Section V
Confirmation of the Structure of C


  1. Nuclear Magnetic Resonance (NMR) Spectra

    Note: Background material on NMR Spectroscopy should be presented in this section.

    The four-line IR spectrum for C, as reported by Krätschmer et al. (Krätschmer & Lamb, 1990), supported the proposed truncated icosahedron structure. However, even more convincing was the C NMR spectrum of the purified C, reported by Kroto et al. (Taylor, 1990). The NMR spectrum contained a single peak at 142.7, as expected for the highly symmetrical truncated icosahedron structure in which all carbons are identical (see Figure V.A). This result eliminated planar graphite fragments and fullerenes of lower symmetry as possible structures for C. A sixty-membered polyalkyne ring would also be expected to exhibit one C NMR signal but the observed chemical shift position (142.7) was inconsistent with this possibility. (Alkyne carbons generally resonate between 50 and 100.)

    The C NMR spectrum of purified C was also reported by Kroto and, as expected, it contained five peaks (Figure V.A). The proposed football-shaped C fullerene possesses five sets of inequivalent carbon atoms in a ratio of 10:10:20:20:10. This is precisely the ratio of the line intensities observed in the C NMR spectrum.

    Figure V.A:
    Idealized C NMR spectra and structural drawings of C(top) and C (bottom). In C, all carbon atoms are identical and a single C NMR peak is observed. In C, there are five sets of inequivalent carbon atoms (labelled a-e), giving rise to five C NMR signals.


  2. Single Crystal X-Ray Structure Determinations

    Note: Background material on X-ray crystallography should be presented in this section.

    Definitive proof for the structure of C came in 1991 when Joel Hawkins (Univ. of California-Berkeley) synthesized and crystallized an osmium derivative of C, (C)OsO(4-tert-butylpyridine) (Hawkins, 1991). Single crystal X-ray analysis of this compound yielded the first atomic-resolution picture of the carbon framework of C (Figure V.B.1). (Note: Because of its high symmetry, C itself is orientationally disordered in the solid state. In fact, at ambient temperature, the C molecules rotate rapidly in the solid state. Derivatization with OsO breaks the nearly spherical symmetry of C, allowing it to crystallize with orientational order.)

    Figure V.B.1:
    X-ray crystal structure of (C)OsO(4-tert-butylpyridine)
    Note: The carbon atoms are green, hydrogen atoms are white, nitrogen atoms are dark blue, osmium atom is magenta, and oxygen atoms are red in the ball-and-stick representation.

    Osmylation occurs selectively across a 6-6 ring junction rather than a 6-5 ring junction. Not surprisingly, bond distances in the osmylated portion of C are dramatically affected (C1-C2, C1-C3, C1-C4, C2-C5, and C2-C6 bond lengths average 1.55Å, comparable to normal C-C single bonds). However, the remainder of the C structure is not significantly perturbed by the OsO unit. Interestingly, there are statistically-different average bond lengths for the 6-6 and 6-5 ring fusions. Excluding bonds to C1 and C2 (the osmylated carbons), the average C-C bond lengths are 1.386(9) Å for 6-6 ring fusions and 1.434 Å for 6-5 ring fusions. (Note: these average bond lengths have subsequently been confirmed in low-temperature neutron powder diffraction and gas-phase electron diffraction studies of C itself.) Hence, the bonding in C is not completely delocalized as it is in graphite. Rather, the dominant resonance structure (see Figure V.B.2) is one in which the double bonds are located "exocyclic" to the five-membered rings and between the six-membered rings. This explains the regiochemistry of the osmylation reaction; it occurs preferentially on the more electron-rich 6-6 ring fusion.

    Figure V.B.2:
    Dominant resonance structure for C, in which formal double bonds are located at 6-6 ring fusions.

    Structural proof for C also came in 1991 when Balch and coworkers succeeded in crystallizing and obtaining the single crystal X-ray structure of (C)Ir(CO)(Cl)(PPh) (see Figure V.B.3) (Balch, 1991, JACS). This adduct is particularly interesting from the point of view of regiochemistry because there are four types of 6-6 ring fusions in C as well as four 6-5 fusions. The metal selectively positions itself across one of the 6-6 fusions near the elongated end of the football-shaped molecule.

    Figure V.B.3:
    X-ray crystal structure of (C)Ir(CO)(Cl)(PPh).
    Note: The carbon atoms are green, chlorine atom is yellow, hydrogen atoms are white, iridium atom is light blue, oxygen atom is red, and phosphorus atoms are magneta in the ball-and-stick representation.

    In general, metal binding is accompanied by local distortion so that the two carbon atoms involved in the coordination are pulled out from the fullerene surface. Simple geometric considerations show that the 6-6 ring fusion at the elongated end is the most accessible. The other 6-6 ring fusions all have a more flattened local structure and would require much larger distortions to accommodate metal coordination.


Return to Fullerene Science Module Table of Contents

[To Fullerene Science Module Page]