我們的研究探討了二維（2D）半導體的電子特性，深入探索其基本屬性。我們專注於設計和製造創新性的元件，以在創造高性能的超薄和超短元件為目標，以促進電晶體和計算技術的持續微型化。幾個關鍵的研究領域包括：
改進 p 型二硒化鎢（WSe2）電晶體的性能。根據不同製程方式使WSe2擁有n與p的性質，很適合研究基本的CMOS技術應用在計算方面。我們著重在電性傳輸上的金屬絕緣體接面與研究提高載子遷移率的方式藉此提升元件表現。
p 型WSe2的摻雜做法。通過使用簡單的氧化技術進行電荷轉移摻雜，可以提高電接觸品質（electrical contact quality）並降低電阻。我們利用這些摻雜方式來創建新穎的裝置，包括選擇性區域摻雜和提高金屬-半導體介面。
二維記憶體。在應對人工智慧不斷發展的格局和龐大的處理需求時，我們探索可以模擬人類大腦神經元突觸行為的電子裝置，特別是在“類神經計算”方案中。這涉及替代傳統的計算架構例如von Neumann 結構，其中記憶體和處理器是分開的，解決了處理速度的瓶頸。我們的調查集中在2D材料作為記憶元件並評估氧化物在元件性能中的作用。
Our research investigates the electronic characteristics of two-dimensional (2D) semiconductors, exploring their fundamental properties. We focus on designing and fabricating innovative devices with the goal of creating high performing ultra-thin and ultra-short devices, that facilitate the ongoing miniaturization of transistor and computing technologies to continue. Several key research areas include:
Improved performance of p-type tungsten diselenide (WSe2) transistors. WSe2, a versatile 2D material, exhibits both n and p-type characteristics depending on fabrication methods. This versatility is fundamental for complimentary metal oxide (CMOS) technologies, forming the basis for computing technologies. We focus on understanding transport processes at the semiconductor-dielectric interface and identifying mobility-limiting processes to enhance device performance.
Doping strategies for p-type WSe2. Enhancing the electrical contact quality and reducing resistance can be achieved through charge transfer doping, using straightforward oxidation techniques. We use such doping strategies to create novel devices, including selective area doping and enhancing the metal-semiconductor interface.
Two-dimensional memory. Addressing the evolving landscape of artificial intelligence and the substantial processing requirements, we explore electronic devices that mimic the synaptic behavior of neurons in the human brain, particularly within "neuromorphic computing" schemes. This presents alternatives to traditional computing schemes in von Neumann architectures, where memory and processing elements are separate creating bottlenecks in processing speed. Our investigation centers on 2D materials as memory elements and evaluates the role of oxides in device performance.