Classification and Structure of Doped Fullerenes

Fullerene solids are unique insofar as they can be doped in several different ways, including endohedral doping (where the dopant is inside the fullerene shell), substitutional doping (where the dopant is included in the fullerene shell), and the most commonly practiced exohedral doping (where the dopant is outside or between fullerene shells). Doped fullerenes in the crystalline phase are often called fullerides, in contrast to the term fullerite, which refers to the crystalline phase prior to doping. In this chapter, the various approaches to doping are classified and reviewed (§8.1). In many instances, the doping of fullerenes with guest species leads to charge transfer between the guest species and the host, while in clathrate materials, such charge transfer does not occur. Examples of both charge transfer and clathrate fullerene-based compounds are given in this chapter (§8.4). Studies of the structure and properties of the doped molecules and of the materials synthesized from doped fullerenes are expected to be active research fields for at least the near future.

Several stable crystalline phases for exohedrally doped (or intercalated) C60 have been identified. At present there is only scanty structural information available for crystalline phases based on endohedrally doped fullerenes (§8.2) or for substitutional^ doped fullerenes (§8.3), such as the fullerene BC59. Most widely studied are the crystalline phases formed by the intercalation of alkali metals (§8.5), although some structural reports have been given for fullerene-derived crystals with alkaline earth (§8.6) [8.1-3], iodine [8.4,5], and other intercalants (§8.7). In §8.8, we review the relatively sparse information on the doping of C70 and higher-mass fullerenes.

The practical aspects of the preparation of doped fullerene solids are reviewed in Chapter 9 on "Single Crystal and Epitaxial Film Growth." The preparation, extraction, purification, and separation of endohedral metallofullerenes are discussed in §5.4. The electrochemistry of fullerenes and fullerene-derived compounds is reviewed in Chapter 10 on "Fullerene Chemistry."

8.1. Classification of Types of Doping for Fullerenes

In this section the various types of doping are classified from both a structural standpoint [8.6] and an electrical point of view. Regarding the structure, we distinguish the doping according to the location of the dopant. The principal methods of doping include endohedral doping (whereby the dopant goes into the hollow core of the fullerene), substitutional doping (whereby the dopant replaces one or more of the carbon atoms on the shell of the molecule), and exohedral doping (whereby the dopant enters the host crystal structure in interstitial positions of the lattice; this method of doping is often called intercalation). Each of these types of doping is discussed in more depth in the following sections. The dopant can be further classified according to whether charge is transferred upon doping. Since charge transfer can modify the properties of fullerenes in scientifically interesting and practically important ways, there is a large literature on the synthesis, structure, and properties of doped fullerides, especially for charge transfer dopants. Dopants for which there is no charge transfer form clathrate compounds. In a clathrate structure, the fullerenes appear on a sublattice, and other molecules reside either on other sublattices or at random lattice sites.

Each carbon atom in a C60 molecule is in an identical environment, has four valence electrons, and bonds to each of the three nearest-neighbor carbon atoms on the shell of the molecule, as discussed in §3.1. Since all the intramolecular bonding requirements of the carbon atoms are satisfied, it is expected that C60 is a van der Waals insulator (semiconductor) with an energy gap between the occupied and unoccupied states, consistent with the observed electronic structure (see Chapter 12). To make C60 (and also other fullerites) conducting, doping is necessary to provide the charge transfer to move the Fermi level into a band of conducting states.

As mentioned above, crystalline compounds based on C60 can be subdivided into two classes: charge transfer compounds and clathrate compounds. In the first class, or "C60 charge transfer compounds," foreign atoms, e.g., alkali metals or alkaline earths, are diffused into solid C60, donating electrons to the filled shell C60 to form C£0~ molecular anions on which the transferred charge is mainly delocalized over the molecular shell.

The resulting dopant cations, required for charge neutrality, reside in the interstitial voids of the C60 sublattice. Upon doping, Fig. 8.1 shows that the solid either may retain the fee structure of the pristine crystal, or may transform it into a different structure [e.g., body-centered tetragonal (bet) or body-centered cubic (bcc)], because of steric strains introduced by the dopant [8.8].

