Crystallization and non-crystallization of Lennard-Jones particles studied by molecular dynamics simulation

TL;DR

This paper uses molecular dynamics simulations to study the crystallization behavior of Lennard-Jones particles, identifying lattice types and microstructures without predefined initial lattices, and introduces a new microstructure analysis method.

Contribution

It presents a novel approach for identifying lattice structures in MD simulations without preset initial lattices and explores microstructure evolution in large-scale particle systems.

Findings

Simulated systems form fcc or hcp lattices.

Non-crystalline microstructure resembles LJ liquid near crystallization.

New method enables large-scale MD simulations with over one million particles.

Abstract

What lattice Lennard-Jones (LJ) solid favors, the lattice identification of simulated system and the microstructures of liquid and non-crystalline solid are three important questions in condensed physics and material science and are addressed in this paper. Both the crystallization and non-crystallization of LJ particles have been investigated by molecular dynamic (MD) simulation without setting any initial Bravais lattice. To identify the Bravais lattice of simulated system, two distribution functions of both the angles between one particle and its nearest neighbors and the distances between particles have been proposed. The final identification can be made by comparing these two calculated distribution functions with those of ideal Bravais lattices and checking the particle arrangement of simulated system. Our results have shown that simulated systems show either the face-centered cubic (fcc) lattice or the hexagonal close-packed (hcp) lattice. The microstructure of non-crystalline system is similar to that of LJ liquid at a temperature near the crystallization temperature, and shows no order of the second nearest neighbors in comparison with that of crystalline system. This paper has proposed a new way of investigating the microstructure of material and its evolution, and paved the way for MD simulation of large scale particle system consisting of more than one million particles.