Unveiling the thermal transport mechanism in compressed plastic crystals assisted by deep potential

The unique properties of plastic crystals highlight their potential for use in solid-state refrigeration. However, their practical applications are limited by thermal hysteresis due to low thermal conductivity. In this study, the effect of compressive strain on the thermal transport properties of plastic crystal [(CH3)4N][FeCl4] was investigated using molecular dynamic simulation with a deep neural network potential. It is found that a 9% strain along [001] direction enhances thermal conductivity sixfold. The underlying mechanisms are analyzed through vibrational density of states, spectral energy densities, and mean square displacements. The enhancement in thermal conductivity is primarily due to increased group velocity and reduced phonon scattering, driven by volume compression within the 0-1 THz. These findings offer theoretical insights for the practical application of plastic crystals in thermal management systems.