News

Graphene may challenge silicon after all

Paul van Gerven
Reading time: 3 minutes

Not all researchers have given up on graphene for use in mainstream electronics. A new method to create a semiconducting version of the material opens up new horizons.

Researchers at Georgia Tech and Tianjin University have shown that a high-quality layer of graphene is formed when silicon is evaporated from the surface of a silicon carbide wafer under carefully controlled conditions. The resulting epitaxial graphene features a bandgap of 0.6 eV and electron mobilities ten times greater than silicon.

As soon as graphene was discovered, researchers theorized it could replace silicon in transistors on the basis that electrons zip through it at lightning speeds. The issue is that pure graphene isn’t a semiconductor but a bandgapless semimetal. To be able to switch conductivity on and off, it’s therefore necessary to induce a bandgap. After various strategies failed to do so in a commercially viable manner, the research community moved on to other 2D materials, such as molybdenum disulfide.

But Georgia Tech’s Walter de Heer, who has made a career out of exploring carbon-based materials for semiconductor applications, and his former post-doc Lei Ma, now working at Tianjin, have never given up on graphene. Now that their resolve has led to a breakthrough, the duo envisions “a paradigm shift in electronics,” De Heer told Bits&Chips in an e-mail, adding that they’re well aware of the challenges posed by having to orchestrate such a dramatic turnaround in the industry. “But we’re scientists, not industrialists.”

Incidentally, De Heer grew up in Aruba and is still a Dutch national, despite spending most of his life in the US.

Credit: Georgia Tech

A very, very long time

When silicon carbide (SiC), a semiconductor known for its large bandgap, is heated until silicon atoms (Si) on its surface sublimate, the remaining carbon atoms (C) arrange into the hexagonal graphene structure, with some atoms bonded to the substrate underneath. It’s these bonds that give so-called epigraphene its bandgap.

De Heer and collaborators are the first to produce epigraphene on wafer scale. Their innovation was to grow the monolayer while having a silicon-terminated SiC chip face a carbon-terminated counterpart inside the furnace. When heating this system, atoms are displaced in a well-controlled manner, resulting in well-ordered epigraphene.

“Our motivation for doing graphene electronics has been there for a long time, and the rest was just making it happen. We had to learn how to treat the material, how to make it better and better, and finally how to measure the properties. That took a very, very long time,” De Heer says.

Wright brother

In a commentary published alongside the Nature article, two researchers not involved in the work write that the new material “isn’t intended to replace silicon-based electronics.” They suggest a much more niche application: fabricating logic circuits alongside power circuits, for which SiC is a well-known substrate, and other components.

De Heer confirms he’s not trying to replace silicon but believes his brainchild does have potential in mainstream electronics. “We don’t want to simply replicate silicon electronics by doing the same thing, only better. To me, this is like a Wright brothers moment. Skeptics asked why the world would need flight when it already had fast trains and boats. But they persisted, and it was the beginning of a technology that can take people across oceans.”

Main picture credit: Georgia Tech

Related content