Firefly and PC GAMESS-related discussion club

Learn how to ask questions correctly  
We are NATO-free zone

Re^9: Problems with SA-MCSCF Geometry optimization



sorry for being so late in my reply.
I have used firefly with half success. I mean, the MCSCF+PCM
geometry optimization works but in my case I really think that one needs state-specific gradients since we have a significant
charge migration between the two electronic states.
It this feature planned in the future?

Kind regards,

On Fri Jul 19 '13 7:08pm, Alex Granovsky wrote

>Do you have some news and follow ups on this issue?

>I'm sorry for long delay on my side and for overlooking your question.

>As to equilibrium vs. non-equilibrium solvation, the solvation model
>used by Firefly is the equilibrium solvation. This is why one needs
>to provide static dielectric constant in $PCM.

>The non-equilibrium solvation model is not implemented in Firefly.
>In my opinion, for MCSCF it should be close to state-specific PCM
>but using different epsilon. The difference in implementation
>should be more significant for time-dependent theories like TD-DFT.

>Kind regards,
>On Fri Jun 28 '13 10:32am, lello wrote

>>thanks for your work. I am testing the new build on my system.

>>Meanwhile I'd like to have a comment on two aspects of PCM.
>>The first is the state-specific solvation (i.e. having a reaction
>>field consistent with the density of a specific state.)
>>The second one is about the concepts of equilibrium
>>and non-equilibrium solvation.
>>Is it possible to have something like equilbrium and non-equilibrium
>>solvation models implemented in MCSCF+PCM?
>>Basically in an exact equilibrium solvation model one uses the static dielectric constant in PCM equation whereas in a non-equilibrium
>>solvation one uses the "optical" part of the dielectric constant.

>>I am very interested in this because I am trying to simulate the absorption lineshapes of organic dyes in solutions.


>>On Sun Jun 23 '13 2:52am, Alex Granovsky wrote

>>>thy the build # 7797 which has been posted today.

>>>Please note the following:

>>>1. The default MCSCF's PCM model was redefined to be more variational
>>>and consistent with the way how gradients are computed.

>>>One can resort to the older model by setting


 $pcm imcpcm=0 $end

>>>(the default is imcpcm=3. Setting imcpcm to 1 or 2 is not recommended)

>>>2. Some nasty bugs in computation of required matrix elements were
>>>found and fixed. This means that if one was using PCM with Firefly
>>>v. 8.0.0 RC it is a good idea to redo calculations.

>>>Kind regards,
>>>Alex Granovsky


[ Previous ] [ Next ] [ Index ]           Sun Aug 25 '13 5:32pm
[ Reply ] [ Edit ] [ Delete ]           This message read 724 times