My Community

Hi,
I want to calculate the migration energy for an atom moving into the nearest vacant site in a solid. I also need to find the minimum energy path (MEP).

I tried using NEB with full structural relaxation of the images (ISIF=3). I find that it takes awfully long time for the convergence and the structure seems to be distorted significantly along the MEP.

Is it necessary to relax all the cell parameters to obtain correct MEP? Or can one fix the cell shape and volume (ISIF=2)? Which among the two will give the correct energy barrier for migration?

regards
MC

In my opinion, the cell shape and volume should not be changed during the NEB calculations. Because of the elastic forces, neither the force nor the stress tensor is ''real'' for the simulation cell; if you allow a fully structural relaxation, the resultant cell could not be correct.

How does one account for the local structural distortion which would occur during the migration of an atom?

[quote="Mahesh_chandran"]How does one account for the local structural distortion which would occur during the migration of an atom?[/quote]

Keep it local: use a bigger supercell!

Using the NEB method, how does one extract the migration energy from the output files?

Thank you,
torey