Thermal conductance of nanostructured interfaces from Monte Carlo simulations with ab initio -based phonon properties

Abstract

Interfaces are ubiquitous in electronics, photonics, and advanced materials. Interface engineering has become an essential strategy for developing functional materials with low thermal conductivities such as thermoelectric materials and thermal barrier coatings. On the other hand, interfaces are becoming a bottleneck for thermal management in electronic devices. Recent experiments have shown that a fin-like nanostructured interface with a size of 30-100 nm could enhance thermal transport across interfaces. Since phonon mean free paths span from several nanometers to dozens of micrometers, depending on the material, the size of the interface features may significantly affect the phonon transport regime and interface conductance. Here, the Monte Carlo simulation, with ab initio-based phonon properties as input parameters, was developed to study thermal conductance of a fin-like nanostructured interface. Simulated results indicate that the nanofin size (i.e., width, spacing, and height) significantly affects interface thermal conductance. Interface conductance is found to first increase and then decrease with increasing width of the nanofin when its height is 100 nm. This phenomenon is attributed to competition between the enlarged interface area and increased backscattering of transmitted phonons. This study demonstrates the existence of an optimal nanofin size for maximizing interface conductance, which could be important for thermal management of high-power electronics using nanostructured interfaces.