Method for making single-atom-thick, wafer-scale materials opens up opportunities in flexible electronics

2 mins read

A technique to harvest 2-inch diameter wafers of 2D material within just a few minutes which can then be stacked together for form an electronic device within an hour, has been developed by MIT.

Since the 2003 discovery of the single-atom-thick carbon material known as graphene, there has been significant interest in other types of 2D materials. These materials could be stacked together like Lego bricks to form a range of devices with different functions, including operating as semiconductors. In this way, they could be used to create ultra-thin, flexible, transparent and wearable electronic devices. However, separating a bulk crystal material into 2D flakes for use in electronics has proven difficult to do on a commercial scale.

The existing process, in which individual flakes are split off from the bulk crystals by repeatedly stamping the crystals onto an adhesive tape, is unreliable and time-consuming, requiring many hours to harvest enough material and form a device.

The MIT Department of Mechanical Engineering researchers’ technique could open up the possibility of commercialising electronic devices based on a variety of 2D materials, according to to Associate Professor Jeehwan Kim, who led the research.

"We have shown that we can do monolayer-by-monolayer isolation of 2D materials at the wafer scale," Assoc Prof. Kim says. "Secondly, we have demonstrated a way to easily stack up these wafer-scale monolayers of 2D material."

The researchers first grew a thick stack of 2D material on top of a sapphire wafer. They then applied a 600-nanometer-thick nickel film to the top of the stack.

Since 2D materials adhere much more strongly to nickel than to sapphire, lifting off this film allowed the researchers to separate the entire stack from the wafer. What's more, the adhesion between the nickel and the individual layers of 2D material is also greater than that between each of the layers themselves.

As a result, when a second nickel film was then added to the bottom of the stack, the researchers were able to peel off individual, single-atom thick monolayers of 2D material. That is because peeling off the first nickel film generates cracks in the material that propagate right through to the bottom of the stack, Assoc Prof. Kim says.

Once the first monolayer collected by the nickel film has been transferred to a substrate, the process can be repeated for each layer. "We use very simple mechanics, and by using this controlled crack propagation concept we are able to isolate monolayer 2D material at the wafer scale," he says.

The universal technique can be used with a range of different 2D materials, including hexagonal boron nitride, tungsten disulfide, and molybdenum disulfide.

In this way, it can be used to produce different types of monolayer 2D materials, such as semiconductors, metals, and insulators, which can then be stacked together to form the 2D heterostructures needed for an electronic device.

"If you fabricate electronic and photonic devices using 2D materials, the devices will be just a few monolayers thick," Assoc Prof. Kim says. "They will be extremely flexible, and can be stamped on to anything," he says. The process is fast and low-cost, making it suitable for commercial operations, he adds.

The researchers are now planning to apply the technique to develop a range of electronic devices, including a nonvolatile memory array and flexible devices that can be worn on the skin.

They are also interested in applying the technique to develop devices for use in the "internet of things," Assoc Prof. Kim says. "All you need to do is grow these thick 2D materials, then isolate them in monolayers and stack them up. So it is extremely cheap - much cheaper than the existing semiconductor process. This means it will bring laboratory-level 2D materials into manufacturing for commercialisation. That makes it perfect for IoT networks, because if you were to use conventional semiconductors for the sensing systems it would be expensive."