“What’s new about our method is that we know how to set a certain length of the nanotube protruding from the support tip in a convenient manner,” Niels de Jonge of Philips told nanotechweb.org. “Others mount nanotubes on atomic force microscope tips and shorten the nanotubes after mounting, but this is a very delicate procedure.”

De Jonge and colleagues used a piezo-driven nanomanipulator in a scanning electron microscope (SEM) to mount each nanotube onto a tungsten wire tip with a curvature radius of 50 nm. “This instrument allows good control over the manipulation - imaging at video frame-rate with a magnification of x40,000 and manipulation with steps of 50 nm,” said de Jonge.

First, the scientists stuck the tungsten tip into a piece of carbon tape to act as a glue, before bringing the tip close to a bundle of multiwalled carbon nanotubes grown by an arc discharge method. The team picked a single nanotube in the bundle to approach by looking for one that was slightly apart from the other tubes, pointing in the same direction as the tip, and with the desired diameter. As the tip came within a few hundred nanometres of the chosen nanotube, the tube bent towards the tip and stuck to it. Pulling the tip then caused the nanotube to align with the tip, with a contact length of about 1 micron. All that remained to do was break the nanotube away from the bundle, a process the team achieved by Joule heating or by applying a mechanical force.

To control the length of tube protruding from the tip, the scientists exploited the small region on many nanotubes that is thinner than the rest of the tube. They found that if they attached the support tip close to such a region, which is visible as a dark spot in the SEM, the nanotube was much more likely to break at this point under the influence of a current or mechanical force. The team mounted 19 carbon nanotubes with lengths of 0.27-9.5 microns. For 12 of the tubes the difference between the desired length and the obtained length was less than 50% - a figure that is sufficient for the construction of electron sources and scanning probe tips. So the scientists say that the success rate for the procedure was about 2/3.

“I think that the result will speed up the development of better tips for electron microscopes and scanning probe microscopes,” said de Jonge. “Improved microscopy techniques are demanded by several fields, such as basic materials science, nanotechnology and life science. Additionally, the details of the experimental methods we described may prove useful for other researchers working on the manipulation of nano-objects.”

The researchers reported their work in Nano Letters.