Silicon shortfall begets home-grown components

Scientists say the IT industry is less than five years away from manufacturing mass-market computer memory and transistors based on organically grown components.

While the emerging field of nanotechnology has already led to prototype computers and processors built from molecules, researchers

in Canada and the United States are now expanding its scope. With little more than organic polymers or bacteria, they are exploring ways that molecular biology could store vast amounts or data and offer new applications for smart cards, LCD panels and electronic paper.

Rafik Loutfy, a scientist with the Xerox Research Centre of Canada (XRCC), says the company has been working on organic electronics for almost 20 years, concentrating primarily on areas that would benefit its core business of printing and copying. In the last five years, however, Loutfy says the group has decided to exploit its knowledge in the field of organic electronics by growing thin-film transistors.

Over time, experts say there will be considerable challenges in using silicon to make increasingly smaller transistors. There has been considerable concern over the potential power consumption of devices using these components. Intel Corp. recently announced a product roadmap that would see the processor maker move to Terahertz transistors by 2005 that would include an ultra-thin layer of silicon on top of an embedded layer of insulation to switch them on and off faster while reducing leakage. IBM, meanwhile, has announced its own plan to move to a double-gate transistor.

“”All these solutions are viable, and usually they will keep pushing the existing technology all the way to the limit,”” Loutfy says. “”But we believe the limit will be reached in 2010-2015, where with silicon, you cannot just make it smaller anymore.””

That’s because as silicon-based transistors become smaller, the amount of noise and extraneous signals could make it difficult to extract data, according to Loutfy. It could also exorbitantly increase the costs of manufacturing. “”Now it might be $60 billion just to build a manufacturing line for these kinds of devices.””

Nanotechnology, which refers to devices with dimensions built within billionths of a metre, or a nanometre, holds great promise because it means hardware could be constructed at a molecular level. Last year IBM unveiled a simple circuit made out of carbon, for example, while scientists in Israel made a tiny computer built out of DNA.

Loutfy’s focus, on the other hand, has been in the development of printed organic electronics. “”We’re not going after packing many transistors into a small area,”” he said. “”We want to almost grow the transistor to apply them in applications that would require flexibility, low cost and we think the adoption of this approach will come much earlier.””

The XRCC scientists foresee the more flexibile printed organic transistors creating digital paper that would be used for e-books, or in product tracking and inventory control for radio-frequency ID tags.

At the University of Conneticut, Professor Robert Birge is also thinking big. For years he has been working on a project in which lasers are turned on a protein called bacterio-rhodopsin to write data into slices on a transparent plastic tube, or cuvette. These cuvettes can store about 7GB of data.

Birge got started on his work after taking part in a consulting project for the United States Navy, which was spying on a similar protein-based memory project in Russia.

The U.S. air force is already trying out the memory when conducting reconnaissance missions, but Birge’s goal is to bring it into the mainstream. Like Loufty, he says this may be only three or four years away. The big disadvantage right now is that the hardware necessary to read and write into the memory is still very expensive.

“”We can’t make anything that works for under $25,000,”” he says, adding that a third party makes the hardware based on his specifications for the military. “”We simply don’t have the capability to make the kinds of devices that are necessary for commercial applications. We’re experts on the memory medium, not the total system.””

Two components — blue lasers and spatial light modulators — contribute to the memory’s expense. Once these drop in price, Birge says it may be more possible to see the organic memory end up in enterprise systems, or even the PC desktop.

“”If anything, I should be further along than I am,”” he says, adding that scientific advancements in the last few years have allowed him to modify proteins to suit specific applications. “”The type of genetic engineering that we can do today was not possible three years ago.””

Share on LinkedIn Share with Google+