ricc2 and cosmo solvation model

[oliver]

Dear Turbomole Team,

I'm interested in calculations of excitation energies in different solvents
and my question is as follows: is it possible to perform RI-CC2 calculations
employing a COSMO solvatation model in Turbomole 5.9 version ?
Many thanks in advance.

Oliver

[christof.haettig]

Hi Oliver,

the answer is no. Sorry...

Christof

[Jonas]

Dear all,

shouldn't it be possible to address for bulk solvation effects via COSMO at the SCF stage of a calculation, and then use these SCF orbitals in a subsequent CC2 calculation?
Then, COSMO should enter CC2 via the orbitals?
I just set up a small number of CIS and CIS(D) calculations like this, which actually use the ricc2 module. Although I didn't have any output so far, it appears to work correctly so far ...?

However, maybe I'm confusing something and am wrong with that ... Can somebody please comment on this?

Best regards, Jonas

[Arnim]

Hi Jonas,

it is indeed possible to run CC2 on HF/COSMO orbitals. The shift will be small, however,
and I am not sure, if this shift will always go in the right direction.
You will neglect the COSMO contributions to the correlated ground state energy as well as
the contributions to the excitation energy. Since there is not much known about these
contributions, it is not clear, how much you are neglecting.

A first step to include some solvation effect, that is maybe a bit more safe, could be to
optimize ground state geometry on DFT/COSMO or MP2/COSMO level, instead of using
gas phase geometries.

Cheers,
Arnim

[Jonas]

Arnim,

Ok, this sounds reasonable to me ... it does not solve my problem, but this cannot be done in a minute anyway. 
Thanks for the education, 

Best regards, Jonas