Speaker: Prof. Akira Nakayama, the University of Tokyo

Time: 15:00 p.m., July 2, 2024, GMT+8

Venue:  Room B229, College of Chemistry and Molecular Engineering

Abstract: 

The rise of neural network potentials (NNP) for interatomic interactions has enabled atomistic simulations of complex systems with unprecedented system sizes and long simulation times.  The NNP provides a reliable description of interatomic interactions for various types of bonding with high fidelity, and by efficiently acquiring and adaptively updating training data, usually obtained by the first-principles calculations such as DFT, the NNP-based simulations are becoming a versatile tool for investigating the microscopic structure and dynamical behavior of various systems.
We present our recent works using NNP on the following topics.
　　(i)Long-range proton and hydroxide ion transfer dynamics at the water/CeO2 interface in the nanosecond regime: reactive molecular dynamics simulations and kinetic analysis.[1]
　　(ii)On-the-fly kinetic Monte Carlo simulations for surface diffusion and reaction. [2]
　　(iii)Grand canonical Monte Carlo simulations for adsorption of hydrogen atoms on metal surfaces and single-atom alloys. [3,4]

　　[1] T. Kobayashi, T. Ikeda, and A. Nakayama, Chem. Sci. 15, 6816 (2024).
　　[2] T. Yokaichiya, T. Ikeda, K. Muraoka, and A. Nakayama, J. Chem. Phys. 160, 204108 (2024).
　　[3] T. Ikeda and A. Nakayama (submitted)
　　[4] T. Kanno, T. Ikeda, and A. Nakayama (in preparation).

Source: College of Chemistry and Molecular Engineering, PKU