spikes in effective potential in slab simulations using MGGA

[Dhirendra]

Hi,
In the effective potential, why the spikes are observed near the edges of the slab (vacuum/material interface) with MGGA functional. Is it fine to observe such spikes?

[Jess Wellendorff]

Do you find the same in a GGA calculation?

[Anders Blom]

If you have vacuum in the system you should fix the "c" parameter in MGGA, and not let it adjust self-consistently. First determine it from a bulk calculation of the same material, then use that value for the confined system

[Dhirendra]

Yes Anders. I am fixing the c parameter calibrated to provide the expected bandstructure.

[Anders Blom]

And you have passivated the surface?
MGGA has a problem with vacuum, so one can perhaps see these spikes, but hopefully it doesn't influence the electron density too much.

[Dhirendra]

Yes Anders. I have passivated the surface.  I will compare the electron difference density with GGA and will check its not weird .

[Anders Blom]

As mentioned, vacuum and TB09 are a bit at odds with each other so there might not be so much you can do about the spike except hope that it doesn't mess up the other quantities you are interested in computing.