how a soccer-ball-shaped molecule was discovered

The Mechanism

The five were *Harold Walter Kroto* (born Wisbech, England, *7 October 1939*; died Lewes, Sussex, *30 April 2016*, aged 76), professor of chemistry at the University of Sussex in Brighton; *Robert Floyd Curl Jr.* (born Alice, Texas, *23 August 1933*; died Houston, *3 July 2022*, aged 88), professor of chemistry at Rice University; *Richard Errett Smalley* (born Akron, Ohio, *6 June 1943*; died Houston, *28 October 2005*, aged 62), professor of chemistry and physics at Rice University; and the two graduate students *James R. Heath* and *Sean C. O'Brien*, plus the postdoctoral researcher *Yuan Liu*. The experiment took place between *1 September and 11 September 1985* on the second floor of the *Space Sciences Building* at *Rice University* in Houston, Texas, using a custom-built molecular-beam cluster-source apparatus known in the laboratory as *AP2*. The AP2 apparatus had been built by Smalley's group over the previous five years to study small clusters of metal atoms produced by *laser vaporisation*: a high-power pulsed Nd:YAG laser would strike a rotating disc of solid metal inside a vacuum chamber, vaporising a thin layer of the metal into a plasma; a stream of cold helium gas would then sweep the plasma through a small nozzle into a low-pressure chamber where the atoms would cool and aggregate into clusters; the resulting cluster beam would be detected by *time-of-flight mass spectrometry*. By 1985, Smalley's group had used the apparatus to characterise clusters of more than a dozen different metals, work that established Smalley as the leading cluster-chemistry experimentalist in the United States. *Harold Kroto* had been corresponding with Smalley and Curl from Sussex throughout the early 1980s. Kroto's interest was different: he had been studying *long carbon-chain molecules* — molecular acetylene chains of the form HC≡C–C≡C–C≡C–H, denoted *cyanopolyynes* when terminated in a CN group — that he had detected by *radio astronomy* in the cold dense interstellar clouds around the carbon-rich red-giant star *IRC+10216*. The chains he had seen in space were as long as *eleven carbon atoms* (HC₁₁N), making them the longest molecules then known in the interstellar medium. Kroto wanted to know how these chains formed. His hypothesis was that they were assembled in the *atmospheres of carbon-rich red-giant stars*, where the temperature is high enough to vaporise solid carbon and cool enough that the vapour can condense back into long molecules. The AP2 apparatus, with graphite as the target instead of metal, would simulate exactly that chemistry. Kroto flew to Houston in late August 1985 and arrived at Rice on *1 September 1985*. The graphite-target experiments began the next day, with Kroto, Curl, Smalley, Heath, O'Brien, and Liu working in shifts in the AP2 lab. The mass spectrum produced by the laser vaporisation of graphite was *crowded* — a forest of peaks corresponding to all sizes of carbon cluster from C₂ to C₁₀₀ and beyond — but two peaks stood out as anomalously tall: one at *720 atomic mass units* (mass of exactly 60 carbon atoms) and a smaller secondary peak at *840 atomic mass units* (exactly 70 carbon atoms). The C₆₀ peak was *roughly four times taller* than its nearest neighbours. The first hypothesis was that 60 atoms was simply a *kinetically favoured size* — that the vapour-condensation dynamics happened to spend more time at that cluster size than at others. But when the team tuned the helium-flow conditions to favour the formation of larger clusters, the C₆₀ peak grew *more* dominant, eventually accounting for most of the total ion current above 500 amu. That ruled out a kinetic explanation: under conditions of higher-temperature longer-time vapour annealing, the C₆₀ cluster became progressively *thermodynamically* favoured over all other cluster sizes. There had to be something about exactly sixty carbon atoms that made them unusually stable. By the *fifth day*, the five had eliminated every two-dimensional graphene-sheet structure as the source of the stability (a flat sheet of 60 carbon atoms would have a large number of reactive edge atoms, which would make it less stable than a sheet of 80 or 100 atoms with proportionally fewer edge atoms). The remaining possibility was that the 60-atom cluster was a *closed-cage structure* — a three-dimensional shell with no edges. Smalley spent the evening of *9 September 1985* at his kitchen table at home in Houston, cutting paper hexagons out of construction paper and attempting to assemble them into a closed cage. He could not make a closed cage out of hexagons alone — sixty hexagons would tile a flat plane, not a sphere. He gave up and went to bed. The next morning, Kroto and Curl recalled that the architect *Buckminster Fuller* had built the geodesic dome of the United States pavilion at *Expo 67 in Montreal* — a 76-metre-diameter spherical shell — using a mixture of hexagonal and pentagonal panels. Smalley returned to his kitchen table and tried adding pentagons to his model. With *12 pentagons and 20 hexagons*, the structure closed into a sphere with 60 vertices, 90 edges, 32 faces — the exact topology of a soccer ball, and the exact shape topologists had been calling a *truncated icosahedron* since Archimedes. The geometric closure was unmistakable. On *11 September 1985*, the five wrote up the result in a *242-word note* and submitted it to *Nature*. Kroto proposed the name *buckminsterfullerene* in honour of Buckminster Fuller's geodesic-dome architecture. The paper, *"C₆₀: Buckminsterfullerene,"* by Kroto, Heath, O'Brien, Curl, and Smalley, appeared in *Nature* on *14 November 1985*, vol. 318, pp. 162-163, *eight weeks after submission*. The structural inference was based on geometric reasoning and the mass-spectrum data alone; there was no NMR data, no X-ray crystallography data, and no microscopy data — those would not arrive until 1990, when *Wolfgang Krätschmer* in Heidelberg and *Donald Huffman* in Tucson independently developed a *graphite-arc-discharge method* that produced macroscopic quantities of solid C₆₀ that could be crystallised. The Krätschmer-Huffman synthesis opened the field of *fullerene chemistry*: by 1991 fullerenes had been isolated in milligram-to-gram quantities, the C₇₀ peak had been confirmed as a *rugby-ball-shaped* fullerene with 12 pentagons and 25 hexagons, the larger fullerenes C₇₆, C₈₄, and so on had been characterised, and the related family of *carbon nanotubes* — long cylindrical fullerene-related structures — had been discovered by *Sumio Iijima* at NEC in 1991. The *1996 Nobel Prize in Chemistry* was awarded jointly to *Kroto, Curl, and Smalley* "for their discovery of fullerenes." Graduate students *Heath* and *O'Brien* were not on the citation despite both being first authors on follow-up papers; Heath went on to a distinguished career in nanotechnology at Caltech and is now President of the Institute for Systems Biology in Seattle. The molecule *C₆₀ itself* — sixty carbon atoms on the vertices of a truncated icosahedron, 90 sigma bonds along the edges, plus a delocalised pi-electron system across all 60 atoms — is now one of the most extensively-studied molecules in chemistry: the *third allotrope of elemental carbon* after diamond and graphite, naturally occurring in trace amounts in *shungite* mineral deposits and in *certain meteorites*, detected in 2010 by NASA's *Spitzer Space Telescope* in the interstellar medium around the planetary nebula *Tc 1*, the most complex molecule ever identified in space.