"The processes that govern fluid transport in pipes are well understood for diameters in the range of micrometres and above," Yury Gogotsi of Drexel University told nanotechweb.org. "As the diameters diminish [e.g. in the range of a few nanometres], the roles of surface tension and capillarity seem to vary. Thus, the expected promise of carbon nanotubes in technological applications is in urgent need of a well-documented basic understanding of such forces, especially since no consistent experimental data have been collected until recently."

Gogotsi and colleagues forced water into closed nanotubes grown by chemical vapour deposition (CVD) or arc evaporation by treating them in an autoclave at high temperatures and pressures. It's likely that the liquid entered the nanotubes through defects in their walls: the team found it harder to fill the nanotubes grown by arc evaporation, which contained fewer defects than those grown by CVD.

Transmission electron microscopy of the structures revealed a disordered gas/liquid interface at the edge of the water regions. This contrasts with previous studies of water in multiwalled nanotubes with a diameter of more than 10 nm, which showed smooth curved menisci.

The team also used the transmission electron microscope beam to heat up regions of the nanotube that contained water. The water responded less quickly than expected, indicating that its fluidity was lower than that of bulk water.

"The results of our work suggest that when ultra-thin channels like carbon nanotubes contain water, fluid transport is greatly retarded compared to the macroscale," said Gogotsi. "Anchoring of aqueous fluid and slow response to thermal stimuli in thin tubes suggests obstacles in the transport of large volumes of liquids through nanofluidic devices implementing nanotubes or nanopipes."

Gogotsi reckons that the efficient transport of aqueous liquids through nanometre channels might require perfect tubes with no wall defects, which is unlikely to be practically feasible. But nanotubes could come in handy as a way of achieving "fluid release with a molecular level of control" because of the very slow flow rate observed.

"Nanotubes offer a unique opportunity for investigation of physical and chemical processes in confined systems," said Gogotsi. "On this basis, we are developing a research programme that will thoroughly explore the various aspects of phase-interfacing in different nanotube situations. We are studying the behaviour of biofluids, salt solutions and organic liquids inside nanotubes, as well as in situ polymerization and other processes."

The researchers reported their work in Nano Letters.