>2.00 occupation value in OUTCAR!

Message

hosfa · #1 Post by **hosfa** » Thu Jul 25, 2013 10:30 pm

I did relaxation for an identical phase by PBE and HSE06 functionals.
In Gamma point for all my calculations I have the following values for the energies and occupations for the bands 19 and 20:
19 -3.6224 2.06818
20 -3.6103 1.93182
What's wrong here?

<span class='smallblacktext'>[ Edited ]</span>

Torbjorn · #2 Post by **Torbjorn** » Fri Jul 26, 2013 8:22 am

What Brillouin zone integration technique are you using? Both higher orders of Methfessel-Paxton smearing and Bl?chl corrections to the tetrahedron method give oscillating occupancies near the Fermi level and can give higher occupancies. This is because they attempt to make a better approximation for the step function (for integration purposes) and the price you pay is that you get some ringing.

This is normally not a problem as long as you have a converged k-point mesh, but if you really need "physical" occupancies and convergence of total energy w.r.t. the k-point set, adaptive gaussian smearing is an alternative technique that can be used by hand without implementing anything in the code (just read the valence bandwidth from a DOS plot, the rest are quantities known from the input, so you can calculate a smearing width appropriate to the k-point mesh).