Design Optimization of Gas Distribution System for Large‐Scale Capacity GaAs‐MOCVD

Abstract

In the flexible photovoltaic (PV) industry, increment of the metal‐organic chemical vapor deposition (MOCVD) deposition zone is crucial to reducing the cost of ownership and increasing the power conversion efficiency (PCE). This implies that the larger closed coupled showerhead (CCS) of MOCVD system should be used in the manufacture of flexible gallium arsenide (GaAs) PV thin‐film solar cells. Currently, Aixtron Crius II is the largest commercial CCS reactor for GaAs deposition with 55 × 2 inches circular wafers. As the size of the carrier and showerhead increases, the deposition uniformity and gas flow stability become more difficult to control. To address these issues, previous studies have investigated the effects of geometric and process parameters of MOCVD on the vertical process gas inlet, with either a rotating carrier model represented by Veeco using high‐speed rotating carrier (Turbo Disc), and Aixtron using low‐speed rotating planetary carrier, or a horizontal process gas injector with a rotating carrier. However, these efforts have limited further capacity scale‐up. In this study, numerical simulations and fluid visualization experiments are conducted, based on previous research on thermal uniformity, for design optimization of a gas distribution system in a large square GaAs‐MOCVD reactor. The achieved reactor capacity is 36 × 4 inches square wafers in one process cycle, and the optimal average deposition rate and deposition uniformity of GaAs film are ≈0.430 µm min−1 and 0.93%, respectively.

Investigation of Hydrogen Flux Influence on InGaP Layer and Device Uniformity

“…This study not only lays a solid experimental foundation but also provides theoretical insights for refining the InGaP epitaxial growth process, offering valuable guidance for the fabrication of superior InGaP-based electronic devices. Future research will delve further into the interplay between H 2 flux and other growth parameters, aiming for a more precise and efficient epitaxial growth methodology [32][33][34][35][36][37][38].…”