Orbitals+predicting+oxidation+states

Whenever I think of the octet rule, I think of all of the other rarer oxidation states. Especially troubling are the considerations for transition metals with multiple common oxidation states. So, I loaded up Excel, typed in all of the individual orbitals (each orbital, not level) in the order they're traditionally filled then started a simple exercise: I first estimated the ground state configuration simply by adding one additional electon to the next available spot until each orbital is filled, then I looked at all of the possible cofigurations if I emptied, half filled, or completely filled each nearby orbital. For an atom, this produced 1-6 estimated oxidation states. Checking against known levels, I predicted well more than half. For instance: With Nitrogen I predicted +5, +3, +2, and -2. With Copper, I predicted +6, +2, +1, and -1. With Zinc, I predicted only +2. With Gallium I predicted +3,+2, and +1. With Fluorine I predicted +7,+5,+4,+2,+1, and -1.

It's a little hit and miss, especially for the Lanthenide series, but a simple algorithm did predict some of the less common, non-octet rule oxidation states.

Richard Daines