Date of Award
Open Access Thesis
As semiconductor electronics, in particular high electron mobility transistors, get hot in operation, people are trying to incorporate different semiconductor materials to solve the problem of heat dissipation to maintain the device efficiency and performance. In this master thesis, two different packages, LAMMPS and almaBTE, that are focusing on micro-scale and atomic scale respectively, are used to simulate the heat transfer process across interfaces in two different scales. In the LAMMPS simulation, the research is exploring how the location of atoms and the thickness of interfacial layers will influence the interfacial thermal conductance. On the other hand, using the almaBTE package, the simulation is focused on how different combinations of materials result in different interfacial thermal conductance. In addition, the phonon density of states mismatch model is applied to search new materials with high heat flux. The CGCNN machine learning technique is used for training the model of interfacial thermal conductance, and finally the trained model is used to screen new materials with higher performance of interfacial thermal management. This thesis work demonstrates the power of machine learning models and paves the way for high-throughput screening of novel crystalline materials with desirable interfacial thermal conductance for phonon-mediated thermal management of wide bandgap electronics.
Chen, H.(2022). High-Throughput Computation of Interfacial Thermal Management of Wide-Bandgap Semiconductors. (Master's thesis). Retrieved from https://scholarcommons.sc.edu/etd/7075