[Wien] Negative Fermi energy

[Stefaan Cottenier]

> I'm  trying  to  calculate  silicon  wires. The calculation of Si(100)
> converged successfully, but the Fermi energy (:FER) is negative (-0.38
> Ry).  However,  when  I  put  it  to  .insp  file  and calculated band
> structure,  zero  level  corresponds to valence band maximum and bands
> are  correct.  When  I've  ran  Si(110),  the Fermi energy was smaller
> (-0.08  Ry).  The  band  structure also looks reasonable, but now zero
> level  corresponds  to conduction band minimum. As there are Si and H,
> I've  used small RKMax=3. However, this situation is rather stable and
> doesn't depend on number of k-points, Lmax or potential choosen. So, I
> wonder what can be wrong or everything is OK?

A partial answer: 

The negative Fermi energies are probably OK. The Ry energy scale of wien2k has 
its zero at the average potential in the insterstitial region. For surfaces 
and wires, you have a lot of interstitial that is vacuum, which leads to 
negative Fermi energies. But that doesn't mean anything, it just depends on 
the choice of what you call zero.

If you plot band structures and select to display the energy axis in eV 
(instead of Ry), then "0" is printed at the energy which corresponds to the 
Fermi energy (change of units Ry -> eV, combined with change of zero-point).

Stefaan