Inventing a new material, say a new polymer or metal alloy, has never been easy-and is often more a case of serendipity than theoretical design. Even at leading industrial facilities such as DuPont’s central research lab, which has invented many of today’s most important polymers and high-tech materials, the process is often a hit-or-miss endeavor governed more by the instincts of experimentalists than the calculated predictions of theorists. At the nanoscale, however, the instincts of researchers often begin to fail, as the strange rules of the quantum world take over. “We’re getting into areas where our intuition, which is often based on our prior experiences, is really quite deficient,” says Ed Wasserman, science advisor to DuPont central research and development.It is, however, a place where theoretical quantum physicists feel right at home-a realm where they can play god, creating new materials by simply shuffling around atoms. One of those gods is Nanomix cofounder Cohen, a pioneer in quantum modeling. In the 1960s, Cohen had the insight that since it is only an atom’s outermost electrons that interact to give a material its properties, it is possible to greatly simplify the calculations needed for simulations by dealing only with those electrons. This mathematical shortcut revolutionized quantum-mechanical modeling. “If you do an all-electron calculation, it’s almost like trying to find the weight of the captain of a ship by weighing the ship and then weighing the captain with the ship,” explains Berkeley physicist Steven Louie, an advisor to Nanomix. By in essence weighing the captain directly, Cohen’s software now makes it practical to use a PC to quickly crank out predictions on nano materials with novel combina-tions of electronic, magnetic and light-handling properties.
But Cohen also realized that such simulations are only as real as the skills of the experimentalists down the hall. So in the early 1990s, he began collaborating with Berkeley colleagues, like Alex Zettl, who were experienced in synthesizing novel materials. The discovery-and actual synthesis-of boron nitride nanotubes in 1994 was one early success. Cohen, pondering the atomic structure of boron nitride sheets, realized that it should be possible to roll up the sheets to form cylindrical molecules. Other researchers had recently made similar molecular tubes from carbon, but no one had extended the technique to boron and nitrogen. Cohen and his students booted up their models and calculated that boron nitride nanotubes would not only be stable but would exhibit interesting electrical properties. Within a year, experimentalist Zettl had found a way to make the novel nanotubes and affirmed the predicted properties.
A neat scientific trick, for sure, but how do you turn that type of physics wizardry into a business?
The immediate challenge for researchers at Nanomix is to perfect the sensors that Janac has promised to begin selling next year. So far, Nanomix’s theorists have identified the specific types of nanotubes that would react most sensitively to the hydrocarbons that can permeate refineries, chemical plants and pipelines. They’ve also predicted how each gas would alter conductance through the tubes, providing a voltage “footprint” that would identify it. Now the experimentalists must figure out how to mass-produce the nanotube-based devices. And they must deliver a product reliable enough to safeguard the lives of plant operators.
That type of development of a new material can easily take a decade. But if Nanomix-using its modeling and theoretical expertise to shortcut the process-can pull off the trick in the next year or so, it will validate its strategy for nano invention. It will also open the door to a whole world of sensors. For starters, Nanomix should be able to tailor nanotube sensors to sniff out just about any gas, from carbon monoxide to the compounds used in chemical weapons. And the company is sponsoring academic research aimed at tuning nanotubes to sense the biochemical whiffs of disease, measuring insulin levels in diabetics or detecting antigens that indicate infection. Nanotubes are the ultimate sensing platform, asserts Janac. “You don’t need more than single-molecule sensitivity, your power consumption is going to be nanowatts, and in terms of size you can’t get smaller-at least in the foreseeable future-than a nanotube. We think we’re going to revolutionize the sensor business with this architecture.”