Graphene: The New Silicon

What is graphene?

Put simply, graphene is the future of silicon and since graphene isn’t silicon at all, it’s going to be curtains for silicon - or at least that’s how it looks right now. Moore’s Law, as I’ve noted several times in other postings, is currently broken, mainly because it’s not possible to make the circuits on silicon chips much smaller and faster. Which in turn is why the chip vendors now are selling multicore chips, so they can claim further increases in speed. However there’s a limit to how far that road can stretch.

Just about everyone in the nanotech world is hoping that graphene will step in and fix the problem and it’s beginning to look like it might. At a molecular level, graphene is one big sheet of chicken-wire built from carbon atoms, as illustrated in the diagram. This means that it is one atom thick and (in theory) as long and broad as you please. Graphite, which puts lead in your pencil, is made of graphene. When you write on paper you mark it by transferring layers of graphene to the paper. And as graphite occurs naturally, the raw material is cheap.

The neat thing about graphene is that it conducts electricity at room temperature better than any other known material. The electrons travel up to 100 times faster through graphene than silicon and it offers low resistance, 35% less than copper. But there is a catch. If it is going to be used for chips then raw graphene needs to have its impurities removed.

However, collaborative work by a number of heavyweight researchers (Robert K. Prud’homme, Ilhan A. Aksay, Roberto Car, Je-Luen Li and Konstantin Kudin - mainly from Princeton) has established a way of refining graphene and now there’s a commercial company, Vorbeck Materials, which provides it.

This is just the beginning. The chemical process invented by Aksay and Prud’homme extracts individual sheets or ribbons of graphene a few nanometers high and 200 to 500 nanometers wide. What now needs to be done is to use these to create high quality wafers and then etch circuits on them, using nanolithography, to create chips. This has not yet been demonstrated, but the consensus view is that it’s only a matter of time.

The potential is to have chips that give out much less heat and have the potential to work up to 100 time faster. That would reestablish Moore’s Law and keep it going for at least another decade. Naturally all the chip vendors, IBM, Intel, AMD et al, are experimenting with graphene, and a contact I have in Intel recently advised me to “watch this space”.