Computational prediction of the molecular configuration of three-dimensional network polymers

The three-dimensional arrangement of natural and synthetic network materials determines their application range. Control over the real-time incorporation of each building block and functional group is desired to regulate the macroscopic properties of the material from the molecular level onwards. Here we report an approach combining kinetic Monte Carlo and molecular dynamics simulations that chemically and physically predicts the interactions between building blocks in time and in space for the entire formation process of three-dimensional networks. This framework takes into account variations in inter- and intramolecular chemical reactivity, diffusivity, segmental compositions, branch/network point locations and defects. From the kinetic and three-dimensional structural information gathered, we construct structure-property relationships based on molecular descriptors such as pore size or dangling chain distribution and differentiate ideal from non-ideal structural elements. We validate such relationships by synthesizing organosilica, epoxy-amine and Diels-Alder networks with tailored properties and functions, further demonstrating the broad applicability of the platform.