Heath should know, however. He is one of the rare scientists to participate in two potential revolutions before the age of 40. As a graduate student at Rice University, Heath ran the experimental apparatus that generated the first C60 molecules and, ultimately, Nobel Prizes for the three senior members of the collaboration: Robert F. Curl, Harry W. Kroto and Richard E. Smalley. The 1985 Nature article describing the discovery has garnered over 3,000 citations, while Heath himself has since published seven other papers with over 100 citations each. UC Berkeley chemist Richard J. Saykally, with whom Heath did a post-doctoral fellowship, calls Heath "the most brilliant experimental scientist" he’s ever worked with.

Heath, 39, earned his undergraduate degree in chemistry from Baylor University in 1984 and obtained his Ph.D. in 1988 with Smalley at Rice. After two years at Berkeley, and a four-year stint at IBM’s T.J. Watson Research Laboratory, Heath arrived in 1994 at UCLA, where he is currently a professor of chemistry.

You’ve been described by Harry Kroto as having green fingers, an almost uncanny ability to get experiments to work. What does that take?

I don’t know. I just have much more success getting science to work than the average person. People have skills, right? And there’s some stuff I’m not very good at. I’m not very good at theory, but when it comes to doing really hard experiments on things that people care about, I’m good at making the experiment work. It’s a kind of intuition. I always had good luck figuring out what will work and what won’t, and what’s interesting and what’s not, and making it work.

Where do you think nanotechnology is going? What’s the ultimate promise?

That eventually we’ll have a bottom-up manufacturing technology instead of the present top-down technology.

Could you define those terms?

"Top-down" means you take a material like silicon and whittle away almost all of it–using lithography, for instance. Maybe you paint the whole thing and then remove the paint you don’t want, and end up with a computer chip. "Bottom-up" is more of a biological approach where you somehow bring small amounts of stuff together and chemically synthesize the computer from the bottom up.

How did you get onto this idea?

I started thinking about the problem when I was at IBM. One of the first things I was interested in understanding was how silicon, germanium, and similar materials behave when they get very, very small–a few hundred atoms worth. I developed a few synthesis techniques for controlling the size and shape of small bits of germanium and silicon, and did some single-particle spectroscopy to get pictures of the stuff. Then I came to UCLA and just continued in this vein of working on nano. I worked on quantum dots, thinking all the time about how to make a computer using this technology.

What crystallized it for you?

A few years ago, I was doing work at Hewlett Packard with a colleague named Stan Williams, helping them set up a basic research effort, when Phil Keukes, a computer architect at HP, told us about this computer called Teramac. "Tera" means "ten to the twelfth" and "mac" stands for "multiple architecture computer." Teramac could do 10¹² operations per second, despite the fact that it had a quarter of a million defects in it. And these were hardware defects, not software bugs. If a pentium computer has even one, it’s trash. So the very fact that this machine worked suggested that we ought to take some time and learn about it.

What was the lesson?

There are two issues. For one, the machine had a lot of extra wires, so it was possible to wire around the defects, but that wasn’t the major point. Rather, think about trying to make a computer from a chemical point of view. You might somehow, for instance, imagine making a molecule that has all kinds of arms and branches, and you can somehow connect to these and make a circuit that adds two sixteen-bit numbers together. That’s not that ridiculous to think about in terms of the technology. But there’s a second approach, which is much simpler chemically, and that’s to just make the elements you need for computation–the wires and the switches and the gain elements–and then use another computer to electronically wire those things together. And that’s more or less how Teramac worked–it had the wire and switches and gain elements, and software was used to figure out how to turn them into a computer. That’s the road it led us down.

The other aspect was that Teramac was essentially laid out in a crystalline lattice, and that gave it a lot of redundancy to deal with the defects. What it suggested was that you can take a system that was defective and ordered, and if you actually do things right, you can still turn it into a computer. We then spent two years writing a paper for Science on it. (See J.R. Heath, et al., "A defect-tolerant computer architecture: Opportunities for nanotechnology," Science, 280[5370]:1716-21, 1998). It took that long because Stan and I are chemists and had to learn about computer architecture, and Phil Keukes is a computer architect who had to learn about chemistry. We spent two-plus years thinking about the architecture of the machine, and then we started making one.

Did you utilize the quantum dot work?