[Pw_forum] smaller band gaps

[Nicola Marzari]

Naol Regassa wrote:
> Dear all
> 
> I was using QE.4.1 to calculate the electronic structure of SrMnO_3
> employing GGA+U
> but I got small band gap of 0.7 eV which is semi-metal (experimentally
> insulator). My main concern is that DFT calculations reduce band gaps
> by 50%. Does this mean that I can add this error to get the correct
> band gap so that any body can understand it? How can one argue about
> this condition?
> 
> Tamene
> Student at AAU
> Ethiopia


Dear Naol,

it's ok to list what the GGA+U band gap is (0.7) and maybe even
the plain GGA. Everyone understands that band gaps are not meant to
be described correctly by DFT - the "band gap" that DFT codes calculate
is that of the fictitious Kohn-Sham states, that have nothing to do,
in principle, with the energy levels of real electrons or of
quasiparticles.

In practive it is found that the Kohn-Sham states do resemble a little
bit the true band structure - albeit with a lot of variability in the 
band gap. LDA/GGA tend to make it smaller, hybrids larger, HSE probably 
quite the right ballpark.

			nicola


---------------------------------------------------------------------
Prof Nicola Marzari   Department of Materials Science and Engineering
13-5066   MIT   77 Massachusetts Avenue   Cambridge MA 02139-4307 USA
tel 617.4522758 fax 2586534 marzari at mit.edu http://quasiamore.mit.edu