We demonstrate that a simple Coulomb-energy model can be used to predict th
e vertical electron detachment energy of an anion of charge (-n) given the
detachment energy of the corresponding anion having one less charge (-n + 1
). This model was applied earlier by other workers to:dianions in which the
two charged sites are quite distant. In this paper we show that it can als
o be applied to more spatially compact species as long as the two orbitals
from which the electrons are removed are sufficiently noninteracting. We fi
rst demonstrate how to use this model by applying it to a series of electro
nically stable dianions (MgF42-, BeF42-, TeF82-, SeF82-, and TeCl82-) for w
hich the (-2) to (-1) and (-1) to (neutral) electron detachment energies ha
ve been evaluated using conventional ab initio methods. These test calculat
ions allow us to assess the predictive accuracy of the Coulomb model. We th
en extend the model's use to predict the energies of dianions and trianions
that are not electronically stable (SO42-, CO32-, PO43-, and PO42-) and fo
r which application of conventional quantum chemistry methods will not yiel
d reliable predictions. That is, we predict at what energies metastable res
onance states of these species will occur. Finally, we use the Coulomb natu
re of the long-range part of the electron-anion potential to estimate the l
ifetimes of these resonance slates with respect to electron loss.