Simulating Crystallization in a Colloidal System Using State Predictive Information Bottleneck based Enhanced Sampling

TL;DR

This paper uses machine learning-enhanced molecular dynamics simulations to study crystal nucleation in colloids, introducing a novel reaction coordinate based on the State Predictive Information Bottleneck to effectively capture phase transitions.

Contribution

It presents a new application of the State Predictive Information Bottleneck to derive an effective reaction coordinate for enhanced sampling of crystallization in colloids.

Findings

Successful simulation of multi-barrier phase transitions in colloids.

Quantification of free energy differences between phases.

Identification of molecular features driving phase changes.

Abstract

We investigate crystal nucleation in supersaturated colloid suspensions using enhanced molecular dynamics simulations augmented with machine learning techniques. The simulations reveal that crystallization in the model colloidal system studied here, with particles interacting through a repulsive screened Coulomb Yukawa potential, proceeds from vapor to dense liquid droplet to crystalline phases across multiple high barriers. Employing a one-dimensional reaction coordinate derived from the State Predictive Information Bottleneck framework, our simulations capture backand-forth phase transitions across multiple barriers effectively in biased metadynamics simulations. We obtain relative free energy differences between different phases and also quantify the roles of different molecular level features in driving the phase changes.