Fullerenes: reactivity

Since efficient syntheses have been available, fullerenes (in particular C₆₀) have been the focus of an explosion of research. We provide a brief introduction to the chemical properties of C₆₀; organometallic derivatives are covered in Section 23.10, and the reading list at the end of the chapter gives more in-depth coverage.

  The structural representation in Figure 1.1b suggests connected benzene rings, but the chemistry of C₆₀ is not reminiscent of benzene. Although C₆₀ exhibits a small degree of aromatic character, its reactions tend to reflect the presence of localized double and single C^_C bonds, e.g. C₆₀ undergoes addition reactions.

Fig. 1.1 (a) The structure of the fullerene C60; the approximately spherical molecule is composed of fused 5- and 6-membered rings of carbon atoms. [X-ray diffraction at 173K of the benzene solvate C60   4C6H6, M.F. Meidine et al. (1992) J. Chem. Soc., Chem. Commun., p. 1534.] (b) A representation of C60, in the same orientation as is shown in (a), but showing only the upper surface and illustrating the localized single and double carbon–carbon bonds.

Birch reduction gives a mixture of polyhydrofullerenes (equation 1.1) with C₆₀H₃₂ being the dominant product; reoxidation occurs with the quinone shown. Reaction 1.2 shows a selective route to C₆₀H₃₆; the hydrogen-transfer agent is 9,10-dihydroanthracene (DHA).

Fig. 1.2 Halogenation reactions of C60. Although the number of possible isomers for products C60Xn where 2 ≤ n ≤ 58 is, at the very least, large, some of the reactions (such as fluorination using NaF and F2) are surprisingly selective.

   In addition to being a selective method of hydrogenation, use of 9,9’,10,10’-[D₄]dihydroanthracene provides a method of selective deuteration.

              (1.1)

          (1.2)

 Additions of F₂, Cl₂ and Br₂ also occur, the degree and selectivity of halogenation depending on conditions (Figure 1.2). Because F atoms are small, addition of F₂ to adjacent C atoms in C₆₀ is possible, e.g. to form 1,2-C₆₀F₂. However, in the addition of Cl₂ or Br₂, the halogen atoms prefer to add to remote C atoms. Thus, in C₆₀Br₈ and in C₆₀Br₂₄ (Figure 1.3a), the Br atoms are in 1,3- or 1,4-positions with respect to each other. Just as going from benzene to cyclohexane causes a change from a planar to boat- or chair-shaped ring, addition of substituents to C₆₀ causes deformation of the near-spherical surface. This is illustrated in Figure 1.3 with the structures of C₆₀Br₂₄ and C60F₁₈. The C₆₀-cage in C₆₀Br₂₄ includes both boat and chair C₆-rings.

Addition of a Br to a C atom causes a change from sp2 to sp3 hybridization. The arrangement of the Br atoms over the surface of the C60 cage is such that they are relatively far apart from each other. In contrast, in C₆₀F₁₈ (Figure 1.3b), the F atoms are in 1,2-positions with respect to each other and the C₆₀-cage suffers severe ‘flattening’ on the side associated with fluorine addition. At the centre of the flattened part of the cage lies a planar, C₆-ring (shown at the centre of the lower part of Figure 1.3b). This ring has equal C–C bond lengths (137 pm) and has aromatic character. It is surrounded by sp3 hybridized C atoms, each of which bears an F atom. The ene-like nature of C₆₀ is reflected in a range of reactions such as the additions of an O atom to give an epoxide (C₆₀O) and of O₃ at 257K to yield an intermediate ozonide (C₆₀O₃). In hydrocarbon solvents, addition occurs at the junction of two 6-membered rings (a 6,6-bond), i.e. at a C=C bond, as shown in scheme 1.1. Loss of O₂ from C₆₀O₃ gives C₆₀O but the structure of this product depends on the reaction conditions. At 296 K, the product is an epoxide with the O bonded across a 6,6-bond. In contrast, photolysis opens the cage and the O atom bridges a 5,6-edge (scheme 1.1).

Fig. 1.3 The structure of C₆₀Br₂₄ determined by X-ray diffraction at 143K [F.N. Tebbe et al. (1992) Science, vol. 256, p. 822]. The introduction of substituents results in deformation of the C₆₀ surface. (b) The structure (X-ray diffraction at 100 K) of C₆₀F₁₈ [I.S. Neretin et al. (2000) Angew. Chem. Int. Ed., vol. 39, p. 3273]. Note that the F atoms are all associated with the ‘flattened’ part of the fullerene cage. Colour code: C, grey; Br, gold; F, green.

                          (1.1)

Other reactions typical of double-bond character include the formation of cycloaddition products (exemplified schematically in equation 1.2), and some have been developed to prepare a range of rather exotic derivatives.

