NEB Settings for materials with complex potential energy surfaces.

[philippedavid]

Hello Henkelman Group,

I'm wondering what NEB setting were used in your paper studying the perovskite to post-perovskite phase transformation in CaIrO3. I'm alsocurious if there there any difficulties as I'm having trouble getting convergence studying a solid-solid phase transformation in a similar material. Is a good NEB strategy for dealing with similar materials with complex potential energy surfaces?

Thank you

[graeme]

I'm attaching an example ssneb calculation for the perovskite work that you mentioned. Any detailed questions should be directed to the primary author of that work, Penghao Xiao.

Your second question is a little harder to answer. The NEB is a good method to find the reaction pathway and barrier for a reaction with a known initial and final state on smooth energy landscapes. By smooth, I mean reactions that are dominated by one or a few enthalpy barriers. For rough landscapes with many possible pathways between an initial and final state, sampling methods are more appropriate.

So the issue is about complexity. If the potential is smooth but the reaction is complex in terms of (for example) the number of atoms moving - that is no problem for the NEB. If by complex you mean rough, then finding any one path with the NEB may not adequately quantify the reaction pathway(s) or rate.

One more point about complexity: a key challenge with the ssneb is determining the labeling of the atoms in the final state as compared to the initial state. In many cases, we assume some ordering, but then if the reaction takes place in a way that is different than expected, it may be necessary to change the ordering. Penghao would often make use of the ssdimer method, in which you do not have to specify the final state. Then, once you know the mechanism, the ssneb can be used to fill in resolution of the reaction mechanism.