Systematic Coarse-graining of Epoxy Resins with Machine Learning-Informed Energy Renormalization

TL;DR

This paper presents a machine learning-informed coarse-graining method for epoxy resins that accurately predicts their properties across different degrees of crosslinking with high computational efficiency.

Contribution

It introduces a novel energy renormalization approach combined with Gaussian process models for calibrating coarse-grained force fields in epoxy resins.

Findings

Excellent agreement between all-atom and CG predictions for key properties.

Simplification of calibration parameters using surrogate models.

Framework enables large-scale, chemistry-specific investigations.

Abstract

A persistent challenge in predictive molecular modeling of thermoset polymers is to capture the effects of chemical composition and degree of crosslinking (DC) on dynamical and mechanical properties with high computational efficiency. We established a new coarse-graining (CG) approach that combines the energy renormalization method with Gaussian process surrogate models of the molecular dynamics simulations. This allows a machine-learning informed functional calibration of DC-dependent CG force field parameters. Taking versatile epoxy resins consisting of Bisphenol A diglycidyl ether combined with curing agent of either 4,4-Diaminodicyclohexylmethane or polyoxypropylene diamines, we demonstrated excellent agreement between all-atom and CG predictions for density, Debye-Waller factor, Young's modulus and yield stress at any DC. We further introduce a surrogate model enabled simplification of the functional forms of 14 non-bonded calibration parameters by quantifying the uncertainty of a candidate set of high-dimensional and flexible calibration functions. The framework established provides an efficient methodology for chemistry-specific, large-scale investigations of the dynamics and mechanics of epoxy resins.