Nevertheless, debate continues over whether the billions budgeted for nanotechnology research will be well-spent, helping to move the technology closer to production systems.

The National Nanotechnology Initiative that President Bush signed into law last December earmarks almost $1 billion for nanotech in the 2005 federal budget, to be divided among 10 government agencies. As the bill was making its way through Congress, proponents offered grand estimates that nanotechnology would be a trillion-dollar business by 2015. Many in the electronics industry say those forecasts are unrealistic.

Still, "There are lots of opportunities for nanotechnology to be used in the semiconductor industry," said Richard Knipe, an engineering manager at Texas Instruments Inc. who has worked for a decade on TI's digital light processors (DLPs). "I view nanotechnology as a means to an end. I build displays, and to do that I use nanotechnology." Making the hinges that control the micromirrors in the DLPs involves the self-assembly of layers of an acid that is only five atomic layers thick.

Intel said last week that its enterprise laboratory will investigate whether polymers built of carbon nanotubes could serve as a thermal-interface material between a microprocessor and its heat sink. A company spokesman said the lab will source its carbon-nanotube material from Zyvex Corp. (Richardson, Texas), one of the early nanotech startups, to achieve commercial revenues from both materials and tools. Zyvex engineers have figured out how to suspend high concentrations of carbon nanotubes in solvents so that customers can disperse the material into polymers, such as epoxies.

Also last week, Infineon Technologies AG said it would use carbon nanotubes in a line of power devices.

A broad view of nanoelectronics activity encompasses not just products but also equipment, such as atomic-layer deposition tools and nanomanipulators and probes. Semiconductor capital equipment companies will come together with the fledgling nanoelectronics research community Nov. 14-17 in Austin, at a conference organized by the Semiconductor Equipment and Materials International trade group. The NanoForum is expected to draw 400 to 500 people seeking to leverage the commercial muscle of the established semiconductor tool vendors.

Chemists wanted

"Researchers are doing a lot of work with specialized chemicals so they can identify molecules," said veteran process engineer Roc Blumenthal, the conference co-chairman. "How do you make a machine with multiple channels that can flow those chemicals? Maybe they can leverage equipment that is already out there in the MEMS [microelectromechanical-systems] world to build those channels." Meanwhile, equipment vendors are interested in new markets, such as the machinery needed to create "Buckyballs"--the hollow Fullerene structures that could be infused with drugs and directed to particular sites inside the body, Blumenthal said. In fact, "chemists with experience in nanomaterials are among the most difficult people to find right now," said TI's Knipe.

Zyvex, for one, is attacking the problem of nanotechnology tools by fielding its own probes and manipulators while also teaming up with an instrument maker and an equipment manufacturer.

Zyvex and Keithley Instruments aim to combine the former's S100 nanomanipulator with Keithley's electrical-characterization system, which can sense electrical currents in the femtoampere regime. The S100, which sells for about $140,000, attaches as a module onto a scanning electron microscope to move micro- and nanoscale materials into position. Plans call for Keithley's Asian sales force to represent Zyvex in Asia, said Tom Cellucci, president and chief operating officer of Zyvex.

The company is also working with "a very large semiconductor equipment manufacturer" to combine its manipulators with that corporation's focused-ion-beam equipment, which is used to repair defective chips. Cellucci did not identify the potential partner. Micrion Corp. (Peabody, Mass.) is the largest vendor of focused-ion-beam equipment.

Cellucci said the deals are part of a broad restructuring of Zyvex, from a company that had 51 research projects and 49 employees two years ago to a more profit-oriented operation with a product-creation mentality. Zyvex, he said, expects instrument and materials sales of $10 million this year, with a further $50 million in government research grants. Nevertheless, the company's main goal is a project, expected to take 12 years, to create self-assembled circuits.

Zvi Yaniv, CEO of Applied Nanotech Inc. here, said he believes many people define nanotechnology too broadly, as simply the scaling of materials and devices down into nanometer dimensions. "It is not about getting smaller," he said. "It is looking for new properties" that occur in new materials, such as carbon nanotubes, or for new possibilities that emerge at the nanoscale with silicon. Yaniv said Applied Nanotech is developing emission display technology from carbon nanotubes.

The convergence of biology and electronics is an area ripe for nanotech, said Louis Brousseau, founder of Quantum Logic Devices Inc. here. He pointed to the promise of "personalized medicines" as one example. Another would be in lab work. A drop of blood could be placed on a quantum-logic device and characterized "in 30 minutes instead of 24 hours," Brousseau said at a nanotechnology forum sponsored by the Austin Technology Council last week. Brousseau said he relocated his company from North Carolina to take advantage of the medical-research community in Houston as well as the chip design expertise in Austin.

For all the excitement that nanotechnology is beginning to generate, the discipline remains in its infancy. Cellucci of Zyvex said the effort to move nanotechnology-related research into the commercial realm is inhibited by universities, most of which are slow to license their intellectual property. "Our biggest challenge is that university technology transfer is in such a dismal state," he said, describing what he said were unrealistic demands for upfront fees on top of royalties in negotiations that can drag on for six months or longer.

Reliability at issue

Elsewhere, Stefan Lai, an Intel vice president in charge of memory technology and manufacturing, said the research into single-electron and molecular memories faces two major problems: interconnection and reliability. "A single-molecule memory is only good if you can sense it," Lai said. "How would you connect them in a working-level system?" That area, he suggested, is where research should be steered, "not into fancy molecules."

Also, single-electron transistors and molecular devices are inherently fragile, and sensitive to temperature changes. "These devices all have high failure rates, so there would need to be a major change in the computing paradigm. We would need machines with high fault-tolerance," Lai said.

None of the caveats, however, is slowing the research community or the growing number of small companies eager to identify themselves with anything "nano."