Structural Transformations in Single-Crystalline AgPd Nanoalloys from Multiscale Deep Potential Molecular Dynamics.

Longfei Guo,Tao Jin,Shuang Shan,Quan Tang,Zhen Li,Chongyang Wang,Junpeng Wang,Bowei Pan,Qiao Wang,Fuyi Chen
DOI: https://doi.org/10.1063/5.0158918
IF: 4.304
2023-01-01
The Journal of Chemical Physics
Abstract:AgPd nanoalloys often undergo structural evolution during catalytic reactions; the mechanism underlying such restructuring remains largely unknown due to the use of oversimplified interatomic potentials in simulations. Herein, a deep-learning potential is developed for AgPd nanoalloys based on a multiscale dataset spanning from nanoclusters to bulk configurations, exhibits precise predictions of mechanical properties and formation energies with near-density functional theory accuracy, calculates the surface energies closer to experimental values compared to those obtained by Gupta potentials, and is applied to investigate the shape reconstruction of single-crystalline AgPd nanoalloys from cuboctahedron (Oh) to icosahedron (Ih) geometries. The Oh to Ih shape restructuring is thermodynamically favorable and occurs at 11 and 92 ps for Pd-55@Ag-254 and Ag-147@Pd-162 nanoalloys, respectively. During the shape reconstruction of Pd@Ag nanoalloys, concurrent surface restructuring of the (100) facet and internal multi-twinned phase change are observed with collaborative displacive characters. The presence of vacancies can influence the final product and reconstructing rate of Pd@Ag core-shell nanoalloys. The Ag outward diffusion on Ag@Pd nanoalloys is more pronounced in Ih geometry compared to Oh geometry and can be further accelerated by the Oh to Ih deformation. The deformation of single-crystalline Pd@Ag nanoalloys is characterized by a displacive transformation involving the collaborative displacement of a large number of atoms, distinguishing it from the diffusion-coupled transformation of Ag@Pd nanoalloys.
What problem does this paper attempt to address?