Predicting Microstructure-Property Correlations of Polymer Nanocomposites

Polymer nanocomposites are well-known for their tunable rheological, mechanical, and electrical properties. These tunabilities are attributed to their rich phase behaviours. However, the phase change during critical operations reduces their efficiency in a given application. Moreover, the addition of nanoparticles typically increases the viscosity of polymers, making their processing considerably difficult. In this study, we investigate the formation of various microstructures within a polymer matrix and examine how their stability and processability relate to nanoparticle shape, size, concentration, and surface functionality, using large-scale coarse-grained molecular dynamics (CGMD) and non-equilibrium molecular dynamics (NEMD) simulations. We further utilize our MD simulation data to build machine learning models that rapidly predict a wide range of microstructures that are possible in a polymer nanocomposites and how they are connected to their overall rheological and electrical properties. The work provides important design rules for highly stable and easily processable polymer nanocomposites with target properties.