[Pw_forum] Fermi energy in 'scf' and 'nscf'(XQ Wang)
hande at newton.physics.metu.edu.tr
Wed Aug 13 14:22:16 CEST 2008
> If the difference in Fermi energy is quite large, what I would suggest is to
> increase the no. of k-points in the nscf cal. & redo it.
The fact that Guoying's nscf calculation was conducted along high symmetry
points suggests that it was to be followed by a band structure run. In that
case, it is not necessary to try to obtain a "good" Fermi energy from the
nscf run other than convince oneself since this is not the point
here. Besides, I don't think increasing the number of kpoints along these
high symmetry directions would mean much because since you would still have
a severely heterogeneous sampling. It should just sufficient to just draw
the band structure from nscf along with the Fermi energy from the scf
calculation. But you would be absolutely right if what we were trying to do
here was to, say, increase resolution in a DOS plot.
On a separate note, I suggest that you try bands rather than nscf if you
are trying to draw a band structure since nscf worries about symmetries and
creates extra points whereas bands just does the ones you ask it to do.
Department of Physics
Middle East Technical University
Ankara 06531, Turkey
Tel : +90 312 210 3264
More information about the users