"We wanted to know if a base growth mechanism is maintained even if a carpet [of nanotubes] of several tens of microns has grown," Martine Mayne-L'Hermite of CEA's Francis Perrin Laboratory told nanotechweb.org. "With this aim in view, and knowing that the growth rate was constant whatever the synthesis duration, we thought that sequential synthesis with different durations for each sequence would reveal the mechanism."

Mayne-L'Hermite and colleagues injected a liquid hydrocarbon – either toluene or benzene – and ferrocene (Fe(C5H5)2), which acted as a source of catalyst, into a reactor in the form of an aerosol. Decomposition of the hydrocarbon created reactive carbon species which deposited as multiwalled carbon nanotubes.

For one sample, the team deposited two layers for ten minutes each from a 5 wt% ferrocene solution in toluene followed by three layers grown for five minutes under the same conditions. The finished product had two thicker layers on the top, with the three thinner layers underneath, next to the substrate.

"The growth of a new layer always occurs directly on the substrate surface under any pre-existing one by lifting it up," said Mayne-L'Hermite. "Carbon species diffuse through the whole carpet thickness and feed the origin of the nanotube growth located on the substrate surface."

The researchers also used benzene enriched with the 13C isotope as a carbon source for some layers. Examining the position of the layers containing the isotope confirmed that later layers were depositing next to the substrate.

"The formation of a layer [of nanotubes] on the substrate surface is stopped only when ferrocene is not fed into the reactor, attesting the important role of continuous feeding of the catalyst," said Mayne-L'Hermite. "The catalyst originating the growth is only acting on the substrate surface."

Now the scientists are focusing their efforts on understanding the role of the catalyst in aligned nanotube growth.

The researchers reported their work in Nano Letters.