"We thought it would be great to get selectivity [the percentage produced with one specific structure] greater than 80%, but we achieved 92%," says Yan Li, professor at the Beijing National Laboratory for Molecular Science at Peking University in China. The key to their success was a nanocatalyst made from a tungsten-cobalt alloy, which has a unique crystal structure and is very stable even at the high temperatures required for growing carbon nanotubes.

Although carbon nanotubes are already used in applications such as composites, touch panels for smartphones and lithium-ion batteries, it is much more difficult to exploit their electronic properties – mainly because they are highly dependent on the exact nanotube structure. For a start, the structure of a carbon nanotube determines whether it is metallic or semiconducting, and this in turn depends on the "chirality" – essentially a measure of the "twist" in the honeycomb carbon lattice. The chirality also determines the size of the electronic bandgap for semiconducting nanotubes, which means that nanotubes must always have the same structural properties to produce devices with reproducible performance characteristics.

Different groups have tried to separate nanotubes with different chiralities by using an external voltage to burn off the metallic nanotubes, or by dispersing them in a solvent. However, these methods are either limited in their specificity or introduce impurities, which impair the nanotubes' electronic properties.

The nanotubes synthesized by Li and her team have a chirality described as (12, 6), which means they are metallic – useful for nanowires but not field effect transistors. Li tells nanotechweb.org that using different faces of the catalyst crystal should yield tubes with chirality either (16, 0) or (14, 4), and both these types are semiconducting. What's more, the bandgap of these types of carbon nanotubes is ideal for electronic applications.

Catalysts feel the heat

Some attempts have already been made to exploit catalyst crystals as templates for growing nanotubes with a specific structure, but until now the proportion of nanotubes with the same chirality was limited to just 55%. "People did not realize that if you use a catalyst as a template you need to make the structure of the catalyst stable," explains Li.

Li’s background in inorganic chemistry and metals directed her to tungsten, known to have exceedingly high thermal stability. Tungsten itself is not a catalyst so the researchers investigated alloys with cobalt, which is catalytic. "At first we were just trying to find a very stable catalyst, but on the way we found the structure is also very important," Li tells nanotechweb.org.

Li explains that the atoms are very densely packed in monometallic nanoparticles, which allows a number of different carbon nanotube structures to fit onto the atoms – which in turn lowers the chiral specificity of the tubes grown. "Our new catalyst is an alloy and the crystal structure has very low symmetry. The arrangement is very unique, so only one structure of carbon nanotube can fit on a particular face."

Next steps

Li’s team collaborated with researchers at the Hong Kong Polytechnic University, the Institute of Physics in China, Shanghai Institute of Applied Physics and the Electron Microscopy Laboratory at Peking University on the current work. They are now trying to design new catalysts to make tubes with different chiralities. They are also still optimizing the process to improve the selectivity. "We think we can now improve it to 98%," she adds, although this work is still in relatively preliminary stages.

The researchers are also exploring the potential of the approach for bulk synthesis. "The challenge is that we made these nanotubes on a substrate. In bulk the conditions and chemical vapour deposition are different."

Further details are available in Nature .