Dr. Velev and fellow researchers at the University of Delaware sparked the growth of the tiny wires by applying an alternating electrical current to minuscule gold flakes suspended in liquid. The flakes, each measuring several billionths of a meter in diameter, assembled themselves into microwires too small for the eye to see but able to carry a current. The process is known as dielectrophoresis.

The gold flakes can self-assemble into wires very rapidly: They grow faster than 50 millionths of a meter per second to lengths topping 5 hundredths of a meter. The wires are about one millionth of a meter in diameter, a fraction of the diameter of a human hair.

"When you miniaturize things, self-assembly is important," Velev says. There is no chemical reaction, no need for soldering, and the electrical connections are made spontaneously, he explains. Typically, only one wire grows in a solution, but more than one could be grown if the voltage were increased.

"One of the bottlenecks in microchip fabrication is connecting wires to all of the elements inside the chip," says Thomas Jones, a professor of electrical and computer engineering at the University of Rochester in New York, who is familiar with Velev's work.

He explains that there are physical limits to how small chips can be made, because there is a limit to how close to one another wires can be placed without causing interference.

"But if something can self-assemble, we could potentially develop microchips that are more complex, using this better way of connecting things together," Mr. Jones says. "This is exciting."

Jones explains that the way the microwires are grown or assembled is analogous to the well-known phenomenon of iron filings forming chains under the influence of a magnet.

Velev says the assembly process for the microwires is relatively simple and easy to control. "We have a fairly good degree of control, because the electric voltage can control the direction of the wire's growth, and the distance between electrodes can control its length." Another advantage: The wires can automatically repair themselves.

The research still is in its early phases, so the scientists are looking at controlling the wires to make rudimentary electrical circuits. They also are studying other materials, such as silver, platinum, plastic, semiconductors, and possibly carbon, for dielectrophoretic assembly.

Velev says the gold microwires could be used in biochemical sensors, and they already can detect cyanide and thiol.

A prototype sensor device could be built in a year or so, but commercial application of the technology is likely five or more years away.

Any device would include a wire or wires in a solution, and encased in glass.

"There are a lot of efforts now to develop new sensors to detect bad things," says Eric Kaler, dean of the college of engineering at the University of Delaware and a co-discoverer of the microwires with Velev. "This technology is not yet mature. We need to see work on an engineering prototype device that could take advantage of it."

Initially, Dean Kaler sees the technology being used to detect contamination of water, such as reservoirs. Eventually, it also could be used with living cells as a sensor of biological or chemical agents, which cause changes in the surface of the gold. Kaler says he already is working with the Army on nerve or chemical warfare agent detection.

"Most people think water and electricity don't go together, but they do in the biology of the body," Kaler says. "We've developed a way to assemble electrical connections in a water environment. This offers a new way to potentially interface living systems with electronic readouts, so we could use a cell as a sensor."

(c) Copyright 2001. The Christian Science Monitor