Hi Steve:
I was checking through the archives of the LAMMPS mailing list, and came across a post by Aidan Thompson 2 years back about polarizable core/shell models. Basically, these are Buckingham potentials + a Coulombic term in which each atom has 2 degrees of freedom instead of 1, i.e. a core and a shell. LAMMPS has the right potential (Coulomb + Buckingham), but I haven't seen any documentation where this can be used for the polarizable core/shell, while also accounting for the harmonic potential that governs interactions between each core and shell.
Thanks,
Harold