Scholarlink

<p>Metal-Organic Chemical Vapor Deposition (MOCVD) is critical for high-quality thin film fabrication in semiconductors, yet film uniformity remains challenging due to complex interactions between reactor design and process parameters. While previous studies focused on two-dimensional models or small reactors, comprehensive three-dimensional (3D) analyses of horizontal MOCVD (H-MOCVD) reactors are limited. We present a systematic optimization of design variables for uniform GaN deposition using a COMSOL-based automated Computational Fluid Dynamics (CFD) model combined with Design of Experiments and Response Surface Methodology. Our approach identified optimal parameters that simultaneously achieve uniform concentration distribution and earlier plug flow formation under non-rotating susceptor conditions. The simulations revealed that larger mixing regions promote earlier plug flow, providing quantitative insights into design parameter effects. This methodology offers fundamental understanding of MOCVD reactor optimization applicable to industrial processes and experimental efficiency improvement.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Numerical analysis and optimization of horizontal MOCVD reactor design parameters for uniform GaN thin films

  • 期刊论文

  • 发表日期：

    2026年3月31日

查看全文
  • Haneul Kim,
  • Roy Byung Kyu Chung,
  • Hyunseok Ko

摘要

Metal-Organic Chemical Vapor Deposition (MOCVD) is critical for high-quality thin film fabrication in semiconductors, yet film uniformity remains challenging due to complex interactions between reactor design and process parameters. While previous studies focused on two-dimensional models or small reactors, comprehensive three-dimensional (3D) analyses of horizontal MOCVD (H-MOCVD) reactors are limited. We present a systematic optimization of design variables for uniform GaN deposition using a COMSOL-based automated Computational Fluid Dynamics (CFD) model combined with Design of Experiments and Response Surface Methodology. Our approach identified optimal parameters that simultaneously achieve uniform concentration distribution and earlier plug flow formation under non-rotating susceptor conditions. The simulations revealed that larger mixing regions promote earlier plug flow, providing quantitative insights into design parameter effects. This methodology offers fundamental understanding of MOCVD reactor optimization applicable to industrial processes and experimental efficiency improvement.