Atomistic study of the influence of lattice defects on the thermal conductivity of silicon

Abstract

Lattice defects such as vacancies, voids and dislocations are inevitably present in any material of technological interest. In this work, non-equilibrium molecular dynamics simulations are conducted to investigate how the monatomic vacancies and nanovoids influence the lattice thermal conductivity of silicon. The results show a clear non-linear decrease of the thermal conductivity with increasing defect volume fraction. Furthermore, it is found that for a given volume fraction of defects, a random distribution shows a lower lattice thermal conductivity. To develop a fundamental understanding of these observations, the spectral energy densities for all phonon branches obtained from 2D Fourier transformations of the atomic trajectories are analyzed. This yields the mean phonon group velocities and relaxation times, which are the main physical quantities contributing to the lattice thermal conductivity. Our analysis reveals that the phonon relaxation time is the most important parameter for describing the degrading of the thermal transport behavior in the defected structures.