              (1.2)

Reactions of C₆₀ with free radicals readily occur, e.g. photolysis of RSSR produces RS^* which reacts with C₆₀ to give C₆₀SR˙, although this is unstable with respect to regeneration of C₆₀. The stabilities of radical species C₆₀Y˙ are highly dependent on the steric demands of Y. When the reaction of t-Bu^* (produced by photolysis of a tert-butyl halide) with C₆₀ is monitored by ESR spectroscopy (which detects the presence of unpaired electrons), the intensity of the signal due to the radical C₆₀ t-Bu^* increases over the temperature range 300–400 K. These data are consistent with equilibrium 1.3, with reversible formation and cleavage of an inter-cage C^_C bond.

        (1.3)

   The formation of methanofullerenes, C₆₀CR₂, occurs by reaction at either 5,6- or 6,6-edges in C₆₀. For the 6,6- addition products, the product of the reaction of C₆₀ with diphenylazomethane is C₆₁Ph₂ (equation 1.5) and, initially, structural data suggested that the reaction was an example of ‘cage expansion’ with the addition of the CPh₂ unit being concomitant with the cleavage of the C^_C bond marked a in equation 1.5. This conclusion was at odds with NMR spectroscopic data and theoretical calculations, and a low-temperature X-ray diffraction study of compound 1.1 has confirmed that 6,6-edge-bridged methanofullerenes should be described in terms of the C₆₀ cage sharing a common C^_C bond with a cyclopropane ring.

                   (1.5)

(1.1)

Theoretical studies on C₆₀ show that the LUMO is triply degenerate and the HOMO–LUMO  separation is relatively small. It follows that reduction of C₆₀ should be readily achieved. A number of charge transfer complexes have been prepared in which a suitable donor molecule transfers an electron to C₆₀ as in equation 1.6. This particular product is of importance because, on cooling to 16 K, it becomes ferromagnetic (see Figure 20.25).

   (1.6)

The electrochemical reduction of C₆₀ results in the formation of a series of fulleride ions, [C₆₀]^n-  where n = 1–6. The midpoint potentials (obtained using cyclic voltammetry and measured with respect to the ferrocenium/ferrocene couple, Fc⁺‏/Fc = 0 V, ferrocene; see Section 23.13) for the reversible one-electron steps at 213K are given in scheme 1.7.

    (1.7)

By titrating C₆₀ in liquid NH₃ against an Rb/NH₃ solution (see Section 8.6) at 213 K, five successive reduction steps are observed and the [C₆₀]^n- anions have been studied by vibrational and electronic spectroscopies. At low temperatures, some alkali metal fulleride salts of type [M‏]³⁺[C₆₀]^3- become superconducting (see Section 27.4). The structures of the M₃C₆₀ fullerides can be described in terms of M‏⁺ ions occupying the interstitial holes in a lattice composed of close-packed, near-spherical C₆₀ cages. In K₃C₆₀ and Rb₃C₆₀, the [C₆₀]^3-  cages are arranged in a ccp lattice, and the cations fully occupy the octahedral and tetrahedral holes (Figure 1.4). The temperature at which a material becomes superconducting is its critical temperature, Tc. Values of Tc for K₃C₆₀ and Rb₃C₆₀ are 18K and 28K respectively, and for Cs₃C₆₀ (in which the C₆₀ cages adopt a bcc lattice), Tc = 40 K. Although Na₃C₆₀ is structurally related to K₃C₆₀ and Rb₃C₆₀, it is not superconducting. The paramagnetic [C₆₀]^2- anion has been isolated as the [K(crypt-222)]‏ salt (reaction 1.8 and Section 10.8).

                     (1.8)

Fig. 1.4 A representation of the structures of K₃C₆₀ and Rb₃C₆₀ in which the [C₆₀] ^3-  cages are arranged in an fcc lattice with the M‏ ions occupying all the octahedral (grey) and tetrahedral (red) holes. Some of the cations in the unit cell shown are hidden by [C₆₀]^3- anions.

In the solid state, the [C₆₀]^2- cages are arranged in layers with hexagonal packing, although the cages are well separated; [K(crypt-222)]‏ cations reside between the layers of fulleride anions. The coupling of C₆₀ molecules through [2 ‏+ 2] cycloaddition to giveC₁₂₀ (1.2) can be achieved by a solid state reaction that involves high-speed vibration milling of C₆₀ in the presence of catalytic amounts of KCN. When heated at 450K for a short period, the C₁₂₀ molecule dissociates into C₆₀.

(1.2)

    Endohedral metallofullerenes are a remarkable series of compounds in which metal atoms are encapsulated within a fullerene cage; the general family is denoted as Mx@C_n. Examples of these compounds include Sc₂@C₈₄, Y@C₈₂, La₂@C₈₀ and Er@C₆₀. In general, the larger fullerenes produce more stable compounds than C₆₀. The compounds are prepared by vaporizing graphite rods impregnated with an appropriate metal oxide or metal carbide. By use of ¹³C and ¹³⁹La NMR spectroscopies, it has been shown that the two lanthanum atoms in La₂@C₈₀ undergo circular motion within the fullerene cage.