Nanotech revolution

When it comes to matter, size really does matter. The properties of materials that we notice color, hardness, electrical conductivity and so on all depend on the nature and structure of the constituent atoms and molecules. With increasing ability to design and build on an atomic and molecular scale a reasonable definition of nanotechnology we are becoming better and better at developing materials with entirely new properties. Those materials, in turn, become the building blocks for more complex systems and entirely new products. But when an emerging technology is the subject of as much hype as nanotech, it's easy to tune out and stop listening. That would be a big mistake. If we ignore the unsupported claims and misguided speculation, especially about what might be achieved in the near term there remain solid reasons to expect significant long-term developments in what the National Science Foundation estimates to be a trillion-dollar-plus industry during the next 10 to 15 years.

Why nanotech? With it, today's supercomputer could become tomorrow's wristwatch personal assistant. Buildings and machines could signal when they need maintenance and perhaps repair themselves. Our clothing could monitor our health and alert us to environmental hazards. All of these wonders, and many more, are scientifically possible. The difficulty comes in figuring out how and when these things will happen.

The biggest nanotech market now, and probably for some time, is in materials. The use of nanosize particles in products such as cosmetics, sunscreen, paints and a host of other products is already commonplace. Specialized nanotech startups are emerging in droves, and new industries could grow up around them. Right now, for example, it is difficult to count the number of startups selling carbon nanotubes (the strongest and most conductive fibers known), even though there are, so far, no proven commercial applications of nanotubes. Researchers are also learning to assemble molecule-size components into complex composite materials and "smart" materials. Several laboratories, for example, are developing nanostructured membranes for efficient filtering of pollutants from water or air.

But while materials are essential, it is the prospect of new devices and gadgets that truly captures our imagination. Research on nanosensors, for example, is burgeoning. The potential applications of sensitive, selective, inexpensive sensors are almost unlimited from medical diagnostics to chemical and biohazard detection. No wonder many startup companies and university research groups are playing in this arena.

Will we someday integrate functional nanocomponents into the complex systems of IT? In a sense, we already do. Even today, transistor dimensions are measured in nanometers. IBM Corp. has shipped more than 5 million disk drives with a new nanostructured magnetic coating (humorously referred to as Pixie Dust), which will allow us to quadruple the data storage capacity of our products.

There are, however, looming obstacles to further progress. First, as the components have shrunk, manufacturing costs have risen. And even if manufacturing costs can be capped, there are physical limits to the minimum size of a useful silicon transistor or the data storage density of a magnetic disk.

That is why many industry and university laboratories are exploring entirely new devices to process and store information. IBM scientists, for example, have been exploring the potential of carbon nanotube transistors. We are certain that these can be made smaller than any possible silicon transistor, and they may beat silicon in performance as well. We are also exploring exotic new ways of storing information. One, a nanomechanical system called Millipede, stores data as minuscule, erasable indentations in a thin plastic film. Millipede might someday allow a trillion bits of information to fit inside a thin wristwatch.

However, don't expect these new nanodevices in your computer right away. Hundreds of millions of transistors are integrated in the latest microprocessor chips. In contrast, last year IBM built a two-transistor logic circuit from a single carbon nanotube and that was the world's first. It will take some time probably more than 10 years to learn how to integrate billions of nanotube transistors into a useful system. But whether or not new nanodevices eventually displace the silicon transistor and the magnetic disk, nanotechnology is the future of information technology. Stay tuned.