A new carbon nanotool springs to life.

Multiwall nanotubes contain several concentric cylinders of carbon atoms, one packed inside another, like nesting wooden dolls. Now, physicists have managed to peel back the outermost layers on one end of such a structure just 100 atoms or so wide and pull out the inner cylinders.

The finding suggests that these nanotubes could, one day, be the tiny bearings and springs of nanosize machines.

John Cumings, a physics graduate student at the University of California, Berkeley, performed the experiments by incorporating a scanning tunneling microscope (STM) into a transmission electron microscope. While the electron microscope took high-resolution pictures, he used a tough nanotube attached to the STM's probe to manipulate a multiwall nanotube fixed to the microscope stage.

After removing the outer few layers of atoms from the fixed tube's end, Cumings found he could slide its inner cylinders--perhaps four or so--in and out of the outer shell.

During one test, the connection between Cumings' probe and the telescoping nanotube broke. To his surprise, the extended layers snapped back into their sheath, as a nearly frictionless spring would. Cumings and Alex Zettl, also at Berkeley, described the research in the July 28 SCIENCE.

"It's interesting because they're directly imaging the structure of the nanotube during a dynamic process," comments Richard Superfine of the University of North Carolina at Chapel Hill.

Cumings and Zettl suggest that van der Waals forces, which lubricate graphite sheets, also operate between the layers of multiwall carbon nanotubes. The forces, which would snap wayward tubes back inside, also would permit nearly perfect bearinglike movement, they say. Although Cumings wasn't able to test rotation of one tube within another, he believes an extended section should easily turn about its long axis, and he hopes to prove it soon.

Cumings extended and retracted layers on several nanotubes 10 to 20 times each. The nanotubes that Cumings manipulated show no signs of wear and tear under nanoscale magnification, he and Zettl report.

"This is really like molecular perfection," Cumings says. When developers incorporate parts onto a microchip--or potentially a nanochip--they want them to last. "It's like having an automobile that you know you can t take into the shop," he says.

Building up a "nanotechnology toolbox" is an active area of research, says Superfine. "We're trying to develop and understand the fundamental elements that go into any kind of nanomachine, whether it's bearings or springs or gears," he adds.

The next challenge will be to integrate components, such as the one that Cumings found, into functional devices, Superfine says.

"That's going to be the next fun part," he says. "I think over the next couple years, what you'll start seeing is people assembling devices by hand. The next step [after that] ... is going to be understanding how you can set up a process where you can make 10,000 of these or so on a chip."

"I think that the report of Cumings and Zettl will send shock waves in the nanoscience community, and even larger," adds Laszlo Forro, of the Ecole Polytechnique Federale de Lausanne in Switzerland. "It will open new avenues."