Ion-Beam Chemistry of Fullerenes.

Goals: Primary interest is in comparing the energetics of the three different types of binding sites (surface, substitutional, and endo-hedral) and in exploring the mechanisms and scattering dynamics for reactive and non-reactive collisions with these large and very stable molecules.

Techniques: Use of a unique triple sector guided ion beam instrument and beam scattering techniques.

Example Results

We have looked at reactions of alkali ions, rare gas ions, most first row atomic ions, and some transition metal ions with C₆₀ and C₇₀. Quite a lot is learned about the physics and chemistry of bonding, cage penetration, cage disruption, and collisional energy transfer. These results directly impact on current efforts in the US and Germany to prepare endohedral fullerene compounds by ion implantation. For example, the Campbell group has found that up to 50% incorporation levels can be achieved by bombarding C₆₀ films with alkali ions, leading to M@C₆₀ ("@" means that the M atom is inside the C₆₀ cage). The energy dependence of this process exactly matches what we observed 5 years earlier for collisions of M⁺ with C₆₀ in the gas phase.

Shown below is a plot of the collision energy dependence of the total fragmentation and endo-complex formation cross sections for all the rare gases. Argon and Krypton penetrate the most efficiently because they have enough mass to blast some carbon atoms out of the way, but also aren't too large. Similar trends are observed for alkali ions.

The plot below shows the cross sections for all products found in collisions of Mn⁺ with C₆₀. Charge transfer and fragmentation are the dominant product channels. One very interesting channel is MnC₆₀⁺. There are two components. At low energies we are forming a coordination complex with binding energy of about 2 eV. This disappears as the energy is increased because it is too weakly bound to be detected except at low excess energies. At around 15 eV, the MnC₆₀⁺ signal comes back, indicating formation of a different compound that is much more strongly bound. The binding energy can be estimated to be about 6 eV from the disappearance at higher collision energies. This chemically bound compound appears to be a metal-carbide, probably with the Mn bound to two carbon atoms in the fullerene cage. It decomposes primarily by MnC₂ loss. Similar results are found for Fe, Cr, Mo, and W.

FIGURE 7. Typical data. Cross sections for scattering of Mn⁺ from C₆₀. Note formation of an MnC₆₀⁺ adduct in two collision energy ranges. The low energy adduct is a coordination compound bound by about 2 eV. The high energy adduct is a metal-dicarbide bound by about 6.2 eV