Thermal conductivity of nanocrystalline silicon by direct molecular dynamics simulation

The thermal conductivity simulation of nanocrystalline silicon is conducted on a three-dimensional configuration of nanocrystalline silicon with random grain shape for molecular dynamics simulation. The configuration is formed by the Voronoi tessellation method and the thermal conductivity is calculated by the Green-Kubo method. The effects of random grain distribution, periodic boundary, and the simulation system size are examined. Their effects on the simulation results can be neglected. The conductivity at temperature range from 300 K to 1100 K is obtained. The results indicate that the nanocrystalline thermal conductivity of silicon is far below the bulk single crystal and increases quickly with increasing grain size. The average grain boundary thermal resistance varies from 1.0 × 10⁻⁹ m² KW⁻¹ to 1.16 × 10⁻⁹ m² KW⁻¹. The restrain of the phonon mean free path by the nano-grain boundary is responsible for the sharp decrease in thermal conductivity. The effective phonon mean free path plays an important role in determining the thermal conductivity of nanocrystalline materials.