Orbital occupation numbers, transition selection

In Tables XVI and XVII we have collected the natural orbital occupation numbers from different CASSCF calculations on a selection of octahedral and tetrahedral molecules of first-row transition metals. In all cases a CASSCF calculation was performed using a basic active space of 10 orbitals. For some of the molecules additional calculations with larger active spaces (up to 14 orbitals) are also presented. Only ground-state results are presented, except for the molecules CrFg , Cr(CN)g , CoFg, and Co(CN)g , for which we have included a selected number of excited states. 

Atomic ionization potentials I which define atomic orbital levels Ufifi (equation 9) are taken from experimental atomic spectra. We have selected atomic configurations with occupation numbers s and m, which are different from Zemer s suggestions for some transition metals. The actual values for /Is and /Id are given in Table 3. Consequently we have also taken the corresponding set of U, fp, and /d in the.se cases (Table 4). Orbital energies Sp and exponents Xp used in the pseudopotential terms in equations (8), (10), and (27) are taken from ab initio calculations. In some cases the initial values have been modified in order to improve the accuracy (Table 4). Slater-Condon parameters F and G (Table 5) are taken from Bacon and Zemer, ... 

Furthermore, the change in L can come about only if the occupancy change corresponds to a change of 1 in the I quantum number of one and only one electron. That is, A/ = 1 while = 0. For example, a change from the occupancy Is 2s 2p to the occupancy Is 2s 2p is a change in one and only one of the electrons / values of 1. The term symbol for the first occupancy is P, and so transitions are allowed to term symbol states arising from the second occupancy, but only if they are or D. When we consider the fine structure of the spectra, which means the small energy differences due to spin-orbit interaction, a selection on / applies ... 

---