Considerable research activity on fullerene-based materials has been expended on the study of the M3C60 or M3__cM|tC(50 alkali metal compounds since the discovery in 1991 of moderately high temperature superconductivity (18-33 K) in these compounds [8.9]. The AT&T group has also demonstrated charge transfer in ammoniated alkali metal compounds, but in this case, the NH3 molecules are not thought to contribute directly to the charge

Fig. 8.1. Crystal structures for the alkali metal fullerides (a) undoped fee Cm, (b) MC60, (c) M2C 6o> (d) M3c6„, (e) undoped hypothetical bcc CM, (f) M4C60, and two structures for M6C60, (g) MftQ,, (bcc) for M = K, Rb, Cs, and (h) M6CW (fee), which is appropriate for M = Na [8.7], The large balls denote Cm molecules and the small balls are alkali metal ions. For fee M3C(l0, which has four Cw molecules per cubic unit cell, the M atoms can be on either octahedral or tetrahedral sites. Undoped solid C60 also exhibits the fee crystal structure at room temperature, but in this case all tetrahedral and octahedral sites are unoccupied. For (f) bet M4C60 and (g) bcc M6CM) all the M atoms are on distorted tetrahedral sites. For (h) we see that four Na ions can occupy an octahedral site of this fee lattice. The M,^,, compounds (for M = Na, K, Rb, Cs) crystallize in the rock-salt structure shown in (b).

Fig. 8.1. Crystal structures for the alkali metal fullerides (a) undoped fee Cm, (b) MC60, (c) M2C 6o> (d) M3c6„, (e) undoped hypothetical bcc CM, (f) M4C60, and two structures for M6C60, (g) MftQ,, (bcc) for M = K, Rb, Cs, and (h) M6CW (fee), which is appropriate for M = Na [8.7], The large balls denote Cm molecules and the small balls are alkali metal ions. For fee M3C(l0, which has four Cw molecules per cubic unit cell, the M atoms can be on either octahedral or tetrahedral sites. Undoped solid C60 also exhibits the fee crystal structure at room temperature, but in this case all tetrahedral and octahedral sites are unoccupied. For (f) bet M4C60 and (g) bcc M6CM) all the M atoms are on distorted tetrahedral sites. For (h) we see that four Na ions can occupy an octahedral site of this fee lattice. The M,^,, compounds (for M = Na, K, Rb, Cs) crystallize in the rock-salt structure shown in (b).

transfer process. These ammoniated compounds have attracted much attention through their enhancement of the superconducting transition temperature Tc. This enhancement of Tc has been explained by the diffusion of NH3 into M3^M;C60 (M, M' = alkali metal dopants) to increase the lattice constant of the crystalline solid [8.9,10] (see §8.5.5 and §15.2).

In the second class of C60-based materials, or "C60 clathrate solids," the structure is stabilized by the van der Waals interaction between C60 molecules and other molecular species, such as a solvent molecule (e.g., hexane, CS2), 02 [8.11] or S8 [8.12], or mixtures of a solvent molecule and a third species (e.g., C60S8CS2 [8.13]). Since, by definition, no charge is transferred between the C60 molecules and the other molecular species in the structure, the basic physical properties of the clathrate compounds should be predictable from the properties of the constituent molecular species. One can also include in the clathrate class molecular mixtures of C60 adducts (see Chapter 10) and other molecular species. Several fullerene clathrates have been found to exhibit structural order in both the C60 and solvent sublattice, although the fullerene shells display orientational disorder [8.14-17],

Returning to the charge transfer dopants, the electrons transferred to the fullerenes are delocalized on the shell of the fullerene anions, and the distinction between double and single bonds becomes less important as charge transfer proceeds. This effect is supported by both experimental observations and theoretical calculations [8.18,19] which show that the bond lengths for the single bonds along the pentagon edges decrease from a5 = 1.46 A upon charge transfer to form a C^" anion, while the bond lengths for the double bonds between adjacent hexagons increase from a6 = 1.40 A for neutral C60, as shown in Fig. 8.2. The high degree of charge transfer observed for some doped C60 compounds (such as those

0 0

Post a comment