pre-convergence before Cl-NEB

[kambiz]

Hi,

I just have a dummy question. I have three options to run a Cl-NEB

1. Run a NEB with force criterion of 0.4 - 0.5 eV/atom and then run a Cl-NEB
2. Run a NEB with force criterion of 0.05 - 0.1 eV/atom and then run a Cl-NEB
3. Run a Cl-NEB from scratch

My force criterion for Cl-NEB is 0.05 eV/atom. I want to know which way is better way and why.

I d really appreciate if someone help me with that


Kind regards

[graeme]

My vote is 3. The CI should only help with convergence. If you see any counter example (NEB then CI-NEB) I would be interested to see it.

If you are starting from a very high-energy path, you could try Wales' double nudging, to maintain a straighter path during convergence. But again, I don't think there is any down-side to using the climbing image as well.

[kambiz]

Thanks dear Prof. for your reply,

But, is n't the option 2 a faster way? We can decrease the forces acting on images to a specified value using regular NEB. Then we can run our Cl-NEB. I dont think this method, technically,is wrong. Could you please enlighten me about this matter?

I m conducting some examples and I ll let you know about the results.

Thanks

Kind regards