Dear the CI-NEB user:
is it fully reasonable to find the reaction path with multiple saddle points in one CI-NEB run ?
just like the that shown in the following figure attacheched
finding multiple saddle points in one run
Moderator: moderators
finding multiple saddle points in one run
- Attachments
-
- mul-sadle point.jpg (60.88 KiB) Viewed 13824 times
Re: finding multiple saddle points in one run
By default, the climbing image will converge to the highest energy saddle. We recommend breaking up your path into segments to find other saddles. You can also run a min-mode following calculation to rigorously converge to the intermediate saddles along the path.
Re: finding multiple saddle points in one run
thank you profesor! by the way, how can we run "a min-mode following calculation"?
Re: finding multiple saddle points in one run
The dimer min-mode following method is implemented in the VTST code. We have a neb2dim.pl script will will interpolate to the highest point along an NEB calculation and also estimate the minimum mode. You can use the geometry of any local maximum along the path to do a local search for a saddle point using this method.
Re: finding multiple saddle points in one run
thank you professor, thank you for the detailed explation