[Pw_forum] general question on materials physics

[Stefano de Gironcoli]

Dear Pascal
     the density of state of a material can have several gaps.
     for instance the core levels of all materials are well separated 
from the valence bands by large gaps, in many polar III-V or II-VI 
semiconductor the lower part of the valence band is separated by a gap 
from its upper part, etc.

     what matters to determine whether a material is a band insulator or 
a band metal is whether a gap separates the occupied from the empty 
levels. If in your case the Fermi energy is located within a partially 
filled group of bands it is a metal, even if somewhere nearby these 
bands are isolated from the others by energy gaps.

     beware that if the partially filled set of bands determines a 
rather narrow peak in the DOS you need to explore the possibility that 
the material is actually a Mott insulator where the independent-particle 
band picture is actually insufficient.

     best rtegards

stefano de Gironcoli - SISSA and DEMOCRITOS

On 11/07/2011 09:07 AM, pascal boulet wrote:
> Dear all,
>
> Sorry to bother you with this silly question but it puzzles me a bit!
>
> In a recent article, I have seen a DOS in which the Fermi level is
> located in the conduction band. So, it seems to me that the materials is
> metallic. But there is clearly a band gap of about 2 eV. I always
> thought that for metals the gap was closed. Is there a contradiction in
> the paper or am I wrong?
>
> Thank you for your answer.
>
> Best
> Pascal
>