The effects of the Coulomb repulsion on the superconductivity is described by the Coulomb pseudopotential mu*. The strong Coulomb repulsion is reduced by retardation and screening effects. For conventional superconductors retardation effects are believed to be very important in view of the very different energy scales for the electrons and phonons. An important question for the Fullerenes is then what the relevant energy scales are. If the t1u band width W is the appropriate energy scale, one would have to assume that the retardation effects are small. This would then raise the question if screening effects can reduce mu* sufficiently to allow for the large experimental Tc, or if the superconductivity should be understood in a different framework, e.g., in terms of an electronic mechanism.
We have found that the the t1u band width W~1/2 eV indeed is a reasonable electronic energy scale and that retardation effects on the Coulomb pseudopotential should not be very important for these systems in view of W~w_ph.
Assuming RPA screening, it was found that screening reduces mu* sufficiently to allow for a T_c of the right order of magnitude.
Since correlation effects are believed to be important in these systems, it is not clear that RPA is a good approximation. A quantum Monte-Carlo calculation ( Phys. Rev. Lett. , 83 , 620 (1999). ) for a Hubbard-like model of the static screening showed, however, that RPA is a fairly good approximation until the system is rather close to a Mott transition, while for larger values of U/W RPA breaks down completely. For systems very close to the Mott transition, the screening becomes inefficient and mu* becomes large. This may explain why Tc increases under pressure in Cs3C60 and NH3K3C60 while it decreases in A3C60 (A= K, Rb). In the former systems, being close to a metal-insulator transition, the dominating effect of applying pressure is probably a reduction of mu*, while in the latter systems a reduction of lambda dominates.