The feat marks a milestone in efforts to lay the foundations of the next generation of computers when conventional silicon based microchip technology runs out of steam.
A team in Manchester last year announced that it had created transistors that measured 50 atoms across. Now they have slashed the size of the transistors to just 10 atoms, marking the first true electronic nanocomponent, where a nanometre is one billionth of a metre, and a single human hair is 100,000 nanometres across.
The University of Manchester team led by Prof Andre Geim has been fashioning the transistors from the world’s thinnest material, called graphene, consisting of carbon atoms a single layer thick, arranged in a hexagonal pattern like that seen in chicken wire.
Working with Dr Kostya Novoselov, he believes that the world’s smallest transistor, described in the journal Science, could spark the development of super-fast computer chips.
In recent decades, manufacturers have crammed more components on to microchips, with the number of transistors per unit area doubling every two years. This has become known as Moore’s Law.
But the speed of cramming is now noticeably decreasing, and further miniaturisation of electronics is to experience its most fundamental challenge in the next 10 to 20 years, according to the semiconductor industry road map. The problem is that at the nanoscale, materials like silicon react with oxygen, changing their properties, moving (“like water droplets on a hot plate”, as Prof Geim puts it) and decomposing.
Graphene has unusual electrical properties and behaves as if the electrical current is not carried by normal electrons but by charged particles with no mass at all. Graphene brings scientists close to making so called ballistic transistors – ultimately faster than any current technology because electrons shoot through them without colliding with component atoms.
Transistors made of graphene start showing advantages at sizes below 10 nanometres – the miniaturisation limit at which traditional silicon based technology is predicted to fail.
Prof Geim does not expect graphene-based circuits to come of age before 2025 but argues this technology will probably be the only viable way to shrink microelectronics after the silicon era comes to an end. “It is too early to promise graphene supercomputers,” he says.
“In our work, we relied on chance when making such small transistors. Unfortunately, no existing technology allows the cutting of materials with true nanometre precision. But this is exactly the same challenge that all post-silicon electronics has to face. At least we now have a material that can meet such a challenge.”
“Graphene is an exciting new material with unusual properties that are promising for nanoelectronics”, comments Prof Bob Westervelt, of Harvard University. “The future should be very interesting”. – tel