Graphene is a sheet of carbon just one atom thick, with the atoms arranged in a hexagonal grid, like chicken wire. As well as being thin, super-strong and very flexible, this structure allows electrons to zip through at high speeds. It therefore seems poised to become the basis for ultra-fast computers.
But there’s a problem. Materials used to make transistors have to be able to switch current on and off to create logic circuits. This switchability depends on there being a difference, known as a band gap, between the energy possessed by the material’s free electrons, and the energy they would need to move around and conduct electricity. Only applying the right amount of external energy allows electrons to jump the band gap to become conducting.
Semiconductors like silicon have band gaps, making for good transistors. But because of its natural conductivity, graphene has a band gap of zero – making it unsuitable for transistors.
Various solutions have been proposed for turning graphene into a semiconductor, including adding an insulating layer between two layers of graphene to reduce its conductivity, or carving it into very narrow ribbons, which alters its structure, making it easier to turn current off.
However, both of these use conventional etching techniques, placing limits on just how small such graphene transistors could get. For example, creating ribbons any less than 10 nanometres wide has until now left them with edges so ragged that the band gap disappears.
Now Ed Conrad at the Georgia Institute of Technology in Atlanta and colleagues have found a way to create much narrower nanoribbons, without destroying their semiconductor capabilities, in a much simpler way.
The trick was to grow graphene sheets on a rippling surface covered in parallel trenches, each 18 nanometres deep. The team found that where the surface of the sheet dipped into a trench, the graphene’s properties became semiconducting, with a band gap of 0.5 electronvolts, rather like the nanoribbons. Crucially, however, these strips of semiconductor were just 1.5 nanometres wide and sleek, not ragged.
“The results are very important, because the existence of a band gap makes it possible to use graphene in nanoelectronics,” says Luis Brey at the Materials Science Institute in Madrid, Spain. But the reason the gap exists is not yet clear. “More theoretical work is needed in order to understand how a 0.5 electronvolt gap appears in this structure,” he says.
Chris Howard at University College London calls the work a strong development that opens up many potential applications, including high-performance electronics.
Silicon won’t be replaced just yet, adds Conrad. “But graphene could push Moore’s law to its limits.”
Journal reference: Nature Physics, DOI: 10.1038/nphys2487
Syndicated content: Douglas Heaven, New Scientist