Mechanism of Fullerene Formation

Abstract

In 1985, Kroto et al. made the surprising discovery that C60 was unusually stable among the gas-phase carbon ions produced by laser vaporization of graphite.1 They hypothesized that this stability resulted from its truncated icosahedron structure and dubbed the ion “buckminsterfullerene” after the famous architect. Many gas-phase experiments and theoretical investigations that followed supported this claim, and the soccer-ball structure of C60 was finally confirmed in 1990, when Kratschmer and Huffman discovered a method for making macroscopic quantities of C60 and other larger “fullerenes” via resistive heating of graphite.2 Kratschmer and Huffman’s finding opened up new avenues for research and discovery. In addition, it presented a mechanistic puzzle to physical and organic chemists: How can such highly ordered compounds as the fullerenes form in significant yields in the entropic conditions of graphite vaporization? And why do these conditions produce more C60 than any other all-carbon molecule? Before the bulk isolation of fullerenes, Smalley and co-workers postulated a mechanism which they called the “party line” for fullerene formation in their laservaporization/molecular-beam source.3 In this scenario, small carbon particles would come together to form linear species, which would react with other linear species to make rings. Further addition of small linear chains would increase the size of the rings until they reached the 25-35-atom range. The party line mechanism assumes that, in that size domain, polycyclic networks resembling open graphitic sheets become thermodynamically most favorable. Smalley and co-workers hypothesized that these graphitic sheets are more reactive than rings or linear chains because they have more dangling bonds, and that to minimize the number of dangling bonds, the polycyclic network incorporates some pentagons, causing curvature. Occasionally, one of these cuplike pieces of graphite gathers enough pentagons in the right places to force it to close into a hollow cage, thereby forming a fullerene.3 Smalley and co-workers developed this theory to explain the observation of C60 ions in their initial cluster beam experiments. In these studies, only a small fraction of the carbon vapor became fullerenes, while the rest apparently formed large soot particles. The party line scheme therefore offered an explanation of a rare but noticeable event. Most of the nucleating carbon would, in this scenario, create large spiraling particles where the “growth edge” overshot the opposite edge of the graphite cup, forming a new layer of graphite surrounding it.3 Kroto and McKay extended this picture to account for the presence of polyhedral particles in soot, observed by transmission electron microscopy (TEM).4 But this mechanism does not explain how soluble fullerenes like C60 can form in macroscopic quantities, with yields of 20% and more. In searching for a replacement to the party line, it is helpful to understand some general information about fullerene production. Bulk fullerenes form from graphite vapor, produced via either resistive heating or a carbon arc under helium atmosphere in a pressure range of 100-400 Torr.5 Isotope studies involving vaporization of mixed samples of 12Cand 13C-graphite have shown that the reacting material first breaks down to atomic carbon or small fragments (C2 or C3) before recondensing into fullerenes.6 In addition, production of fullerenes in an electric field leads to their isolation almost entirely from the cathode, rather than the anode, suggesting that cations are important to the reaction.7 Although varying the conditions of bulk production can alter the ratio of isolated C60 to C70 considerably, generally C60 is the dominant product, with fullerenes forming in up to 40% overall yield.8 However, C60 is less thermodynamically stable than the larger fullerenes, according to both theoretical9 and experimental10 evidence.