Chip manufacturers have managed to shrink their products regularly for decades and so doubled the circuitry on silicon wafers roughly every two years. Moore’s law describes this long-term trend, so named because Intel co-founder Gordon E. Moore, described it in his 1965 paper, “Cramming more components onto integrated circuits” (If curious, click here to download that paper).
OK, so here is the problem. As the transistors that make up computer chips have shrunk to sizes approaching the atomic scale, a new challenge looms—extremely tiny chip components tend to resist the flow of an electrical current. Since computer chips use electricity to conduct their business, that’s a major problem. In fact, scientists have found that once circuitry gets below 10 nanometers, its resistivity increases exponentially the smaller it gets. Since semiconductor manufacturers are already producing 22nm chips for release this year and the 10nm node should come online in 2015, does this then imply we are fast approaching the end of the road?
The roadblock is not just the resistivity problem, they face other atomic-scale problems like quantum tunneling.
We now have a potential answer to all this, it appears in Science and comes from researchers at the University of New South Wales, the University of Melbourne, and Purdue University’s Birck Nanotechnology Center. In a study published in the current issue of Science, they say they’ve achieved ohmic scaling to the atomic limit—which is to say they’ve managed to get electrical wires just a few atoms in width to work the way they’re supposed to.
That paper its here, and the abstract for it reads …
As silicon electronics approaches the atomic scale, interconnects and circuitry become comparable in size to the active device components. Maintaining low electrical resistivity at this scale is challenging because of the presence of conﬁning surfaces and interfaces. We report on the fabrication of wires in silicon—only one atom tall and four atoms wide—with exceptionally low resistivity (~0.3 milliohm-centimeters) and the current-carrying capabilities of copper. By embedding phosphorus atoms within a silicon crystal with an average spacing of less than 1 nanometer, we achieved a diameter-independent resistivity, which demonstrates ohmic scaling to the atomic limit. Atomistic tight-binding calculations conﬁrm the metallicity of these atomic-scale wires, which pave the way for single-atom device architectures for both classical and quantum information processing.
Simmons and her team basically hand-crafted their circuitry, “covering a silicon crystal with a layer of hydrogen atoms and then carving out several-nanometer-wide channels in the hydrogen using the tip of a scanning tunneling microscope,” according to Nature, which reviewed the Science report.
It is not a done deal, the trick will be to translate the experiment to large-scale manufacturing processes, and here is where there is still a problem.
Ferry [David Ferry, an electrical engineer at Arizona State University in Tempe] agrees that the work has significant implications for the microchip industry. “Before this paper there was perhaps one more generation of microchips, whereas now there might be two or three generations,” he says. He adds that it may be possible to reduce component lengths to as small as 5 nanometres.
But Suman Datta, an electrical engineer at Pennsylvania State University in University Park , doubts that Weber and Simmons’ work will have a direct impact on the manufacture of conventional electronic devices. He praises the study as a “nice science experiment” but argues that making practical chips will probably involve lowering the concentration of phosphorous and therefore increasing resistivity.
Ferry doesn’t think that the low resistivity of the wires is in itself the crucial point. What matters, he says, is that the wires exhibit classical rather than quantum behaviour. “When you have quantum coherence the transistor doesn’t turn on and off like you expect it to,” he explains, “and if the transistor doesn’t work like it ought to then Moore’s law is ended.”
Where will it all end? Gordon Moore stated in an interview that the law cannot be sustained indefinitely, and yet the law has often met obstacles that first appeared insurmountable but were indeed surmounted before long. In 2003 Intel predicted the end would come between 2013 and 2018 with 16 nanometer manufacturing processes and 5 nanometer gates, but this just might change that.
As an aside, Some see the limits of the law as being far in the distant future. Lawrence Krauss and Glenn D. Starkman announced an ultimate limit of around 600 years in their paper (it is only 4 pages), based on rigorous estimation of total information-processing capacity of any system in the Universe.