Mott Transition in Degnerate Hubbard Models: Application to Doped Fullerenes

Olle Gunnarsson (a), Erik Koch (a, b), and Richard M. Martin (a, b)
(a) Max-Planck-Institut fuer Festkoerperforschung, Stuttgart
(b) Department of Physics, University of Illinois, Urbana-Champaign

The Mott-Hubbard transition is studied for a Hubbard model with orbital degeneracy N, using a diffusion Monte Carlo method. Based on general arguments, we conjecture that the Mott-Hubbard transition takes place for U/W ~ sqrt{N}, where U is the Coulomb interaction and W is the band width. This is supported by exact diagonalization and Monte Carlo calculations. Realistic parameters for the doped fullerenes lead to the conclusions that stoichiometric A3C60 (A=K, Rb) are near the Mott-Hubbard transition, in a correlated metallic state.

Physical Review B 54, R11026 (1996).



Back to main page.


Max-Planck Institut für Festkörperforschung
Postfach 800 665 D-70506 Stuttgart