With the rapid growth of the semiconductor market, high-quality thin film deposition has emerged as a critical challenge in both industry and academia. Metalorganic chemical vapor deposition (MOCVD)is widely employed in compound semiconductor processe...
With the rapid growth of the semiconductor market, high-quality thin film deposition has emerged as a critical challenge in both industry and academia. Metalorganic chemical vapor deposition (MOCVD)is widely employed in compound semiconductor processes due to its advantages, such as large epitaxial areas, high deposition rates, and precise controllability. However, thin film quality is significantly affected not only by gas flow rates, operating pressure, and temperature but also by the structural design of the reactor. These factors determine the distributions of flow, temperature, and chemical composition in the growth zone, thereby directly influencing film uniformity and yield. Previous studies have primarily focused on two-dimensional models or small-scale reactors, with limited three-dimensional analyses of horizontal MOCVD reactors. Furthermore, simultaneous consideration of multiple design variables often incurs significant computational costs and time requirements.
In this study, we analyzed and optimized design variables for uniform GaN deposition in a horizontal reactor. The effects of key parameters were assessed under non-rotating susceptor conditions, and a COMSOL-based automated CFD model was developed to enhance simulation efficiency. We systematically investigated correlations among design parameters using design of experiments (DoE) and response surface methodology (RSM). The optimization results yielded a parameter set that simultaneously satisfied the requirements for uniform concentration distribution and early plug-flow formation. In addition, further simulations revealed that a larger mixing region promotes earlier plug-flow formation, thereby providing quantitative insights into the impact of design parameters on flow characteristics. This methodology offers a fundamental understanding of MOCVD reactor optimization and can contribute to industrial applications as well as to improving the efficiency of experimental processes.