Getting the most out of carbon nanotubes

David L. Shenkenberg

Carbon nanotubes have been proposed for imaging and therapy, but the electronic properties and the length of the nanotubes can affect their functionality in these applications.

In the October 2006 issue of Nature Nanotechnology, Michael S. Arnold and colleagues at Northwestern University in Evanston, Ill., described a method of separating carbon nanotubes that was a major breakthrough. They separated semiconducting and metallic nanotubes using density-gradient centrifugation. They put surfactant-coated nanotubes into test tubes and added layers of various concentrations of iodixanol solution to create a density gradient. As the test tubes spin in the centrifuge, the different types of nanotubes float upward or downward and are separated by their individual densities.

Now scientists at the National Institute of Standards and Technology (NIST) in Gaithersburg, Md., have modified the Northwestern technique to separate carbon nanotubes by length. Lead researcher Jeffrey A. Fagan said that this method can process 50 to 100 mg of carbon nanotubes per day with a few bench scale centrifuges, more than enough for many applications. It also could be scaled up to industrial production levels, as with the original Northwestern process.

Although size exclusion chromatography already successfully separates nanotubes by length, the NIST technique has higher throughput, costs substantially less and can separate longer nanotubes. It also achieves almost 100 percent purity, whereas unprocessed nanotubes contain about 60 to 80 percent junk.

In the NIST technique, the longer nanotubes float upward in the centrifuge tubes faster than the shorter ones. Courtesy of NIST.

The NIST scientists used the same materials as the Northwestern ones but increased the density of the liquid, which caused the nanotubes to float upward because they are less dense than the liquid. The longer nanotubes float upward faster than the shorter ones because the buoyancy increases with length faster than the frictional force. The separation takes about 20 h with the centrifuge running at 20,000 rpm at an average force of 32,000 g.

Advanced Materials, May 5, 2008, pp. 1609-1613.