The adage that technology develops at an accelerated rate seems all too true in the realm of microelectronics — soon to be called nanoelectronics. As I discussed in “Tech Page: Faster, Smaller, Better,” in the October 2001 issue, IBM and Intel recently announced breakthroughs in shrinking transistors, breakthroughs that will lead to processors and communications chips far more powerful than those available today. Amazingly, the day that issue went to press, I learned of further developments that will lead to even smaller, faster, and more powerful electronic components.
Those developments are based on carbon nanotubes, in which individual carbon atoms are arranged to form hollow tubes. Scientists at Delft University of Technology in the Netherlands have succeeded in building a transistor from a single nanotube, which measures one nanometer (one billionth of a meter) in diameter — about 10,000 times thinner than a human hair. Using an atomic-force microscope, which can move individual atoms, the Delft team pushed on the nanotube, causing it to buckle in two places. Such deformities restrict the flow of electrons, letting a single electron pass through the tube at a time. That electron can be used to turn the device on or off. Conventional transistors require hundreds to millions of electrons to perform the same task.
IBM researchers also have been working to develop useful carbon-nanotube transistors. One hurdle to practical application of that technology has been reliable replication. Depending on their size and shape, nanotubes can behave as if they are metallic (conducting) or semiconducting. Previous methods of forming large numbers of nanotubes created a mixture of metallic and semiconducting types that stuck together to form ropes or bundles. Those bundles had to be separated manually to isolate the semiconducting nanotubes required to make transistors. That's fine for experimentation but far too slow and tedious for mass production.
To solve the problem, the IBM team developed a method of constructive destruction that selectively destroys the metallic nanotubes in a bundle. Mixed metallic and semiconducting bundles are deposited on a silicon-oxide wafer, and metal electrodes are formed over the nanotubes. Those electrodes “turn off” the semiconducting nanotubes, preventing any electrons from flowing through them. The metallic nanotubes are left unprotected, and applying an appropriate voltage to the wafer destroys them, leaving the semiconducting variety untouched.
IBM's ultimate goal is to create a useful electronic device out of nanotubes, and that goal is close at hand. The company recently announced that it has built a voltage inverter from a single carbon nanotube — the world's first single-molecule logic circuit. Also known as a NOT gate, a voltage inverter is one of the fundamental logic circuits that form the basis of all computers. (The others are AND and OR gates.)
To achieve that monumental feat, the IBM team created an n-type nanotube transistor, in which electrons carry the electrical current. Until now all nanotube transistors have been p-type, in which electron-deficient areas called holes carry the current. Voltage inverters consist of both types of transistors, and IBM found a way to convert a segment of a single nanotube from p-type to n-type, leading directly to the voltage inverter (see Fig. 1). More important, the inverter's gain (the ratio of output to input amplitude) is greater than one, which is essential for assembling gates into useful electronic circuits that will find their way into smaller and more powerful tools for electronic musicians.
For more information on nanotechnology, see “Tech Page: Quantum Mirage” in the May 2000 issue and “Tech Page: Nanocomputers” in the June 1996 issue.