On Tue May 28 '13 8:08pm, lello wrote
>I agree with your statements. However I have in mind a different
>case for which state specific solvation can be of help such as the
>calculation of electronic absorption bands (I am condidering
>the simulation of the entire vibronic profile).spectroscopy
>In that case one needs the equilibrium geometry of the
>solute in both its ground and excited state and in both states
>the solvent should be "in equilibrium" with the solute, i.e. the reaction field should be state specific.
>I agree with you that increasing the weight of the state of interest
>might be a solution, with the limiting case of a state-specific
>MCSCF. This, I think, would be the best option, if possible.
>However, in my experience it can be quite hard to achieve
>convergence on a SS-MCSCF geometry optimization for an excited state,
>unless one has symmetry labels to help the choice of the reference.
>In those cases where SS-MCSCF won't converge, and I bet there
>are many, SA-MCSCF + state-specific solvation gradients can be
>a good compromise.
I agree there exist situations when this model could be useful.
Moreover, as I mentioned earlier, we will consider its implementation
in the near future. It was in our "todo" list for many years but
was not ever considered to be a high priority.