In the def2_tzvp and def2_tzvp basis inputs, it says 60 core electrons are replaced by -ul potential. I run FREEZE_CORE : True to get the count down to Xe but my Ir_2 calculation returns 18 alpha, 16 beta electrons contributing 9 & 8 each but 54 + 9 + 8 = 71, not 77. Reading the number, 60 + 9 + 8 = 77 means the 6s shell is being treated as core, right? The rest is okay with me if it wants to use occupied f-shells in the calculation to occupy 6s but I want to know that it’s doing it justifiably (I know it uses this occupancy in 6s because orbital 1 is an s-s bond in fchk prints).
I solved my own problem, I promise somebody’s going to have the same question I did eventually. 60 = half-filled f-shell (so that ml = 0) + Xe. This causes the atom to have no net l, and you can pair the spins evenly I’m sure for s = 0.
If you run frozen core, it occupies starting at 1s, which is equivalent to simulating electrons always in an immediate sphere to the nucleus, which is the point of frozen core. Keep this in mind, but it doesn’t seem immediately wrong since d-orbitals and core will tunnel or remain this way.