|By Tim Negris||
|June 27, 2010 09:45 AM EDT||
What do the Mayan calendar and silicon-based semiconductor technology have in common? They both end in 2012. That's the year when the laws of physics will likely trump Moore's Law and it simply won't be possible to cram any more transistors into the integrated circuits on a silicon chip without current leaks ruining its ability to process digital information reliably.
In the scrum for a viable alternative to silicon, the smart money for the last few years has been on graphene, a one-atom-thick sheet of carbon with the fantastical attributes of being flexible, transparent, recyclable, many times stronger than steel and, most important, extraordinarily conductive at room temperature. But until very recently, when it came to using graphene to build circuits, there was a huge catch. It had to be cut into microscopic "nano-ribbons" from which the circuits could then be assembled. If such a process sounds clunky and expensive, that's because it is.
But, a brilliantly simple discovery by Paul Sheehan of the Naval Research Laboratory, Elisa Riedo at Georgia Institute of Technology and their colleagues, published in the current issue of the journal Science, paints a much prettier picture of how graphene-based semiconductors might be built in the near future.
It's been known for the last few years that when graphene oxide, an electrical insulator, is heated, the oxygen atoms are removed and it's converted to pure graphene, an unprecedented electrical conductor. A little heat makes it somewhat conductive; a lot of heat makes it very conductive. The researchers developed a way to exploit this property of graphene by using a 12nm-wide stylus borrowed from an atomic force microscope and heating it to temperatures between 150 and 1060 degrees Celsius while dragging it across a sheet of graphene oxide. The result is a conductive path, or nano-wire of pure graphene "drawn" into the insulating surface of the graphene oxide. With the alterations in temperature, they were able to vary the nano-wire's conductivity by four orders of magnitude. Over the course of the experiment they also found that the stylus tip didn't wear out and that the graphene oxide sheet wasn't inclined to tear.
Using this technique, it'll be possible, in a fast, inexpensive, single-step process, to "draw" complex, compact, extremely energy-efficient circuits on material the thickness of a single carbon atom. That's a dramatic improvement over the method currently used to produce silicon-based chips, which involves many, many steps using a huge array of expensive machinery and chemical agents, consumes vast amounts of water and electricity, and produces a variety of toxic wastes and byproducts. On top of which the devices built from those chips consume large amounts of electricity while they're being used and produce mountains of toxic e-waste thereafter.
By comparison, graphene-based computing promises almost unimaginable things, including a simple, inexpensive, efficient manufacturing process open to even small companies, where the principal byproduct is oxygen and the product is a supercomputer the size of a credit card, a television the thickness of a sheet of paper, or a cell phone that runs for months on a single charge, all made largely of fully recyclable material.
It will also mean that Gordon Moore will have to rewrite his law with a steeper curve.