Impact of mass and bond energy difference and interface defects on thermal boundary conductance

The objective of this study is to use molecular dynamics simulation techniques in order to improve the understanding of phonon transport at the interface of dissimilar materials and the impact of different material properties on thermal boundary conductance (TBC). In order to achieve this goal, we investigated the contributions of mass and bond energy difference and interface defects on TBC at the interface of nanostructured materials using non-equilibrium molecular dynamics (NEMD) simulation and phonon wave-packet (PWP) simulation techniques.;NEMD is used to distinguish relative and combined contributions of mass and bond energy difference on TBC. As a result, it is found that the mass has a stronger contribution than the bond energy on lowering the TBC and that the TBC is dependent on the length of interdiffusion region as well as temperature. In addition, evidence of inelastic scattering is observed with interdiffusion regions especially when two materials differ in the bond energy.;A detailed description of phonon interactions at the interface is obtained performing PWP simulations. A frequency dependence of the TBC based on phonon dispersion relation is observed. As it is expected, minimum scattering occurs when there exists only vibrational mismatch at the interface and inelastic scattering is to take place at high frequency region when the bond energy of the two materials is different resulting in the strain at the interface. It is also shown that the level of inelastic scattering is dependent on the length of the interdiffusion region. In addition, the TBC calculated with the results of PWP simulations is compared with that of NEMD simulations as well as theoretical predictions from the acoustic mismatch model and the diffuse mismatch model.;A simple analytical model, which utilizes knowledge of thermal interface resistance and the interface geometry for the prediction of effective thermal conductivity, is developed. This model is generated based on Si-Ge superlattices and embedded nanoparticle and nanowire arrays, and is validated by comparing to existing experimental and computational data. Results shows that the effective thermal conductivity calculated from the analytic model agrees well with that of various systems of different materials and geometries.