I have recently started working with actinide ions, and have therefore started building Hamiltonians directly from the L and S quantum numbers. In this case, the crystal field (CF), in the form of Steven's operators, act on L rather than S. Thus, Easyspin now parametrises the CF using CF_kq constants instead of B_kq that is normally used when Steven's operators act on S.
Is there any way to translate the CF_kq to the B_kq formalism to better compare with literature values for CF parameters?
Assuming that Easyspin uses Wybourne notation for Sys.CF, I tried using table A1 in this paper https://www.sciencedirect.com/science/a ... 492X#s0125 to translate reported Bk_q constants in a Steven's operator formalism to use with Easyspin. Unfortunately, my calculation using the reported parameters is still different than the results reported.
The matrices for the CF parameters Sys.CF1 etc. are implemented in the EasySpin function ham_cf.m. Have a look at the source code of this function here. It looks like these are just Stevens operators, since ham_cf calls stev.
I would second the question above.
Is there a possibility to let stev act on J instead of S and subsiquently calculate CF differently. It seems as L and S are not really suitable to describe 5f spin states correctly as we find significant zero-field contributions and expect a significant contribution from spin-orbit coupling in the actinides. Thus J would be the better quantum number.
Would it maybe be possible to implement a Opt.X setting to choose a different set of Okq or even Wybourne instead of Stevens nomenclature?
I tried to recalculate Wybourne data to Stevens using these parameter: https://www2.cpfs.mpg.de/~rotter/homepa ... de130.html
and subsequently compensate for the (not-suitable) Okq in the second stev step. However, this appears to be not working in my hands.