The development, which was presented both as a paper and a working demonstration at a technical conference here, is the most striking evidence yet that carbon nanotubes may prove to be the material of the future when today’s silicon-based chips reach their fundamental physical limits.
I.B.M., which is one of the biggest proponents of nanotubes for microelectronic applications, has made clear its hope that carbon nanotube technology will be ready a decade from now, when semiconductors are expected to shrink to minimum dimensions of just 5 nanometers. But until now, researchers at universities and chip makers have succeeded in making only individual devices, like transistors, from carbon nanotubes.
The Stanford development is the first time a complete working circuit has been created and publicly demonstrated, suggesting that the material may indeed live up to its promise.
Silicon, a plentiful natural element that functions both as a conductor and an insulator, has already lasted decades longer than computer engineers originally expected, as generations of increasingly smaller transistors have been perfected. It is used by the computer chip industry to etch circuits much finer than the wave length of light, and engineers and scientists say they believe that the material will continue to scale down, at least until the end of the decade.
But sooner or later the shrinking of circuits made from the material will stop, ending the microelectronic era that has been defined by Moore’s Law, the 1965 observation by the Intel co-founder Gordon Moore that the number of transistors that could be placed on a silicon chip doubled at regular intervals.
The Stanford advance seems to hold promise for the belief that whenever the silicon era stalls, the scaling-down process will continue, and permit designers to increase power and capacity of computers far into the future.
The Stanford demonstration came during a session at the International Solid State Circuits Conference, held here annually. A graduate student, Max Shulaker, chose a wooden, human-size hand, connected to a simple motor and gear arrangement on a makeshift stand. Onstage, he threw a switch and the hand shook vigorously.
It was a simple demonstration, but the research group said its goal was to build an entire microprocessor from carbon nanotubes to confirm the potential of the material.
Besides their small size, carbon nanotubes use much less power and switch faster than today’s silicon transistors.
“The bottom line is you can expect an order of magnitude in power saving at the system level,” said Subhasish Mitra, an associate professor of electrical engineering at Stanford and director of the Robust Systems Group. That offers tremendous promise for effectively increasing the battery life in mobile consumer devices in the future, he said.
Other new materials and variations of silicon-based transistors are also being studied to see if they will shrink to smaller sizes. Intel, for example, last year began using a three-dimensional transistor called a FinFET. By turning the device on its side, the chip maker was able to pack transistors more densely on the surface of a chip.
“I’m not saying there is nothing else around,” said H.-S. Philip Wong, a Stanford electrical engineering professor. “It’s just a matter of who wins when you scale down to really, really small dimensions.”
The challenge of carbon nanotubes in their type state is that they form a giant “hairball” of interwoven molecules. However, by chemically growing them on a quartz surface, the researchers are able to align them closely and in regularly spaced rows. They then transfer them to a silicon wafer, where they used conventional photolithographic techniques to make working circuits.
The technological hurdle has been to make reliable circuits even when a small percentage of the wires are misaligned. The Stanford group stated it had perfected a circuit technique that made use of redundancy to work around the imperfectly formed wires.
Dr. Mitra said that “99.5 percent looks very nice on a PowerPoint slide. But when you’re talking about 10 billion things, .5 percent of 10 billion is a really large number, and that completely messes things up.”
Beyond microelectronics, carbon nanotubes are showing promise in commercial applications like rechargeable batteries, bicycle frames, ship hulls, solar cells and water filters, according to an article in the Feb. 1 issue of the journal Science.