In minerals electrons travel back and forth between ions in adjacent sites. In this case crystal field theory (splitting of energy levels in 3d orbitals causing colour) does not apply.
I'm sure the 3d orbitals are still differentiated in a crystal regardless of whether or not the electrons move around in the orbitals. I thought the molecular orbital theory was just another way of visualizing electron configuration. If a crystal absorbs energy wouldn't the electrons in the 3d orbitals absorb and possibly release energy even while molecular orbital transitions occur?
In metals and ionic compounds the electrons are not localized. Therefore, it is not right to think in terms of separate orbitals.
The magnitude of a crystal also causes that there are a large quantity of energy states available ("bands"), with the difference between what in traditional orbital reasoning would be the HOMO-LUMO gap now being defined as the bandgap. In conductors this "gap" is negligent, allowing electrons to move freely; in semi-conductors this bandgap is substantial, but can be overcome by influncing temperature/doping etc.; in insulators this bandgap is too large, stopping any free transport of electrons.
Also where does the energy come from to separate an electron from Fe to another Fe? Light?
for example, but do not think of it in the way that one Fe would now be positively charged, with the other negative. What happens is that there may be an exchange in electrons (swapping) or a transport (for example) from a source at one end of an iron wire to the other.