"For any kind of device it is very important to get low-resistance contact, especially for field-effect transistors, or you lose a lot of power as heat," says Aditya Mohite of Los Alamos National Laboratory in the US, one of the scientists leading the team behind this latest work. "Contacts can be really problematic – that’s why carbon nanotube electronics hasn’t yet achieved what people expected it to. So it’s important that we achieved this low-resistance contact with MoS2."

The researchers modified the MoS2 at the site of the contact to a metallic phase, which allowed good contact with any metal used. They demonstrated a difference in the current for on and off states of six orders of magnitude for their "phase-engineering approach" transistor device. As Mohite points out, this significantly exceeds the requirements of a typical laptop, where a difference of just four orders of magnitude would suffice.

More than just a phase

The idea of modifying material properties at the contact to reduce resistance is not new, Mohite tells nanotechweb.org. "People had tried local doping but it was not very successful because the properties of the doped material were liable to change over time."

He describes how they had been looking at the phase conversion in MoS2 at the bulk level and noticed enhanced chemical activity, which suggested a high electron density. The idea followed that this metallic phase could be used to reduce the resistance between metals and MoS2 in contacts. The metal phase allows good electron injection whatever metal is then used at the contact, and because the energy levels of the metallic phase and semiconductor phase of the MoS2 are very well aligned, there is a low-resistance connection to the semiconductor channel as well.

Production challenges

Mohite collaborated with colleagues at Los Alamos National University and with Manish Chhowalla and his team at Rutgers University. The researchers used a process called lithiation to modify the MoS2 to the metallic phase, by exposing the material to n-butyl lithium.

"In the bulk phase this is straightforward but to get lines around a micron wide you need to get the right concentration and lithiation time – if you do it too long you get diffusion into the semiconducting channel," says Mohite.

The researchers also experimented with MoS2 produced in different ways. While 60–70% phase conversion was achieved on flakes made by exfoliation, they managed just 10–20% phase conversion on flakes created by chemical vapour deposition.

Further work is needed to investigate the stability of the devices. The metallic state of MoS2 is metastable and reverts to the semiconductor phase at temperatures above 150 °C. All of the devices tested proved stable in ambient conditions, which may be testament to the quality of the contacts. "The contact is low resistance," points out Mohite. "If it was high resistance there may be more heating.

Next the team plans to look at the optical properties of MoS2 and how they might be exploited for optoelectronic devices, using this phase-engineering approach to probe intrinsic parameters of the material in isolation from the contacts.

Full details are reported in Nature Materials.