thank you very much Prof. Blaha and Pavel,

Prof. Blaha: Yes, in the paper author reported the sharp peal for lowest
exciton transition using BSE theory.

Prof. Pavel:

I do not have any experience about BSE calculations so for now I want to
bindup the project with DFT only.

I have finished the calculation with -SO now and the shape of the
absorption spectra has changes significantly.

The experimental band gap of the compound is 1.62 eV and I am getting it
1.1 eV with SO.
In the revised absorption coefficient calculated from Wien2k
the absorption peak starts from ~1.0 eV followed by a shoulder like shape
at ~3.eV.
In many literature paper, people reported a small broad peak around  1.7 eV
and then at ~2.5eV.

If I do a tau plot then the onsite absorption edge occurs at ~1.7eV.

My further queries are:

1. Should I report the tau plot to compare my tau plot peak with the the
sharp absorption peak observed in experiment
2. Is it possible to apply scissors correction so that my broad shoulder
like peak may shifts to lower energy side?

Thanks and regards
Ms. Fatima

On Fri, Aug 4, 2017 at 12:10 PM, Pavel Ondračka <>

> fatima DFT píše v Čt 03. 08. 2017 v 15:50 +0530:
> > Dear Users of Wien2k
> >
> > I want to simulate optical properties of a perovskite compound.
> > In the literature people have reported a sharp peak at 1.6 eV while I
> > am not getting it.
> > Here is the file from literature and here is the file that I got from
> > Wien2k.
> > Does including -SO in the calculation may effect the shape of the
> > absorption curve?
> >
> >
> > Regards
> > Fatima
> >
> Dear Fatima,
> it is hard to tell without the knowledge of the compound, however in
> general sharp peaks like this near the absorption edge are usually
> caused by excitonic effects, which are not included in the optic
> package (it uses just random phase approximation without local field
> effects). If you compound is not too complicated and you are skilled
> enough, you can try the BSE unsupported package which can calculate the
> optical spectra including the excitons to see it helps. Just be warned,
> it is unsupported and much less documented than the Wien2k standards.
> Moreover, to get even somewhat converged results requires a LOT of CPU
> time and a LOT of memory.
> Best regards
> Pavel
> _______________________________________________
> Wien mailing list
Wien mailing list

Reply via email to