O2 molecule electronic states (CASSCF/CASCI/other FCI solvers)

Greetings,

First of all, this may be a very naïve question and maybe I am missing out something in the documentation.

I am trying to obtain multistate potential energy curves with multireference methods, such as CASSCF, CASCI or SHCI, but I always have a problem involving state degeneracy.

For example, with this input I should obtain a double-degenerate delta singlet state for the oxygen molecule:

from pyscf import gto, scf, mcscf
mol = gto.M(atom="O 0 0 0; O 0 0 1.2", basis="ccpvdz")
mf = scf.RHF(mol).run()
ncas = 8
nelec_cas = 12
mc = mcscf.CASSCF(mf, ncas, nelec_cas)
mc.kernel()
mci = mcscf.CASCI(mf, ncas, nelec_cas)
mci.mo_coeff = mc.mo_coeff
mci.fcisolver.nroots = 2
mci.kernel()

These degenerate states have an energy of -149.71369601 Ha. The problem is that the program is jumping to the next state, at -149.67773309. How could I obtain multistate energies, taking into account degeneracy?

This calculation would not be the real problem, because we know that state is degenerate. But if I want to calculate the reaction path energy for those 2-fold states as they approach an ethylene molecule, they would be split-up and should be explicitely considered.

Thanks in advance! Pablo Ortega

PS Update: I tried the same calculation, but using DMRGCI as fcisolver, and it did work. I would like to have a way to obtain the same result with CASCI anyway.

PS Update 2: I also used SHCI as fcisolver but it didn’t work either.

[portega-usal]

Great, I finally got the degenerate states I needed. We can close this thread now.

Thanks a lot (again) for your valuable help!

[jamesETsmith]

Happy to help!