Zhihui Cheng1 Hattan Abuzaid1 Shreya Singh1 Yifei Yu2 Linyou Cao2 Aaron Franklin1 3

1, Department of Electrical and Computer Engineering, Duke University, Durham, North Carolina, United States
2, Department of Materials Science and Engineering, North Carolina State University, Raleigh, North Carolina, United States
3, Department of Chemistry, Duke University, Durham, North Carolina, United States

Atomically thin 2D crystals are promising channel materials for extremely scaled field-effect transistors (FETs). However, the challenge of contacting 2D materials, especially at the scaled contact lengths (Lc < 30 nm) required for future technologies, constitutes a major roadblock for realizing their full potential. Two mysteries have emerged from studying contact length scaling behavior in 2D FETs: the impact of 2D material thickness and contact gating. It is unclear how the thickness of 2D materials impacts the transfer length (length over which the majority of carriers are injected at the metal-semiconductor contact). There are many incongruent claims around this mystery, from both experimental and theoretical studies. For example, some theoretical studies claim the transfer length is ~1 nm for monolayer MoS2 FETs, whereas some experimental studies demonstrate a transfer length ranging from 30 nm to 100 nm. The second mystery is the influence of contact gating on contact scaling. Most 2D FETs demonstrated thus far use a back-gate configuration, which allows for electrostatic modulation of the metal-2D as the channel is gated, creating a contact-gating effect. Contact gating could induce stronger carrier injection, but conclusive evidence for its actual impact on the transfer length remains unknown. Unraveling these two mysteries is pivotal in order to understand how carriers are transported in scaled metal-2D contacts.

In order to investigate these mysteries, we fabricated 2D devices having identical channel length, but with different contact lengths (from 15 nm to 100 nm). The channel material is CVD-grown MoS2 with the thickness ranging from 1 to 4 layers, allowing us to study the effects of 2D crystal thickness based on monolayer increments. For this range of MoS2 thicknesses, we compared the devices in top and bottom gate configurations, to understand the impacts of contact gating on the transfer length and the contact scaling behavior. Furthermore, we benchmarked the contact scaling behavior of these various 2D FET configurations against the theoretical and experimental observations in the literature, providing a holistic picture of carrier transport at these metal-2D interfaces.