[Pw_forum] Fermi energy in 'scf' and 'nscf'

[Nicholas E. Singh-Miller]

Hi Jess,

This might not address your issue, but from your k-point sampling it looks 
like you are performing calculations for a slab geometry (i.e. with vacuum 
layer in between).  If this is the case, the Fermi energy (and the 
eigenvalues) will all be shifted due to the leaving out of the ill defined 
g=0 term in the solving of the electrostatic potential in Fourier space. 
This shift will depend on the cell dimension (e.g. two calculations with 
different sized vacuum regions will report different Fermi energies).

again, i am not sure that this was the issue.

-Nicholas


On Thu, 7 Jun 2007, Jess Kondor wrote:

> Paolo,
>
> Yes, FFT grid is the same for scf and nscf, only k-point mesh is different
> (13x13x1  -scf vs. 23x23x1 -nscf). But, I repeated again scf calculation and
> funny enough Fermi level is 3.86 eV (not 5.05 eV as before). I will repeat
> nscf again to see what is going on.
>
> cheers,
> J.
>
> On 6/6/07, Paolo Giannozzi <giannozz at nest.sns.it> wrote:
>> 
>> 
>> On Jun 5, 2007, at 22:43 , Jess Kondor wrote:
>> 
>> >  I performed 'scf' and 'nscf' calculations with 'm-v' smearing
>> > method. In the end of both calculations, the Fermi energy is the
>> > same. But if these calculations are performed using 'tetrahedra',
>> > Fermi levels are different ( 5.05 eV from 'scf' and 3.86 eV from
>> > 'nscf'). Is it normal?
>> 
>> are the grid for scf and nscf the same? are you getting the same
>> Kohn-Sham eigenvalues?
>> 
>> P.
>> ---
>> Paolo Giannozzi, Democritos and University of Udine, Italy
>> 
>> 
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>> 
>
>
>
>

*****************************************
Nicholas E. Singh-Miller
Ph.D. Candidate
Prof. Marzari Group (quasiamore.mit.edu)
Materials Science and Engineering
Massachusetts Institute of Technology
13-4066
(617)324-0372
*****************************************