Molcas Forum

Sihan Zhang

Member

Registered: 2025-05-15

Posts: 5

Spin-Orbital Coupling in RASSI

Dear Developers,

I have some doubts about how the SOC works in RASSI part. When I try to obtain the correct 2p -> 5d transitions for Ce, I calculated both singlets and triplets, and use SpinOrbit in RASSI to obtain correct split of L2 and L3 edges. For these two, it works fine, but I found that there is another group of peaks between L2 edge and L3 edge. They come from the coupling among singlets (L2 and L3 come from the coupling between singlet and triplet with the same configuration). I want to know why there is a difference between simulation and experiment, because this kind of SOC states cannot be seen.

Later, I also tried ORCA to do the same thing, just to clarify whether this is due to my wrong settings for the calculation. In its results, I cannot see the peaks between. The SOC in ORCA is processed by QDPT, while in OpenMolcas, from this paper Chemical Physics Letters 357 (2002) 230–240, it is done by transformation matrix and Wigner-Eckart theory. I am also curious why the different methods of considering SOC, one from perturbation theory, one from second quantization, will cause this kind of different.