Machine-Learning-Assisted Cooling Control for MoS₂ Chemical Vapor Deposition

Chemical vapor deposition (CVD) is a prominent technique for synthesizing high-quality, large-area monolayer MoS₂. Cooling is a crucial post-growth stage in most MoS₂ CVD systems. Despite its significant influence on the material's crystallinity, morphology, and properties, it has garnered limited attention in CVD research.<br/>In our study, we explored the effects of cooling on CVD MoS₂ by incorporating cooling processes both during and after growth. We employed machine-learning-assisted feature identification to rapidly distinguish MoS₂ flakes with varying layers and morphologies in optical microscopy. Additionally, we utilized RAMAN and photoluminescence (PL) spectroscopy to identify layers and crystallinity, scanning electron microscopy (SEM) to examine surface morphology and structures with high solutions, and atomic force microscopy (AFM) to measure the thickness of MoS₂ flakes. Our results demonstrated that we successfully obtained monolayer MoS₂ flakes with diverse hierarchical structures by adjusting the cooling rates. This research endeavors to provide more comprehensive insights into the growth mechanisms of CVD MoS₂.