Glassy Dynamics from First-Principles Simulations

Florian Pabst,Stefano Baroni

2024-08-10

Abstract:The microscopic understanding of the dramatic increase in viscosity of liquids when cooled towards the glass transition is a major unresolved issue in condensed matter physics. Here, we use machine learning methods to accelerate molecular dynamics simulations with first-principles accuracy for the glass-former toluene. We show that the increase in viscosity is intimately linked to the increasing number of dynamically correlated molecules $N^*$. While certain hallmark features of glassy dynamics, like physical aging, are linked to $N^*$ as well, others, like relaxation stretching, are not.

Disordered Systems and Neural Networks

What problem does this paper attempt to address?

The paper aims to study the dynamics of the glassy state, specifically exploring the reasons for the increase in activation energy \(E^*(T)\) during the cooling process of liquids. Specifically, the authors chose toluene as the research subject because toluene is a small molecule containing only two types of atoms, and its chemical structure has been thoroughly experimentally studied. To achieve this goal, the authors used machine learning methods to train a neural network potential, enabling molecular dynamics simulations with first-principles accuracy, thereby significantly extending the simulation time scale.

The main findings of the paper include:
- There is a proportional relationship between the number of dynamically relevant molecules \(N^*\) and the apparent activation energy \(E^*_\text{app}(T)\).
- The study shows that as the temperature decreases, the increase in the number of dynamically relevant molecules \(N^*\) is closely related to the significant increase in viscosity.
- The paper also discusses other features of glassy state dynamics, such as the relationship between physical aging and \(N^*\), and the irrelevance of relaxation stretching to dynamic heterogeneity.

Through these studies, the authors demonstrated all the features of glassy state dynamics observed in real liquids and validated some theoretical models. Additionally, they found that some common assumptions do not hold, such as the non-existence of a transition between high-temperature liquids and low-temperature glassy states. These findings provide new insights into understanding the dynamics of the glassy state.