Dear Joan,
this is a very late reply. But I would stil like to answer your questions. In the two-component implementation in ridft (DOI: 10.1021/acs.jctc.1c01027), we can only partition the tensor into the scalar-relativistic contribution and the spin-orbit part. The latter corresponds to the paramagnetic spin-orbit term, which can also be calculated via perturbation theory in mpshift (DOI: 10.1021/acs.jpca.2c03579) in a non-relativistic framework or the X2C framework. The scalar-relativistic part describes both the FC and the SD term. Unfortunately, these are coupled in X2C, see DOI: 10.1021/acs.jpca.1c07793. The FC term is mainly the result on the diagonal of the scalar-relativistic tensor. However, the FC+SD can only be fully decoupled in the non-relativistic limit. Then, you can separate them by integration by parts after summing the Dirac small-large and large-small part.
If you have access to V7.7, you can also use mpshift for the EPR properties. I have ported all EPR capabilities to this module and added some features for user-friendliness.
Best and happy holidays,
Yannick