I was checking out an extended Hückel (non-iterative) code, and realised that for s-block elements it would be impossible to simply use atomic ionisation energies as an approximation to the orbital potential energies (due to the p orbitals being empty). In this case, what values are used?
Please see the following reference:
A. Herman, Modelling Simul. Mater. Sci. Eng. 2004, 12, 21.
It has $H_{vv}$ terms for different orbitals of all elements of the table.