Mechanical and thermo-mechanical behavior of h-BNNS-reinforced polyurethane nanocomposites: a molecular dynamics and...

The present research systematically investigates polyurethane (PU)/hexagonal boron nitride nanosheet (h-BNNS) nanocomposites with varying h-BNNS loadings using a molecular dynamics (MD) simulation-driven framework, complemented by experimental validation based on the solvent-casting technique. The work aims to elucidate the fundamental structure-property-performance relationships governing the mechanical and thermo-mechanical behavior of PU in the presence of h-BNNS reinforcement.

Atomistic models of PU were constructed in Materials Studio software, and nanocomposite systems containing 0.3, 0.5, and 0.7 wt.% h-BNNS were developed. Structural equilibration was achieved through energy minimization followed by dynamic simulations using the Forcite module with the DREIDING force field. Elastic, shear, and bulk moduli were determined using the constant strain approach with six independent deformation modes. Temperature-dependent behavior was investigated in the range of 300-600 K with 25 K increments. The simulation results demonstrated a monotonic enhancement in elastic, shear, and bulk moduli with increasing h-BNNS content, with the 0.7 wt.% system exhibiting the highest stiffness and resistance to deformation in ambient conditions. Thermo-mechanical analysis revealed gradual modulus degradation with temperature for all compositions; however, h-BNNS-reinforced systems consistently maintained superior stiffness, indicating enhanced thermal stability and delayed softening. Experimental validation was conducted by fabricating PU/h-BNNS nanocomposites via a solvent-casting method using identical filler concentrations. Experimentally, the elastic modulus increased up to 0.5 wt.% reinforcement of h-BNNS due to improved dispersion and strong interfacial bonding, followed by a reduction at 0.7 wt.% attributed to nanosheet agglomeration and stress concentration effects. In contrast, MD simulations predicted a continuous modulus increase with filler loading and yielded higher absolute values due to idealized assumptions such as defect-free interfaces and uniform dispersion. Despite quantitative discrepancies, both experimental and computational results exhibited strong qualitative agreement, identifying 0.5 wt.% h-BNNS as the optimal reinforcement level.