Stoichiometry-dependent network heterogeneity during epoxy curing: Experiments and coarse-grained molecular dynamics simulations

The formation of heterogeneous network structures in epoxy thermosets strongly influences their final properties. However, the link between local curing reactions and mesoscale network heterogeneity remains unclear. In this study, the mechanism of stoichiometry-dependent heterogeneous network formation in an epoxy/amine thermosetting system is investigated by combining experimental characterization and coarse-grained molecular dynamics (CGMD) simulations. Experimentally, the epoxy-based cure conversion increased with an increasing epoxy equivalent ratio (EER), whereas the heterogeneity quantified by the full width at half maximum of the loss tangent, tan δ , peak was significantly higher in low-EER systems. The glass transition temperature reached its maximum at the stoichiometric composition and decreased under both low- and high-EER conditions. CGMD simulations reproduced the experimental EER dependence of cure progression and network heterogeneity. Structural visualization and voxel-based analysis showed that the experimentally observed heterogeneity could be interpreted as mesoscale aggregation of non-percolated molecules embedded in the percolated network. Topological analysis further revealed that low-EER systems preferentially formed highly branched percolated networks dominated by tertiary amines, whereas high-EER systems contained a larger fraction of less-branched and linear segments. The results demonstrate that stoichiometry governs network heterogeneity through the overall extent of cure, mesoscale retention of non-percolated molecules, and topology of the percolated network. This study provides a mechanistic picture of how stoichiometry controls the heterogeneous network formation in epoxy thermosets.