About VASP setting in NEB and Dimer method

[cwandtj]

Hi!

I have a question about EDIFF and EDIFFG for VASP setting.

I actually used following setting for my calculations.

System: LaFeO3

EDIFF: 1E-07
EDIFFG: -0.0005

These were OK but for surface calculations, I can not have convergence with these setting.

How you all set EDIFF and EDIFFG for your surface calculations?

Another question is I try to see what will be the optimum no. of layer for my system. 5L is OK. But If I extend my system as 7L and 9L, no convergence. I try to add dipole correction as my system has tilted octahedrals and I can not have "exact" symmetry even though I have identical surface termination.

Please share your ideas and experince.

Thanks,

Chan-Woo

[graeme]

These questions are probably more appropriate for the general vasp forum because they are not specific to finding transition state.
See: http://cms.mpi.univie.ac.at/vasp-forum/forum.php

Anyways, your ediffg is very low. A strict force criteria is -0.001, and many people would use -0.01 or higher. If you look at the change in the energy as you go from a force of 0.01 eV/Ang to 0.001 eV/Ang, it is usually negligible.

Your results should be converged with respect to the number of layers. If you are interested in a barrier, start with a few layers, then add subsequent layers until the barrier stops changing.

I don't have any experience with your system. Try the general vasp forum to see if others have more specific advice.

[cwandtj]

Actually, my final goal is to see how oxygen molecules dissociate on surfaces and how oxygen adatom will diffuse into the surfaces...... That is why I believed that this is the place. I thought that I need to know how other people who use NEB, Dimer methods, etc set their EDIFF and EDIFFG for thier research.

Anyway, as you said, there is noting about transition state. :) Sorry about that.

Your advice and explanation is truly helpful.

I appreciate your help.

Chan-Woo