Molecular dynamics simulation of thermal conductivity of GaN

Abstract

Wurtzite GaN (gallium nitride) is a technologically significant semiconductor material known for its diverse applications in optoelectronics and high-power electronics. Understanding its thermal properties is crucial for optimizing the performance and efficiency of GaN-based devices. This study investigates the thermal conductivity of wurtzite GaN along the [0001] crystallographic direction at 300 K. We employ two computational methods: Non-Equilibrium Molecular Dynamics (NEMD) and Equilibrium Molecular Dynamics (EMD). NEMD involves applying a heat flux/sink to the system and measuring the resulting temperature gradient to determine thermal conductivity. We introduce a novel interpolation method for predicting thermal conductivity and extend our simulations to sizes up to 8.5 micrometers to explore size effects. Results reveal that linear extrapolation of thermal resistivity versus the reciprocal of system length is not valid for GaN. EMD is employed using the Green-Kubo method, which calculates thermal conductivity by analyzing heat flux autocorrelation functions at equilibrium. We compare the results from NEMD, EMD, various analytical models, experiments, and first-principles calculations. Our results reveal that NEMD provides thermal conductivity values approximately 1.3 times higher than those obtained from EMD. This comparative analysis presents the strengths and limitations of each method and provides a thorough understanding of the thermal transport properties of GaN.