Relaxation in a glassy binary mixture: Mode-coupling-like power laws, dynamic heterogeneity and a new non-Gaussian parameter

TL;DR

This study uses simulations to explore the relaxation dynamics of a binary glassy mixture, revealing mode-coupling-like power laws, dynamic heterogeneity, and introducing a new non-Gaussian parameter to characterize particle motion.

Contribution

It identifies a crossover temperature marking a shift from homogeneous to heterogeneous dynamics and introduces a novel non-Gaussian parameter for analyzing particle displacements.

Findings

Power-law dependence of diffusion and relaxation on temperature.

Crossover temperature separates diffusive and hopping-like motion.

New non-Gaussian parameter peaks at maximum dynamic heterogeneity.

Abstract

We examine the relaxation of the Kob-Andersen Lennard-Jones binary mixture using Brownian dynamics computer simulations. We find that in accordance with mode-coupling theory the self-diffusion coefficient and the relaxation time show power-law dependence on temperature. However, different mode-coupling temperatures and power laws can be obtained from the simulation data depending on the range of temperatures chosen for the power-law fits. The temperature that is commonly reported as this system's mode-coupling transition temperature, in addition to being obtained from a power law fit, is a crossover temperature at which there is a change in the dynamics from the high temperature homogeneous, diffusive relaxation to a heterogeneous, hopping-like motion. The hopping-like motion is evident in the probability distributions of the logarithm of single-particle displacements: approaching the commonly reported mode-coupling temperature these distributions start exhibiting two peaks. Notably, the temperature at which the hopping-like motion appears for the smaller particles is slightly higher than that at which the hopping-like motion appears for the larger ones. We define and calculate a new non-Gaussian parameter whose maximum occurs approximately at the time at which the two peaks in the probability distribution of the logarithm of displacements are most evident.