Transistors Of A Very Different Kind -- More Than Moore!

Today, most transistors are made of silicon, their patterns etched by visible light lithography, and soon by ultraviolet light. But although the coming UV-etched chips promise to eventually create truly tiny
elements (0.03 microns - http://www.compaq.com/rcfoc/20010423.html#_Toc512230209), future
generations of transistors may make these bits of silicon seem large, indeed.

Brought to our attention by RCFoC reader Christian Clauss, an April 26 Reuters article describes how IBM researchers have figured out how to build, and now how to mass-produce, transistors 500-times smaller than today's transistors -- out of carbon nanotubes. One of these transistors is about ten atoms in diameter!

[Image - Example of a carbon nanotube single-electron transistor - http://www.research.ibm.com/nanoscience/fet.html]

Totally Tubular.

It turns out that although these tiny carbon nanotubes are normally extraordinarily good conductors of electricity, if they are bent just so, they develop semiconductor properties -- just what we need to make
transistors. But there was a problem: nanotubes come in bundles containing both conducting and semiconducting tubes -- the trick was figuring out how to separate them so that transistors could be made of only the desired, semiconducting, nanotubes. The answer was both as simple as it is complex: pass just the right current through the bundle of nanotubes. It heats and destroys the conducting tubes, while leaving the semiconducting nanotubes (which exhibit a higher resistance to the current) intact. Cute. (Details are at
http://www.research.ibm.com/resources/news/20010425_Carbon_Nanotubes.sh
tml , and a QuickTime illustration of the process is at
http://www.research.ibm.com/resources/news/images/nanoanimation.mov .)

Of course this is the solution to just one of the many technological and economic problems in bringing such a new technology out of the labs and into our pockets. But it is a good wake-up call, reminding us that we are awfully good at not only finding ways to improve how we do things today (such as moving from visible light to its shorter-wavelength cousin, Extreme Ultraviolet, to etch chips), but in stepping back and completely changing the rules.

Imagine the implications, if carbon nanotubes enable us to produce transistors 500 times smaller (hence closer together, and faster) than we're used to. According to Thomas Theis, director of physical
sciences at the Thomas J. Watson Research Center
(http://www.eetimes.com/story/OEG20010426S0069),

"If we can prove that [carbon nanotube transistors] can outperform any future silicon transistor, then IBM will bring the forces to bear to engineer a technology for mass producing them in complex circuits - hopefully a chemical synthetic technique that grows just the kind of nanotubes that we want, where we want them."

This is an excerpt from the "Rapidly Changing Face of Computing, " a free weekly multimedia technology journal written by Jeffrey R. Harrow, Principal Member of Technical Staff for the Corporate Strategy group at Compaq. A more extensive version of this discussion, as well as others around the innovations and trends of contemporary computing and the technologies that drive them, are available at http://www.compaq.com/rcfoc . Jeff's opinions do not necessarily reflect the opinions of Compaq. The RCFoC is a service of, and Copyright 2000, Compaq Computer Corp."