We arrived at this when using the driving approach, after finding two close breaks in energy profile. We started optimizations to TS from structures closest to each of these two breaks and arrived to almost, but not entirely identical TS (bond lengths differing by 10-4 A) with imaginary frequencies differing by 2 cm-1, see files 3drvD_TS1.log and 3drvD_TS2.log. Neither optimization improving to tight and verytight nor scf tightening have helped (not shown in the attached files). When we removed the "nosymm" option, optimizations arrived at "same", but 10-5 a.u. higher energies.
When trying a third optimization to TS from a structure in-between the two driving breaks, we arrived at an energy lower by 7.0e-7 a.u. than the lower of the first two transition states, and a frequency just in the middle of the first two frequencies.
While the third TS looks as the most reasonable one, we wonder why tightening optimization and scf criteria together with calcall option does not
tighten the structure and its energy?
Thank you for any help,
Markéta Munzarová
Use some professianal software instead of Guessian ;-)
Regards,
Bernd
Hello Marketa,
Did you try to change the grid quality? This may be helpful as differences seem to be exceptionally subtle.
Good luck!
Hi Bernd and thank you, Daniel,
increasing the grid was the next option which came to my mind, we will try. Best regards to Bernd, good to see he is still in a good humor.
Marketa
- Badges
- Science topic