ccsd(t) calculation for singlet o2 (delta and sigma states)

[Florent Louis]

Dear molpro users,

 As a first guess, I want to compute single reference ccsd(t)/avtz for both excited states of singlet molecular oxygen. 

How to distinguish them in each input file if possible in this type of calculation ?

How to describe the wavefunction for each state? 

Thanks for your feedback,

Cheers,

Florent

[jxzou]

Hi,  Florent,

The low-lying excited states of the O2 molecule have significant multiconfigurational characters. It requires linear combinations of (at least) two Slater determinants, in order to correctly describe each state (two degenerate 1delta_g states, and one 1sigma^{+}_{g} state). Single reference methods like, RHF and RHF-based MP2, CCSD(T) are qualitatively wrong for the lowest three singlet states of O2, since the reference wave function RHF contains only one Slater determinant. But the fact is that there exists a second Slater determinant which is equally important to the RHF determinant, so they must be put into the reference wave function together. So, you are supposed to use the multiconfigurational methods or multireference methods, such as CASSCF, CASPT2, NEVPT2. And the minimum active space for O2 is CAS(2e,2o), but one can also use a larger active space (e.g. CAS(4o,6e)) for more accurate results.

One may use UHF-UCCSD(T) as an alternative, but he/she cannot tell what the obtained state stands for, since singlet UHF is symmetry broken.

Best wishes,

Jingxiang