[Wien] about the fermi level

[Stefaan Cottenier]

> for pure TiO2, its total valence band is below the fermi level as it
> should be for a semiconductor.  for a nitrogen doped TiO2, it should
> be
> a semiconductor too, but the calculated DOS shows that some valence
> bands
> are above the fermi level. who know why?

1) Is the percentage of Ni in your unit cell comparable to experiment? 
(with a 2x2x2 supercell you might have e.g. 8% of Ni, while 
experimental doping is maybe 0.5 % ...?) 

2) Do LDA and GGA give the same result? (what you observe might be an 
LDA- and/or GGA-failure)

Stefaan