I am trying to run a EOM-EA-CCSD calculation on the triphenylene anion which has a C2v geometry. Q-Chem does not recognize this symmetry at its default symmetry tolerance settings so I have tried lowering it.
Turns out it would recognize C2v, at the very beginning of the calculations where qchem reads the molecular point group and largest abelian subgroup, only if I set SYM_TOL = 0. But the EOM-EA still wouldn’t go through properly. When I get to the part where qchem solves for transition energies (where qchem prints “solving for EOMEA_CCSD/MP2 [state] transitions.”), qchem fails to recognize the EA_STATES that I had specified. Q-Chem prints instead “solving for EOMEA_CCSD/MP2 ? transitions”.
It probably has to do with the fact that qchem could not determine the irrep of some of the orbitals. But I want to know how to fix this issue.
Probably you need to actually symmetrize the molecule, rather than relying on turning the symmetry tolerance way down (to the breaking point). See comments in this post for how to do that, e.g., using IQmol or via geometry optimization:
That worked for me. I had to use Avogadro to symmetrize the molecule because IQmol didn’t recognize C2v at all. It was giving either D3h or C1. But symmetrizing it worked for Q-Chem eventually. Thank you from the Mabbs Lab.
Yeah. It appears that the symmetry detection in IQmol will detect C2v correctly but only if you selected the correct tolerance threshold on the first time you hit “apply” otherwise the program detects another symmetry AND changes the geometry (the coordinates) according to the wrongly identified symmetry, which you wouldn’t want before you get the program’s symmetry recognition tool to lock on to the correct symmetry. I believe Avogadro doesn’t change the coordinates when adjusting the symmetry tolerance variable.
