Loading-Dependent Structural Evolution and Thermal Transport in BNNS/Epoxy Composites: A Molecular Dynamics Study
The increasing integration density of electronic packaging places growing demands on electrically insulating materials with improved heat dissipation. Here, molecular dynamics simulations were used to investigate randomly dispersed boron nitride nanosheet (BNNS)/epoxy composites and clarify how BNNS loading affects structure, thermal transport, and thermomechanical response under a fixed crosslinked network. Atomistic DGEBF–TETA epoxy models with a crosslinking degree of 35% were constructed with BNNS loadings of 12, 15, 18, and 21 wt.%. Thermal conductivity was calculated using non-equilibrium molecular dynamics, while density, mesh-derived void fraction, elastic moduli, volumetric thermal expansion, and vibrational density of states were analyzed to connect molecular structure with macroscopic properties. The average thermal conductivity increased monotonically with BNNS loading, reaching 0.53 W/(m·K) at 21 wt.%, approximately 119% higher than neat epoxy. This improvement cannot be explained by density alone. Although composite density increased with BNNS content, surface mesh analysis showed that BNNS incorporation also increased the void fraction relative to neat epoxy, indicating that global densification coexists with local disruption of polymer packing. Spatial analysis further suggested that higher BNNS loadings reduce nanosheet separation, providing qualitative structural information for interpreting the thermal conductivity trend. VDOS analysis showed that BNNS suppresses low-frequency collective motions (0–5 THz) while enhancing intermediate-frequency vibrational modes (5–20 THz), which qualitatively accompanies the loading-dependent thermal conductivity enhancement. Thermomechanical calculations showed an overall increase in Young’s modulus and a reduction in volumetric thermal expansion, whereas shear modulus was less sensitive to BNNS loading. These results indicate that BNNS/epoxy performance is associated with coupled changes in density, local packing disorder, nanosheet distribution, and vibrational response, providing molecular-level guidance for designing thermally conductive and electrically insulating epoxy composites.