question about HF exchange correction

Concerns issues with computing quasiparticle corrections to the DFT eigenvalues - i.e., the self-energy within the GW approximation (-g n), or considering the Hartree-Fock exchange only (-x)

Moderators: Davide Sangalli, andrea.ferretti, myrta gruning, andrea marini, Daniele Varsano

Post Reply
rvatanme
Posts: 18
Joined: Thu Mar 26, 2015 5:33 am

question about HF exchange correction

Post by rvatanme » Sat Jun 27, 2015 1:20 am

Dear all,

I'm using yambo-3.4.1-rev61 to make a correction on Si band gap value calculated by Quantum Espresso and also to get corrected DOS plot. After running HF correction calculation on my input file, yambo only produces r_... file as output not l_... and o_... file. My questions are that
1) How can I extract band gap information from the r-... file?
and
2) How can I get corrected DOS plot based on my HF calculation outputs?

Thanks in advances.

Best,
Reza
Reza Vatan Meidanshahi
Electrical Engineering Department
Arizona State University
USA

User avatar
Daniele Varsano
Posts: 4198
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: question about HF exchange correction

Post by Daniele Varsano » Tue Jun 30, 2015 8:19 am

Dear Reza,
it looks you are using an old version of the code. Before there was not output file for HF corrections, but it is present in newer version.
In any case the HF correction are reported in the report file after the line:

Code: Select all

 XC HF and DFT [eV] @ K [1] (iku): 0.00      0.00      0.00
where you have diagonal HF matrix element and the <Vxc> DFT element, so your non self consistent HF energies will be given by:
E^HF=E^DFT+<HF>-<Vxc>

so be aware that these are not true HF energies but HF energies at first order.
About the dos, while it is straightforward to include GW or COHSEX corrections, in HF probably you first need to generate a QP database by hand according to your results by using the

Code: Select all

ypp -q g
utility, and next you will need to add the keyword:
GfnQPdb= "E < SAVE/ndb.QP"
in your ypp dos input file.

You can try by simply adding in your ypp input file:
GfnQPdb= "E < SAVE/ndb.HF_and_locXC"
but I'm not sure it will work.
Best,
Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

rvatanme
Posts: 18
Joined: Thu Mar 26, 2015 5:33 am

Re: question about HF exchange correction

Post by rvatanme » Sat Jul 04, 2015 3:52 pm

Dear Daniele,

Thank you so much for the explanations. That worked fine, but I got different results for band gap value from r-... file and DOS plot. I calculated (based on your explanations) the value of almost 5.5eV from r-... file and 0.7eV from dos plot. Do you have any what is the source of this difference? I attached my r-... file and dos plot.

Thanks in advances.

Best,
Reza
You do not have the required permissions to view the files attached to this post.
Reza Vatan Meidanshahi
Electrical Engineering Department
Arizona State University
USA

User avatar
Daniele Varsano
Posts: 4198
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: question about HF exchange correction

Post by Daniele Varsano » Sat Jul 04, 2015 4:55 pm

Dear Reza,
it looks your dos is the one computed at DFT level (gap=0.900146 ).
In order to read the HF energies you need to generate the qp database by using the ypp utility as described above by filling the correction by hand (or via a script), it could be very boring...). In case of QP energies (gw or cohsex) instead the database can be read directly.

Best,
Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

rvatanme
Posts: 18
Joined: Thu Mar 26, 2015 5:33 am

Re: question about HF exchange correction

Post by rvatanme » Sat Jul 04, 2015 8:27 pm

Dear Daniele,

You were so right. I wasn't able to generate qp database based on HF calculations (but I'm still trying), so I tried getting DOS plot from my COHSEX calculation and I got the corrected DOS which is in agreement with r-... file.

Thank you SO much for your help and comments.

Best,
Reza
Reza Vatan Meidanshahi
Electrical Engineering Department
Arizona State University
USA

Post Reply