My concern here is: Should I try the zfs parameters calculation after geometry optimization of the structure or directly use the coordinates from x-ray diffraction data of the single crystal collected at 140K?
Hi, I'd rather suggest calculation on experimental geometry, as you are not interested in energetics. If agreement is achived, than it is enough, if not lattice has to be taken in account.
1. Units. Get sure, that your experimental and calculated values are expressed in same units. Zfs can usually be given in cm-1, MHz or even Gauss.
2. In Orca it is advisable to use either ROKS, or UKS with UNOs, otherwise you will get crap
3. In orca also spin-orbit contribution to zfs is implemented. With UKS and UNOs it seems to yield crap, though my experience is limited to nitroxide biradicals. Try it in ROKS calculation
I have no experience with gaussian at all, so I can't compare them