Crystallization of the P3Sn4 Phase upon Cooling P2Sn5 Liquid by Molecular Dynamics Simulation Using a Machine Learning Interatomic Potential

Thumbnail Image
Zhang, Chao
Sun, Yang
Wang, Hai-Di
Zhang, Feng
Wen, Tong-Qi
Ho, Kai-Ming
Wang, Cai-Zhuang
Major Professor
Committee Member
Journal Title
Journal ISSN
Volume Title
Research Projects
Organizational Units
Organizational Unit
Ames National Laboratory

Ames National Laboratory is a government-owned, contractor-operated national laboratory of the U.S. Department of Energy (DOE), operated by and located on the campus of Iowa State University in Ames, Iowa.

For more than 70 years, the Ames National Laboratory has successfully partnered with Iowa State University, and is unique among the 17 DOE laboratories in that it is physically located on the campus of a major research university. Many of the scientists and administrators at the Laboratory also hold faculty positions at the University and the Laboratory has access to both undergraduate and graduate student talent.

Organizational Unit
Physics and Astronomy
Physics and astronomy are basic natural sciences which attempt to describe and provide an understanding of both our world and our universe. Physics serves as the underpinning of many different disciplines including the other natural sciences and technological areas.
Journal Issue
Is Version Of
Ames Laboratory

We performed molecular dynamics simulations to study the crystallization of the P3Sn4 phase from P2Sn5 liquid using a machine learning (ML) interatomic potential with desirable efficiency and accuracy. Our results capture the liquid properties of P2Sn5 at 1300 K, which is well above the melting temperature. The phase separation and crystallization are observed when P2Sn5 liquid is cooled down below 832 and 505 K, respectively. The simulation results are in good agreement with the experimentally observed phase transformation behaviors and provide useful insights into the complex nucleation and crystallization process at the details of atomistic scale. Our work also demonstrated that ML interatomic potentials based on neural network deep learning are robust and capable of accurately describing the energetics and kinetics of complex materials through molecular dynamics simulations.