Multiscale Modeling of Defect Generation and Diffusion in Semiconductor Materials
MetadataShow full item record
As semiconductor device technology continues to evolve, great challenges arise in many areas. With the continuous shrinkage of the design window, a better understanding of the detailed physical processes that occur during fabrication is required. This dissertation utilizes the multiscale modeling approach to investigate such processes, focusing on the origin of these processes at the atomic scale and the collective behaviors over the large device scale. In order to gain a fundamental understanding of the diffusion processes that occur during device fabrication, we have developed a kinetic lattice Monte Carlo (KLMC) simulator capable of simulating the self-/inter-/impurity diffusion processes. The findings demonstrate the great potential of KLMC in full scale simulation of strain and composition dependent diffusion processes in strained silicon and silicon-germanium devices. The silicon photovoltaics industry is undergoing a continuous drive of cost reduction and efficiency improvement. The lack of fundamental understanding of the emitter deposition process poses challenges on process optimization. We have developed continuum models incorporating important physical processes during deposition, such as growth of phosphosilicate glass and transport of phosphorus across the glass layer, and diffusion, deactivation and immobilization of phosphorus in silicon. We have simulated the diffused emitter profiles that agree well with experiments. The full modeling of the emitter deposition process allows process optimization to enhance efficiency and reduce cost. Metals are detrimental defects in devices as they introduce traps that limit carrier lifetime. Modern devices require gettering processes to remove metals from active device region. We have built a model for the phosphorus diffusion gettering behavior of various metal species. Calculation results are in agreement with experimental gettering behavior and provide guidance on optimizing gettering efficiency for better device performance. Lastly, we have extended our approach to wide band gap semiconductor β-Ga<sub>2</sub>O<sub>3</sub>, a potential candidate for future electronics devices. We have carried out <i>ab initio</i> calculations on the intrinsic vacancy and transition metal impurity defects in β-Ga<sub>2</sub>O<sub>3</sub> and the results are in good agreement with experimental data. The theoretical analysis, combined with experimental observations, contributes to the fundamental understanding of nano-processes and electrical properties of β-Ga<sub>2</sub>O<sub>3</sub>.
- Electrical engineering