Explaining carbon-cluster magic numbers.

Explaining carbon-cluster magic numbers

When graphite is vaporized by a laser, the liberated carbon atoms are found to be clumped together in remarkably specific numbers: If more than 40 atoms make up a cluster, then an even number will be in the clump, while in smaller clusters, certain "magic numbers'--11, 15, 19 and 23--are most common. For years chemists have debated the origin and structure of these clusters, especially those in the 40-plus range. To explain the high prevalence of C60 clusters, for example, some researchers have proposed a soccer-ball-like structure called buckminsterfullerene (SN: 11/23/85, p.325).

Now two scientists at Oregon State University in Corvallis have come up with a carbon structure that they believe may explain the magic numbers of the smaller clusters. Materials scientist James A. Van Vechten and chemist Douglas A. Keszler also suggest that their new structure could form the basis of a novel and industrially important material that would have all the strength of graphite, but none of its brittleness. Their work will appear in the Sept. 15 PHYSICAL REVIEW B.

Van Vechten and Keszler happened upon the structure when they were analyzing fine "whiskers' of carbon that they had made by "sputtering' or bombarding a graphite target with ions. Transmission electron microscopy revealed that the whisker material is crystalline, but that the carbon atoms in it are arranged in neither a diamond's nor graphite's pattern. Van Vechten says that after spending months thinking up atomic arrangements that could be reconciled with the experimental data, he found only one model that would fit: an 11-atom, paddle-wheel-like structure consisting of two "hub' atoms along an axle surrounded by three paddles, each containing an additional three atoms.

In the midst of the whisker work, Van Vechten realized that this structure could also be "a natural explanation for the strong prominence of 11-atom clusters in laser [vaporization],' since that process is as violent as sputtering. Support for the stability of the structure comes from the recent synthesis of similarly shaped molecules such as propellahexaene (SN: 6/6/87, p.357).

To explain the other carbon-cluster magic numbers, Van Vechten says each addition of four carbon atoms to the 11-atom molecule would allow a stable graphite six-member ring to form along the side of one paddle. The magic number series stops at 23, he believes, because there are only three paddles in the molecule. He also thinks this molecule--unlike the chain and ring structures proposed before to explain small clusters--can account for why 11, 15, 19 and 23 are observed to be magic numbers for neutral and positively charged clusters but not for negatively charged ones.

While Princeton (N.J.) University's Leland Allen says this work is "interesting and impressive,' he and others caution that it is conjectural and that the chains and rings are still very much in the running. Moreover, says Richard E. Smalley at Rice University in Houston, Van Vechten's molecule is much more reactive than the other candidate structures. This property appears to be inconsistent with experiments indicating that the 11-atom clusters are not very reactive.

Van Vechten hopes to use the 11-atom structure as the basis for making a new low-density material that improves upon the properties of graphite. "Graphite is extremely strong, but there are difficulties using it as a structural material because it tends of fracture,' he says.

Van Vechten would like to try to grow carbon crystals in which the 11-atom structure is stacked into a honeycomb pattern, interlocked by carbon chains. This interlocking would prevent the honeycomb planes from slipping past one another, which is at the root of graphite's brittleness. "It looks as though this material would have an order of magnitude higher tensile strength than titanium and about a third the density,' he says.

"Also, the material is clearly a metal and it's nonmagnetic, so it ought to have a superconducting-transition temperature,' says Ven Vechten. "Its features lead one to think that it [the temperature] would be high.' If so, then he says it would have vastly better structural properties than the high-temperature superconducting materials (see p.106) that are getting so much attention now, but that are difficult to form into wires because of their brittleness.

Photo: Two views of the proposed 11-atom paddle-wheel structure.