Plasma-Based Integrated System for Synthesis, Etching and Machine Learning of 2D TMDc

The research presented in this work represents a culmination of over a decade of dedicated efforts toward the development and implementation of a plasma-based integrated system for the synthesis, etching, and machine learning of two-dimensional (2D) transition metal dichalcogenides (TMDc). Our journey began in 2013 with a groundbreaking achievement: the successful synthesis of semiconducting MoS₂ at 300 °C using an inductively coupled plasma (ICP)-plasma enhanced chemical vapor deposition (PECVD) technique (PLASMART, Republic of Korea) on a 4-inch substrate. This milestone was the first step towards realizing the immense potential of 2D TMDc materials in various applications.<br/>Building on this initial success, we continued to advance our capabilities. In 2015, we lowered the synthesis temperature to 150 °C, enabling direct deposition on polyimide substrates. This development not only expanded the range of possible applications but also demonstrated our commitment to environmentally sustainable processes. In 2020, we achieved another significant milestone by successfully synthesizing 2D WS₂, a member of the same TMDc group, in the metallic 1T phase using PECVD.<br/>The year 2021 marked a breakthrough as we overcame previous limitations in heterostructure fabrication by creating a MoS₂-WS₂ heterostructure using H₂S + Ar plasma on a 4-inch hetero-metallic (Mo-W) layer. This achievement unlocked the potential for novel electronic and optoelectronic devices based on these heterostructures. Various forms of TMDc, all consisting of 4-5 layers, were systematically obtained, laying the foundation for further exploration.<br/>By 2023, we had established precise etching conditions at the atomic layer level using ICP-reactive ion etching (RIE) with a carefully designed three-gas mixture (Ar+O₂+CF₄) and optimized RIE utilization. This enabled us to tailor the properties of 2D TMDc materials with unprecedented precision. Furthermore, we employed density functional theory (DFT) calculations and optical emission spectroscopy (OES) diagnostic results to gain deeper insights into the plasma reaction mechanisms, shedding light on the underlying physical processes.<br/>In a remarkable culmination of our efforts, we harnessed the power of machine learning (ML) to predict process outcomes in 2023, revolutionizing the way we approach material synthesis and etching. This achievement represents a significant leap forward in terms of efficiency and precision, with implications spanning various industries and research domains.<br/>In summary, our research journey has led to the development of a comprehensive and versatile plasma-based integrated system for 2D TMDc materials. Over the course of a decade, we have pushed the boundaries of what is possible in material synthesis, etching, and process optimization, ultimately contributing to the advancement of science and technology.