IBM's carbon nanotube FET hints at post-silicon circuits
By Paul Kallender
EE Times
(10/05/01 15:32 p.m. EST)
TSUKUBA, Japan — IBM Corp.'s manufacture of a top-gate carbon nanotube field effect transistor (CNTFET) is a key breakthrough in post-silicon circuit design, according to a leading IBM researcher speaking at this weeks' Nanotube Symposium. Separately, another key researcher challenged the international research community to develop nanometer-scale magnets.
IBM's CNTFET places an independent gate just above a carbon nanotube that connects a FET's source and drain, potentially yielding "more than one order of magnitude better performance" than back-gate CNTFETs, said Phaedon Avouris, manager of nanometer scale science and technology at IBM's T.J. Watson Research Center (Yorktown Heights, N.Y.).
IBM's recently developed p-type and n-type CNTFETs, wherein the gate and nanotubes were separated by about 150 nanometers, were comparable with traditional silicon dioxide FETs. Now the company will work to shrink the gap for top-gate CNTFETs to 2 nanometers, which will increasing the transistor's performance exponentially and possibly fulfill the promise of carbon nanotubes as a nanoscale replacement for silicon circuits.
"The game is to get to 2 nanometers and we're going down," Avouris told EE Times. "The [CNTFET] devices we have now compare well with silicon dioxide and that's at 150 nanometers. Wait until we get to 2 [nanometers]."
The new design builds on several developments IBM has made over the last few months, which Avouris outlined to an audience of about 500.
This spring IBM developed a "creative destruction" method of making nanotubes, which it touted as a major step toward the bulk manufacture of the tubes. Last month, the Yorktown Heights labs revealed it had built molecular-scale logic circuits that mixed p-type and n-type CNTFETs with carbon nanotubes to form logic gate voltage inverters. Until that achievement, researchers had succeeded in developing arrays of only p-type CNTFETs. IBM researchers found that by annealing normal p-type CNTFETs in a vacuum, they could manufacture both p-type and n-type and ambipolar CNTFETs to form complementary intermolecular logic gates, or nanoscale CMOS-like logic.
"So you have FETs along the length of the nanotube, and you can connect them anyway you want and form circuits," Avouris said. "For a long time people had dreamed of making a circuit out of a molecule. But now we can make devices."
Researchers at the conference — which marked the discovery of nanotubes here 10 years ago by NEC Corp.'s Sumio Iijima — sought to map the potential of carbon nanotube technology, which they agree is still years from commercial implementation in circuits. But a series of discoveries have brought researchers to the cusp moving the tubes out of labs and into commercial products, said Leo Esaki of the Science Academy of Tsukuba. In recent years researchers have found they can stuff nanotubes with carbon 60 and a variety of other molecules to form complex structures.
CNTs could produce a clutch of "billionaires" among those making key breakthroughs over the next decade, Esaki said. Japan is developing a nanotechnology industry, with CNTs a key weapon, and the Japanese government last year founded the Center for Advanced Carbon Materials to speed this work, which is centered in Tsukuba. Esaki compared the work to the Apollo space program in the United States in the 1960s.
"The Japanese government has spent $22 billion building Tsukuba, and NEC and a score of other firms have also invested here," Esaki said. "By comparison the U.S. government spent $25 billion on Apollo."
NEC's Iijima called carbon nanohorns the next frontier for developing nanobatteries with up to 10 times the capacitance of the best lithium ion batteries. The horns are easy to make at room temperature, and NEC can produce them with a 90 percent yield at a rate of 50 grams per hour. That's 50 times faster than the production of single-walled nanotubes. NEC is now developing nanohorn cell phone batteries, and Nissan Motors has invested $300 million to develop fuel cells from them, Iijima said.
Challenge issued
Atsushi Oshiyama, a former researcher and currently a professor of the Institute of Physics at the University of Tsukuba, challenged researchers to confirm that CNTs can be made into nanomagnets, and that CNT "peapods" filled with carbon 60 molecules can be made into superconductor wires. Those findings could revolutionize electronics, he said.
"If we manipulate carbon correctly, we can make carbon magnets," Oshiyama told EE Times. That finding could be confirmed "within a couple of years," he said.
Support this view, Steven Louie of the University of California at Berkeley said his studies of carbon-boron-nitrogen nanotubes and peapods confirm that nanotubes could be made into quantum wires.
IBM's Avouris warned the audience not to forget how far nanotube research is from commercial application.
"We still have critical problems related to production and cost. Do you know how much it costs to make a silicon FET?" he asked.
"One five hundredth of a cent." |
