Page 1 of 1
A question about NEB
Posted: Wed Oct 18, 2006 12:52 pm
by thshwy
Hello, everyone! I need your help!
When I found the saddle points using the CI-NEB, I found the coordinates in some images go beyond the coordinates of the initial state and the final state after some iterations. I got the images by linear interpolation. I want to know the reason, and how should I do next?
Thanks!
Posted: Wed Oct 18, 2006 7:58 pm
by graeme
Can you post a link to tar file of the run? This is not enough information to figure out what is going on.
Posted: Thu Oct 19, 2006 3:16 am
by graeme
Posted: Thu Oct 19, 2006 3:49 am
by graeme
Well, the calculation doesn't look too bad to me. It's not fully converged, but the pathway is reasonable. Perhaps the dissociation mechanism is not what you expected?
It looks like there are going to be two equivilant saddle point for dissociation/recombination connecting your initial and final states.
There is some discontinutity between the first and second images, which would be better resolved with more images. But it looks like you've captured the saddle point with the climbing image.
It would also be good to put your intial and final state OUTCAR files in the 00 and 06 directories if you want to use our scripts.
A question about NEB
Posted: Thu Oct 19, 2006 7:17 am
by thshwy
Thank you! It means that my settings in INCAR is right, doesn't it? I will try to calculate it with 8 images.
I want to know that the OUTCAR files of IS and FS must be put the directories of 00 and nn+1 when I start with setting EDDIFG = -0.05. I used your scripts.