I don't have experience in performing ROHF-CCSD via MRCC. It is best to send me the corresponding ROHF-CCSD input file, output file, molden file and CCDENSITIES via E-mail (compress them first), and it is better to also send me RHF-CCSD files for comparison. I will check.

Best,

Tian

I would like to analyze some CCSD wavefunctions which come from the MRCC program.
As described in the manual, I load the MOLDEN file, then the CCDENSITIES file with options "1000" and "97".

For RHF reference, this works perfectly fine.

However, in the case of open-shell systems with ROHF orbitals, it seems that Multiwfn does not change the ROHF wavefunction when loading the CCDENSITIES file. For example, the Mayer bond order does not change compared to loading the original MOLDEN file (while for RHF, I can see a clear difference, as I expect.)

I also noticed that in the case of ROHF, the natural orbital occupation numbers do not appear when loading the CCDENSITIES file, even though the message " Done! Basis function information now correspond to natural orbital cases" is shown.

I suspect that the problem is that Multiwfn expects a CCDENSITIES file which contains alpha and beta blocks separately, but in the case of ROHF-CCSD, there are no blocks (that is the CCDENSITIES file looks similar to the RHF case.)

What can I do?

Best,
Zsolt