Careful construction. Concentrated heat can induce nanowires or nanotubes to sprout from a silicon chip.

Turning Down the Heat on Nanowires

Nanometer-scale wires are stronger and more capable of conducting heat than the average wire, but manufacturing them or their cousins, nanotubes, is difficult and mass production is currently impossible. In an attempt to break that barrier, researchers have taken a small step toward producing nanowires and nanotubes directly on silicon chips.

One reason it's tricky to make and hook up nanowires, made of single-crystal silicon, or nanotubes, made of molecule-thick sheets of carbon, is that they are fragile and can easily break when transported from one substrate to another. Their synthesis also requires temperatures high enough to destroy silicon microelectronics, such as computer chips on which nanowires or nanotubes might carry out faster computations than less dainty connections. But a research group at the University of California, Berkeley, has worked out a method of localizing the heat in order to assemble nanowires right on a chip at room temperature.

The team, led by mechanical engineer Liwei Lin, placed microelectromechanical systems, or MEMS, in tiny bridges on an etched silicon chip. The MEMS bridges acted as resistors, carrying temperatures of up to 1400°C, while micrometers away the chip itself was at room temperature. The experimenters introduced acetylene or silane into the vacuum chamber as sources of carbon or silicon, respectively. Nanowires grew from the MEMS to lengths of 5 to 10 micrometers all anchored at one end to the MEMS bridges, the team reports in the 30 June issue of Applied Physics Letters. The next step, says Lin, is to find a way to get the wires or tubes to anchor at the other end, which would provide a complete nanowire or nanotube right on the chip.

The research "opens up a new possible way that one could pursue in integrating nanostructures with microstructures," says Charles Lieber, a nanotechnologist at Harvard University. In principle, the team could grow silicon nanowires in one region of a chip and then switch to grow carbon nanotubes in another spot, but so far, "they don't have a lot of control over the growth," he says. "The idea is a nice idea," Lieber says. "How important it is beyond that, it's a little early to say."

Related site
Liwei Lin's home page