Skip to main content

Liquid metal deposition process developed for wafer-scale 2D semiconductors

By Tim Palucka May 8, 2017
Schematic showing deposition of patterned, two-dimensional GaS on a substrate. FDTES (1h–1h–2h–2h–perflourodecyltriethoxysilane) creates non-wettable regions on the surface to define the pattern for Ga oxide. Poly(dimethylsiloxane) (PDMS) polymer is used as a squeegee to remove excess liquid Ga metal, leaving behind a Ga oxide layer. Subsequent treatment with HCl vapor to form GaCl3 and S vapor to form GaS completes the procedure. Image courtesy of Kourosh Kalantar-Zadeh, RMIT University, Melbourne, Australia.

Many techniques have been used to deposit wafer-scale, two-dimensional (2D) semiconductors (such as MoS2, MoSe2, and WS2) on a substrate—chemical vapor deposition (CVD), physical vapor deposition (PVD), and exfoliation from a layered bulk material, among others. While successful, these batch processes lack the production speeds of a continuous deposition method, and most require high temperature and/or high vacuum conditions. They also result in the formation of grain boundaries that reduce the quality of the films. Now, researchers at RMIT University in Melbourne, Australia, have developed a room-temperature liquid metal deposition method that continuously deposits and simultaneously patterns a semiconducting metal oxide on almost any substrate over large areas in a very short time. The post-transition metal oxide, Ga oxide in this case, is then converted to GaS at relatively low temperatures via an intermediate step. The technique could lead to high-throughput manufacturing of semiconductor components and major cost savings in the industry.

Kourosh Kalantar-Zadeh, a Distinguished Professor of Engineering at RMIT who specializes in electronics engineering, has been working with liquid metals for years, specifically gallium alloys, which are non-toxic and liquid at room temperature. These metals form a self-limiting oxide across the entire surface in one or several layers when exposed to air. There is no nucleation and growth, as is common for CVD or PVD techniques, so the oxide forms in large, boundary-free sheets, which can extend to wafer-scale. “Nature gives us the perfect two-dimensional material in Ga oxide,” Kalantar-Zadeh says. “It’s liquid at room temperature and basically it can amalgamate many metals into itself. You have a medium you never had before to deposit metal oxides or metal sulfides or other materials perfectly at room temperature.”

Achieving the room-temperature deposition capability required the help of chemist and RMIT Research Fellow Torben Daenake, and the dedicated work of colleague Benjamin Carey. The direct conversion of Ga oxide to GaS takes place at about 900°C. Daeneke knew that converting the Ga oxide to GaCl3 first by reacting it with HCl would allow subsequent sulfurization at a 300°C.  “Now we can suddenly print it on all sorts of materials, like temperature-resistant polymers,” Daeneke says.

In practice, the researchers first patterned the substrate to form areas that were wettable by Ga. Then they applied liquid Ga (melting point 29.8°C) at room temperature to the substrate, which stuck to the wettable patterned areas, and removed the excess Ga using a polymer squeegee. An atomically thin Ga oxide “skin” was left behind due to van der Waal forces. Subsequent treatment at 300°C with HCl formed GaCl3 followed by sulfur vapor to form 2D GaS semiconductor bilayers with a thickness of 1.5 nm. The thin layers of GaS photoluminesce with an optical response that is thousands of times faster than bulk GaS, which can help in switching transistors on and off easier at lower voltages.

“From a chemist’s perspective it’s very exciting because for the last few decades people have been working in solvent environments and ionic liquids, and now we get metallic bonds, so it’s a completely new opportunity here that nobody has actually really looked at,” Daeneke says. “Retrospectively it looks quite obvious that we use these oxide skins on top of metal surfaces to make 2D material. In my opinion it’s a very elegant way of getting around the problem of how we synthesize these 2D semiconductors.”

“The idea of using liquid metal is big,” Kalantar-Zadeh says. “Mercury, which is toxic, was phased out in the 1980s. Now we look at Ga, which is not poisonous, so it brings the whole field back to life, and we’ve got to explore it further. There are many other opportunities for these liquid metals to impact other fields.”

“The proposed method will be a good fit for next-generation electronics on flexible polymer substrates,” says Young Duck Kim, assistant professor in the Department of Physics at Kyung Hee University, Seoul, who was not involved in this research. He cautions that more studies are needed to demonstrate improved material quality and control of doping levels during serial deposition of different types of semiconductors. “After that, this process will allow the development of roll-to-roll printing for 2D, material-based, complex circuit structures, similar to color screen printing with various color inks,” Kim concludes.

Read the article in Nature Communications.