NTUEE

::: Home / Research Areas / Research Spotlight

Incorporating Bismuth (Bi) Electrodes

Research and development of miniaturized devices has long been a goal pursued by many scientists. Since Intel co-founder Gordon Moore proposed Moore’s Law in 1965, predicting that the number of transistors on a chip would double every 24 months, this principle has guided the industry’s advancement toward ever-shrinking chip sizes and decreasing 3C product costs. However, after three decades of silicon-based transistors, continuous scaling has led to severe challenges, such as the short-channel effect, increased contact resistance, and reduced carrier mobility. Various solutions have emerged in response, including multi-gate structures, planar fully depleted devices, and source/drain metal silicides. Yet, as multi-gate fabrication becomes increasingly complex, overly thin fully depleted devices suffer from interface scattering that compromises carrier mobility, and the contact resistance reduction achievable through metal silicides has reached its limit. As a result, the overall performance of silicon-based transistors is nearing its physical boundary. In the post-2 nm era, two promising device architectures remain: gate-all-around (GAA) transistors, and transistors that replace silicon with alternative channel materials.

Consequently, scientists began searching for substitute materials for device channel layers. Compared with traditional bulk semiconductors, transition metal dichalcogenides (TMDs)—such as MoS₂, WSe₂, and PtSe₂—offer stronger confinement of channel charges controlled by gate voltage in atomically thin films with reduced surface scattering. Even after scaling, these materials retain superior short-channel immunity and exhibit low-defect van der Waals interfaces. Moreover, as thickness decreases, their carrier mobility does not sharply drop compared to silicon. Due to their excellent conductive, semiconductive, and insulating properties, various two-dimensional (2D) materials have attracted significant attention in both academia and industry, and are regarded as candidates for extending the mission of silicon materials that have reached their physical bottleneck.

Incorporating Bismuth (Bi) Electrodes: A Breakthrough in Current Deficiency

A collaborative team led by Professor Chih-I Wu at NTU, together with TSMC and MIT, demonstrated that using semimetal bismuth (Bi) electrodes with 2D materials dramatically reduces resistance and enhances current. This discovery is crucial because it addresses the inherent current limitations of 2D materials, enabling performance and energy efficiency nearly on par with silicon, while also unleashing the full scaling potential of 2D materials for realizing sub-1 nm atomic-scale transistors. This international collaboration led to the 2021 publication in Nature titled “Ultralow contact resistance between semimetal and monolayer semiconductors”. The paper, for the first time, showed that using Bi as the contact electrode for 2D materials achieved a record-low resistance in the field, underscoring the potential of 2D materials for future atomic-scale transistor technologies. This groundbreaking outcome was both unexpected and remarkable: Bi’s semimetallic properties drastically reduce the energy barrier between metal and semiconductor, while preserving the atomic integrity of 2D materials during deposition. Applying this method to several 2D semiconductors—including MoS₂, WS₂, and WSe₂—produced excellent electrical characteristics.

Showcasing NTU’s Academic Strength and Sustaining Next-Generation Industrial Competitiveness

Beyond the discovery of Bi electrodes, Professor Wu’s laboratory further employed helium-ion beam lithography to shrink device channels to nanometer dimensions. Professor Wu noted that this latest breakthrough raises the transistor performance of monolayer 2D materials to levels comparable to silicon semiconductors, while maintaining compatibility with mainstream silicon processes. Not only does this help push beyond the limits of Moore’s Law and open up new possibilities for future technologies like 5G and AI, it also carries two major implications. First, in terms of academic competitiveness, Taiwan’s unique technological education and complete industrial ecosystem give it an edge over other countries in advanced R&D. Second, in terms of industrial competitiveness, research outcomes such as this, once they achieve commercial breakthroughs, will help Taiwan’s semiconductor and tech supply chain maintain its global leadership.

This success demonstrates how the project established a robust model of academia-industry collaboration, bridging basic and applied sciences. Co-first authors Dr. Pin-Chun Shen and Dr. Ang-Sheng Chou remarked that the opportunity for NTU to collaborate with MIT was made possible by the Industry-Academia Collaboration Alliance supported by TSMC and the Ministry of Science and Technology. Because of time zone differences, every email and phone discussion was time-critical, but the chance to exchange ideas with world-leading scholars was an invaluable academic experience. Interestingly, both Shen and Chou were once master’s students at NTU’s Graduate Institute of Photonics and Optoelectronics—one pursued further studies abroad while the other remained in Taiwan—yet both eventually joined TSMC years later to continue advancing semiconductor research, a remarkable twist of fate.