If you have convergence problems, firstly, you can try to optimize the molecule in gas phase, and then, in a second step, you can use the gas phase .chk file as guess to start the optimization in a solvent by using the IEF-PCM aproach (guess=read). The open-shell structure of the anionic form is difficult to converge in some cases, and quadratic convergence is dissabled in g09; however, you can use some keywords as scf=(cdiis, damp,novaracc) int=ultrafine. If the latter is a option, you must take into account that all your energies must to be computed in the same integration grids to have consistent energies.

Hi all, 

I tried the propositions of Diego, but it was not successful although for the gas phase optimization of anion radical molecule (chloroform or diethylether plus one electron). And the problem consists in the geom optimization. 

However, i tried with the frozen technique in freezing some specific bonds and atoms, and then it worked well. However, i wonder if the electronic and thermal enthalpies values obtained are correct?

Thanks in advanced for all the discussions.