[Wien] Refractive index calculation using kram

[Pavel Ondračka]

---------- Původní zpráva ----------
Od: Peter Blaha <pblaha at theochem.tuwien.ac.at>
Komu: A Mailing list for WIEN2k users <wien at zeus.theochem.tuwien.ac.at>
Datum: 24. 3. 2017 12:01:18
Předmět: Re: [Wien] Refractive index calculation using kram

"No, you cannot do that. 

You can run x optic on any k-mesh you like and get the momentum 
matrix elements. 

But you cannot run x joint with an arbitrary k-mesh (I'm even 
surprised that it did not crash). joint performs an integration over the 
BZ using a tetrahedron method. So it need a tetrahedral mesh (from x kgen). 


(You cannot calculate a DOS using x tetra either with an arbitrary 
k-mesh (except when using a simple sample method and Gauss broadening). 

On 03/24/2017 09:59 AM, Yong Woo Kim wrote: 
> Dear Wien2k users, 
> 
> Hello, I am currently working on obtaining transition dipole moment 
> values of wurtzite ZnO along a path using optic. 
> 
> I have obtained some results but to check whether they are reasonable, I 
> have been trying to obtain the refractive index values using x kram. 
> 
> It seems that if I perform x kram after obtaining the momentum matrix 
> elements along a certain path, (specified by xcrysden.klist for 
> bandstructure calculations) I get incorrect refractive index values. 
> That is I performed something like the following commands. 
> 
> x lapw1 -band 
> x optic 
> x joint 
> x kram 
> 
> and opticplot_lapw to see the refractive index curve. Am I not supposed 
> to do this and perform x optic on a regular k mesh? 
> 
> I also tried the same thing without the bandstructure calculation part, i.
e. 
> 
> x kgen 
> x lapw1 
> x lapw2 -fermi 
> x optic 
> x joint 
> x kram 
> 
> and the results seemed to be more reasonable but still the refractive 
> index values were about 0.2 lower than the known values. Is this within 
> errors of DFT calculation? Or is it too large and I may have done 
> something wrong? "



Dear Yong,

since prof. Blaha already answered the first part of the question, I'll just
add some notes on this part.




To get a good absolute refractive index value from DFT is quite hard.

If your gap is too low (as usual with LDA or GGA) the transitions are 
shifted lower in energy and you will get high refractive index. Even if you 
correct this eg. with scissor operator then you can have some excitonic 
effects in play which would (besides changing the overall shape of the 
dielectric function) also shift the spectral weight to lower energies => 
since those are not calculated in optic (you would need BSE module for 
this), your calculated refractive index decreases.




Also, all possible transitions contribute to refractive index, so it is 
important that you have high enough emax values in lapw1, optic and joint 
input files, otherwise you will get also lower then expected value due to 
missing high energy transitions.





Sometimes those errors cancel, and you get correct refractive index. For 
example with the original mBJ parametrization which underestimated little 
bit the quasiparticle band gaps the downshift of the gap caused by this 
compensates the missing downshift from excitonic effects and you can get 
quite good refractive index for lot of oxides (such as TiO2, HfO2, SiO2). 
Don't know about ZnO though.




Hope this helps.

Best regards

Pavel





 
"
> Also in the end I need to obtain the momentum matrix elements using the 
> first way, and when I do it should I use a finer k mesh before the band 
> calculation? That is, 
> 
> x kgen 
> x lapw1 -band 
> x optic 
> 
> Right now I am just using a value of 1000 used in init_lapw. Could there 
> be a reasonable value (or minimum value) of k points that should be used? 
> 
> Thank you very much for reading the long post, and any help will be 
> greatly appreciated. 
> 
> Yours Sincerely, 
> Yong Woo Kim 
> 
> 
> 
> _______________________________________________ 
> Wien mailing list 
> Wien at zeus.theochem.tuwien.ac.at 
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien 
> SEARCH the MAILING-LIST at: http://www.mail-archive.com/wien@zeus.
theochem.tuwien.ac.at/index.html 
> 

-- 

P.Blaha 
-------------------------------------------------------------------------- 
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna 
Phone: +43-1-58801-165300 FAX: +43-1-58801-165982 
Email: blaha at theochem.tuwien.ac.at WIEN2k: http://www.wien2k.at 
WWW: http://www.imc.tuwien.ac.at/TC_Blaha 
-------------------------------------------------------------------------- 
_______________________________________________ 
Wien mailing list 
Wien at zeus.theochem.tuwien.ac.at 
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien 
SEARCH the MAILING-LIST at: http://www.mail-archive.com/wien@zeus.theochem.
tuwien.ac.at/index.html 
"
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20170324/fe6b1e20/attachment.html>