Tuesday, February 21, 2006


IBM Squeezes More into Microchips

I have previously discussed the nature of Moore's law and the reasons why this empiracle law may soon need drastic new techniques in order to survive.

Integrated circuits have been effectively made by the same process for the last 30 years, photolithography. In this process the feature size of the components of the circuit are determined by the ability of a light source to harden a photoresist material laid on a metal surface, a few nanometers thick, which itself is laid on the silicon substrate. The entire surface is then exposed to a chemical etching agent, which removes all except the hardened photoresist, that not exposed to light, and the metal underneath. The photoresist that remains after the first etching is then itself removed using a different chemical treatment, and all that remains is a layer of metal in the same shape as the components required. To reduce the size of the components, it is simply necessary to reduce the wavelength of the light that is used to mark the pattern, and hence reduce the diffraction of the light that results from its wave nature.

Present fabrication techniques rely on light in the ultraviolet region, allowing the production of feature sizes ~100 nm, and this is about as far as present technology can take the technique; the next step down the wavelength chart takes us into the domain of x-rays. The problems here arise from the difficulty associated with finding sources of x-rays, e.g. synchrotrons, and the even greater challenge of focusing and reflecting the beams, the cost, even by the standards of the electronics industry, would be huge with predictions of a fabrication facility costing in the region of $200 billion by 2015.

IBM squeezes more into microchips is the title of an article appearing on todays Technology news on the BBC website (the story also appears on several other sources today). The technology giant claims that they now possess techniques that will help overcome these fundermental problems, without resorting to costlier and unproven chip-making methods. They say they have been able to etch circuits on silicon wafers that are a third of the width of those produced using existing technology. The methods used by the scientists at IBM's Almaden Research Center in San Jose, California, uses a method called deep-ultraviolet optical lithography. "Our goal is to push optical lithography as far as we can so the industry does not have to move to any expensive alternatives until absolutely necessary," said Dr Robert D Allen, manager of lithography materials at IBM's Almaden Research Center.

The IBM scientists have created the tiniest, high-quality line patterns ever made using deep-ultraviolet optical lithography, a technology currently used to "print" circuits on chips.
The distinct and uniformly spaced ridges are only 29.9 nanometres wide.

This is less than one-third the size of the 90-nanometre features now in mass production and below the 32 nanometres that industry consensus held as the limit for optical lithography techniques. According to the researchers, "This result is the strongest evidence to date that the industry may have at least seven years of breathing room before any radical changes in chip-making techniques would be needed."

Does this provide a death blow for molecular electronics? On the surface of it, it would appear that this is nothing more than a short extension to the dominance of silicon components, with molecular electronics still possessing the strongest case for succession. But with work continuing at beathtaking pace in the field of quantum computing (something i intend to talk about in the future), this delay could well provide the time required for chip technology to 'skip a generation' and consign molecular electronics to the great what if bin of science and technology.

This page is powered by Blogger. Isn't yours?