Slater integrals and SOC for system with only t2g d orbitals (l_eff = 1)

Hi everyone,

I would like to calculate the XAS and RIXS L3-edge of a trimer composed by two $4d^2$ ions and one $4d^3$ ion. I considered only pure ionic calculations (no ligands). But the calculations are still too heavy even for XAS.

A trick to reduce the heavy calculations could be to consider only the t2g orbitals and/or project out the empty eg orbitals. Is there a way to do this in Quanty ?

Shall I consider NF = 6 p orbitals + 3 x 6 d orbitals = 24 orbitals? But then how to rewrite the SOC and Slater Integrals F0, F2, F4 (GS) for system with only t2g d orbitals (l_eff = 1) ?

Or shall I still consider NF = 6 p orbitals + 3 x 10 d orbitals = 36 orbitals? And restrict the coulomb interaction to t2g orbitals by using a restricted rotation matrix? For example, Hamiltonian = Chop(Rotate(Chop(Rotate¹⁾,ConjugateTranspose(rotmatKd ))) ? where rotmatKd rotates the 3d states to a (k tau s sz) basis and rotmatKdt2g is defined such that we only keep the orbitals belonging to the t2g irreducible representation, similar to what is explained in quanty.org/documentation/tutorials/nio_crystal_field/xas_l23_partial_excitations.

Thank you very much for your help!! Stefano