Researchers watch nanofibers grow.

Carbon nanotubes only a few atoms in diameter could become the basis of important new technologies ranging from molecular-scale transistors to elevators capable of ascending to space. Yet longstanding uncertainties about how nanotubes and other ultrathin carbon fibers grow are slowing efforts to realize such visions (SN: 10/5/02, p. 218). That knowledge gap has also stymied engineers trying to prevent the unwanted formation of nanoscale carbon whiskers in various industrial chemical processes where the fibers inactivate catalysts.

Now, a team of researchers in Denmark has made atomic-scale movies of the formation of carbon nanofibers. This most-detailed-ever look at the growth of such filaments bucks the prevailing wisdom about atomic migrations during the process. It also reveals unexpected behaviors of the metal catalysts, such as nickel, that facilitate fiber development.

In their experiment, described in the Jan. 29 Nature, Stig Helveg of Haldor Topsoe A/S, a catalyst firm in Lyngby, and his coworkers flowed carbon-rich methane gas over nanoscale clusters of nickel atoms on a ceramic base. All this occurred at furnace temperatures within a custom-designed transmission electron microscope.

As the methane molecules disintegrated, the researchers collected evidence that carbon atoms scuttled across a nickel cluster's surface to where the leading edge of a new nanofiber was picking up additional atoms. There, the carbon atoms appeared to displace metal atoms from the surface, leading to remarkable changes in the clusters shape.

As expected, each developing carbon fiber trailed from a single nickel cluster--much as a sock emerges from a set of knitting needles. Startling, however, was the discovery that the nickel nanoclusters undergo wild and rapid pulsations as they fulfill their catalytic duties. In a repetitive cycle, they start as spheres, suddenly morph into tiny pillars, and then convert back into spheres, Helveg says. The morphing apparently results from repeated intracluster migrations of nickel atoms dislodged by fiber growth. Researchers at the Technical University of Denmark, also in Lyngby, calculated certain parameters of the process. Their results contributed to the new picture of fiber formation.

Helveg and his colleagues have provided a "long-awaited solution to the mystery of [carbon] nanofiber growth," says Pulickel M. Ajayan of Rensselaer Polytechnic Institute in Troy, N.Y., in a commentary in the same issue of Nature.

"It was previously accepted that carbon-fiber growth occurs by diffusion of carbon through the nickel particles," notes Abhaya K. Datye of the University of New Mexico in Albuquerque. "This study challenges that model and instead proposes that carbon diffuses over the surface."

The new findings should lead to both improved fiber manufacturing and better catalyst preservation, Helveg says. As they form, carbon nanofibers can assume many different structures, including carbon nanotubes (SN: 11/22/03, p. 324), and researchers now want to learn how widely the new findings apply.