Slashdot Mirror
Replacing Copper With Pencil Graphite
Late-Eight writes "A key discovery at Rensselaer Polytechnic Institute could help advance the role of graphene as a possible heir to copper and silicon in nanoelectronics. Researchers believe graphene's extremely efficient conductive properties can be exploited for use in nanoelectronics. Graphene, a one-atom-thick sheet of carbon, eluded scientists for years but was finally made in the laboratory in 2004 with the help of everyday, store-bought transparent tape. The current research, which shows a way to control the conductivity of graphene, is an important first step towards mass producing metallic graphene that could one day replace copper as the primary interconnect material on nearly all computer chips." Researchers are now hot to pursue graphene for this purpose over the previous favorite candidate, buckytubes (which are just rolled-up graphene). Farther down the road, semiconducting graphene might take over from silicon at the heart of logic chips.
The Pencil trick is also useful for reconnecting the bridges on the original Duron/Athlon chips.
Pushing the cpu up from 650 to 800mhz made *all* the difference...
liqbase
Bah, Physicists and their QM simulations! They got it all wrong again. It isn't the length of the graphene ribbon that affects its properties, but the shape of its edges. If you look at benzene ring's molecular orbitals, you'll see that there are two ways to pack them in a ribbon. If they all line up, with resonant transfer going along the ribbon in a straight line, then you have metallic conductivity, with the electron just gliding across all the orbitals without hitting any gaps. If the orbitals don't line up, you end up with little dead ends here and there, which cause "turbulence" and reduce conductivity.
Now, the packing of the orbitals is determined by the edges because of their constraints on orbital orientation. In the middle of the ribbon, you have a pure hex grid, and the orbitals, which can be visualized as taking half of each hex and painting a large C on it (these are not the same as the three bonding pi orbitals). Try it yourself: draw a hex grid and try to pack Cs. To visualize resonance, push on one end of a C and see how to repack the resulting structure. In the middle, you have three orientations at every node, but at the edges you don't. The more edges you have, the more constraints there are on the packing, and the more likely it is that the oribitals in the middle won't be in resonance with each other in a given direction. When you push on a C in such a grid, it will push other Cs sideways instead of along the ribbon, causing "resistance".
There are two types of edges, familiar to tile game developers as the vertical and horizontal orientation. In the horizontal packing, the flat side of each hex is bordering the edge, in the vertical the flat side is perpendicular to the edge. It turns out that if you have horizontal edges on your graphene ribbon, it is metallic; if you have vertical ones, it is semiconductive (which is another way of saying it has more resistance). If the edges are not quite straight, which will quite likely happen if you are making your ribbons via CVD or duct tape or something, you'll see a mix of both behaviors, resulting in a conductivity somewhere in between full-out and almost-nothing.
This is the trouble with modern physics - they just don't care about reality any more. If they only drew a few pictures, like real chemists do, they'd have seen this very easily. Instead they waste their time on simulations that only give them numbers they don't know how to interpret. Sheesh.