Molecular dynamics simulation of thermal conduction across mechanical interfaces with sub-nm roughness

Ben Gao,Ziwen Zou,Menglin Li,Menglong Hao
DOI: https://doi.org/10.1016/j.icheatmasstransfer.2024.107622
IF: 6.782
2024-05-26
International Communications in Heat and Mass Transfer
Abstract:Mechanical interfaces are ubiquitous in devices like the micro-electromechanical system. The surface morphology plays a crucial role in thermal transport across interfaces. However, the mechanism that impacts thermal contact conductance (TCC) remains poorly understood. Molecular dynamics simulations are employed to investigate thermal transport across randomly rough mechanically contacted Si/Si interfaces. Given the random characteristics of rough surfaces, at least 20 contact structures are constructed for each root-mean-square (RMS) roughness. Furthermore, the simulation for each structure is conducted 5 times with distinct random seeds. Correlations between TCC and related factors, such as RMS, contact area ratio, the number of interacting atoms, and atomic pairs, are analyzed. Statistical results indicate a maximum variation of about 21 times in TCC, with a fixed RMS of 5 Å. This contrasts with the commonly held belief that TCC is typically determined by RMS. Interestingly, TCC exhibits a strong linear correlation with the number of interacting atomic pairs per unit area across interfaces regardless of RMS. The phonon density of states (PDOS) at interfaces is further investigated. This work elucidates the underlying mechanism that dictates TCC across mechanical interfaces with sub-nm roughness and provides insight into potential ways to optimize these interfaces for thermal purposes.
mechanics,thermodynamics
What problem does this paper attempt to address?