91av

THE squeaky-clean image of buckytubes has taken a beating. Ever since their
discovery, these carbon nanotubes have dazzled scientists with their near-ideal
combination of strength, heat conduction and electrical properties. But now the
truth is out: research published this week shows they could be one of the
noisiest components ever considered for use in electronics.

A team from the University of California at Berkeley and the Lawrence
Berkeley National Laboratory, led by Alex Zettl, has reported on the first
extensive study of noise in nanotubes. Noise is the name given to voltage
fluctuations that can occur when a constant electric current is injected into a
device.

Zettl found that buckytubes were millions of times noisier than conventional
conductors and semiconductors. This unususpected flaw in their character might
rule them out for some applications.

“It’s good work,” says Hongjie Dai, a chemist at Stanford University in Palo
Alto, California. “If you talk about using nanotubes as transistors in computer
chips, this kind of noise level could be a problem.”

Zettl says: “Some people interpreting my results have said, `Boy, you just
threw cold water on this entire field. It’s dead.’ But for me it’s exactly
opposite.” He takes the view that the phenomenal noise is just another
fascinating and unexplored characteristic of nanotubes.

“If you want to solve the problem you better find the origin first,” says
Dai. But Zettl has a theory about what is happening. In conventional conductors,
noise can be produced when atoms of a contaminant move around, disrupting the
flow of electrons. Zettl says some researchers assumed nanotubes wouldn’t have
this problem because their carbon atoms are tightly bound, leaving no path for
disruption.

He believes the noise arises because nanotubes are built up from rolled
sheets of carbon just one atom thick. Every electron passing through “sees” the
surface of the sheet, and this makes the nanotube vulnerable to the effects of
anything striking the surface. In contrast, in a conventional conductor such as
a copper wire, most electrons flow through the centre of the wire so the surface
is relatively unimportant.

According to Zettl, gas molecules from the environment could be scurrying
about like gnats on the outside of nanotubes, disrupting electron flow. If so,
it might be possible to solve the problem by building nanotube-based electronics
in clean rooms, sealed off from outside air, like those used by chip
manufacturers.

Zettl’s group is also studying multiwalled nanotubes—groups of
concentric tubes resembling closed telescoping antennas. He says there may be a
way to send electric current through the inner tubes to avoid noise.

Dai and Zettl agree that in some applications noise won’t matter. For
instance, when nanotubes are used as chemical sensors
(91av, 25 December 1999, p 12),
the target substances produces such a strong signal
that the noise can’t drown it out.

“I don’t want to pretend this noise is not a big problem,” Zettl says. “But I
don’t think it’s at all a show stopper.”