Zhang's alternative is to grow iron particles, 20 to 50 nanometers in diameter, and graft onto each particle a small amount of palladium to act as a super-catalyst. The particles are small enough to suspend in the water, eliminating the need for pumping water into trenches for treatment. The nanoparticles' larger surface area, when compared with the same amount of conventional cleanup material, also makes them much more reactive.

"In the conventional method the reaction often will go halfway and stop, because there isn't enough energy to complete the reaction," says Zhang. "So you're left with tons of iron filings in the water, which creates another hazard. With the nanoparticles the reactions tend to be very complete. The material all gets used up."

Zhang's method has also been found effective for removing cyanide, and may be adaptable for use in soil treatment and for nuclear waste treatment. One obstacle remains: it takes two weeks to grow enough material in Zhang's lab even for a limited field trial. "The key to commercializing this process will be to make the material in a large enough quantity to get economies of scale," says Zhang, who is in discussions with chemical companies and expects to have industrial partners signed on to do just that by the end of the year. Other nanotechnology research projects being funded by the EPA might lead to particles that treat automobile exhaust gas at the source, even before it disperses into the atmosphere, or the development of highly sensitive sensors that can detect minute amounts of toxic material. As with Zhang's research, these projects seek to exploit the greater reactivity that smaller-sized molecules will provide, as well as the ability to design these molecules in a manner that is uniquely suited to attacking a particular pollution problem.

Ultimately, the real promise of nanotechnology may come not only through treating pollution, but by avoiding its creation in the first place, replacing existing manufacturing processes with cleaner, nano-based alternatives. Intematix, for example, a small company based in Moraga, CA, is developing a method for adding carbon nanotubes to plastics that will give these plastics the properties of a metal. This will allow plastic parts of the automobile to be painted without requiring toxic solvents to bind the paint to the plastic. If the company succeeds in making the process economical for automobile manufacturers to adopt, it will reduce the amount of polluting material used for each car coming down the assembly line.

"Toxic cleanup is useful, but is not really where I set my sights for this technology," says Tina Masciangioli, a technology policy fellow with National Center for Environmental Research, the group in the EPA that monitors the progress of fund recipients. "I like to look further ahead, where we learn to make things that don't pollute to begin with. That's where the potential benefits are just tremendous."