“A self-assembly strategy has been adapted by biological systems for billions of years,” Seunghun Hong, who is now at Seoul National University, told nanotechweb.org. “We were fascinated by its efficiency and precision, and thought that if you can programme the substrate so that individual carbon nanotubes can recognize the location on the substrate and self-assemble, you should be able to mass-produce a large number of carbon nanotube-based circuit structures over a short time.”

To achieve the process, Hong and colleagues directly coated substrates with patterns of organic molecules using techniques such as dip-pen nanolithography and microcontact stamping. They created two surface regions - one patterned with polar chemical groups such as amino or carboxyl, and the second coated with non-polar groups such as methyl. They then added a suspension of single-walled carbon nanotubes: the nanotubes were attracted to the polar regions and self-assembled to form pre-designed structures. The self-assembly process usually took less than 10 seconds.

The researchers assembled millions of individual nanotubes on stamp-generated microscale patterns, covering areas of about 1 square cm, on gold. They achieved a yield of more than 90%. For low concentrations of nanotubes (about 0.02 mg per ml), a single nanotube assembled at the centre of each polar molecular pattern, even though there was enough space for more than one tube.

The team also incorporated the process into conventional microfabrication methods. They assembled individual nanotubes between two polar molecular patterns made by stamping on microfabricated gold electrodes. Atomic-force microscopy using a conducting probe showed that the resulting nanotube circuits were able to conduct small currents.

“We now have a tool to commercialize nanotube-based devices that have been considered merely as an academic interest until now,” said Hong. “Personally, I would really like to see high-school students make nanotube devices and study them in their science class using our method. Considering the simplicity of the technique, I believe this is not just a dream.”

Now, the researchers are working on applying the same strategy to other nanoscale wires such as metal oxide nanobelts and silicon nanowires. “In addition, the development of our method has raised many new scientific questions such as the forces between surface organic molecules and nanowires in the solution,” added Hong. “We will also explore explanations for these new issues.”

The researchers reported their work in Nature.