You can do COSMO with TDDFT in NWCHEM

You can do it in Turbomole

I tried with Turbomole, the result is: "Sorry, COSMO not implemented yet", for the optimization of the 1st triplet excited state.

And Gaussian and NWCHEM didn't converge the optimization.

Did you get any structures from the calculations or did the wavefunction fail to converge? Your project is challenging so expect failures in the beginning.

SCF converges. The problem is that the geometry optimization is oscillating. I tried to change all the things that are possible in optimization parameters, scf, td, solvent, etc using G09. But the same happens with NWCHEM.

I'm running a rotational barrier in excited state (trans-cis), and for all the points in the surface I have this problem.

Without solvent effects I runned with Turbomole and I didn't have problems. But with Gaussian even without solvent no way to converge the optimization.

Thank you for your attention.

My calculations of relaxed rotational barriers for the excited state, including solvation, finally converged using the ADF computational package.

The problems of convergence in several packages exist because at some points in my rotational barrier the ground state is a Transition State. The calculations in these programs finished with different kinds of error, due the instability of the reference state.

If you need to optimize an excited state in which the reference state is a Transition State (Saddle point), I suggest the ADF package (Amsterdam Density Functional).

Thank you for your attention.

You can use Gamess US, but you have to use PCM. For some reasons the COSMO implementation is not working properly in the excited state optimization