NN potential generation, and atomic simulation using the NN potential.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
