optical absorption peak of GW+BSE is on lower energy than that of IP and rim_cut band structure

You can find here problems arising when using old releases of Yambo (< 5.0). Issues as parallelization strategy, performance issues and other technical aspects.

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

AmberLee123
Posts: 35
Joined: Sun Jul 26, 2020 7:00 am
Location: Xiamen, Fujian, China

Re: optical absorption peak of GW+BSE is on lower energy than that of IP and rim_cut band structure

Post by AmberLee123 » Mon Apr 25, 2022 9:14 am

Dear Daniele Varsano,

Sorry for the later reply.

The vacuum layer thickness affects BSE absorption a little bit. And 25 Angstrom would gets to the convergence.
060102_vac_conv_2.png
060102_vac_conv_3.png
060102_vac_conv_4.png
I am testing convergence of the bands range of QPkrange as the web page you recommend
http://www.yambo-code.org/wiki/index.ph ... alues_only

(for a simple test, 15 A vacuum layer and kpoints 6x6x1, QPkrange band range 1-34, GW converge in 4 iterations, and band gap increases about 0.4 eV)

I have two questions here,
1. As it say QPrange increases from the G0W0 calculation here(http://www.yambo-code.org/wiki/index.php?title=Silicon), must the QPkrange larger than GbndRnge for GW calculation?

2. To reduce the memory requirement, I hope to use GTermKind= "BG"(http://www.yambo-code.org/wiki/index.ph ... rial:_h-BN)
There is a parameter in input file but not mentioned in tutorial,
GTermEn= 40.81708 eV # [GW] GW terminator energy (only for kind="BG")
Should I test the value, or just left it as default.

Best,
You do not have the required permissions to view the files attached to this post.
Sincerely,
Dr. Guoyu Yang
Lecturer
Jimei Univ, School of Science, Digital Fujian Big Data Modeling and Intelligent Computing Institute
185 Yinjiang Rd.,
Jimei District, Xiamen,361021
Fujian, China
E-mail: 201961000100@jmu.edu.cn

User avatar
Daniele Varsano
Posts: 3773
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: optical absorption peak of GW+BSE is on lower energy than that of IP and rim_cut band structure

Post by Daniele Varsano » Tue Apr 26, 2022 9:04 am

Dear Guoyu Yang,
1. As it say QPrange increases from the G0W0 calculation here(http://www.yambo-code.org/wiki/index.php?title=Silicon), must the QPkrange larger than GbndRnge for GW calculation?
I'm not sure I have understood your question. If you meant that you want to correct more bands as they enter in the G sum over state, in principle yes, you need to correct all the bands entering in the sum. In practice, Yambo makes an interpolation for the missing bands that usually provide a good approximation


2. You can leave it as the default value, the poles will be set 40 eV larger than the last energy difference you have considering the bands you set in GbndRnge. So, its value change automatically according to GbndRnge.

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/

AmberLee123
Posts: 35
Joined: Sun Jul 26, 2020 7:00 am
Location: Xiamen, Fujian, China

Re: optical absorption peak of GW+BSE is on lower energy than that of IP and rim_cut band structure

Post by AmberLee123 » Wed Apr 27, 2022 2:04 am

Many thanks for your dedication! That is what I mean.

%QPkrange # [GW] QP generalized Kpoint/Band indices
kpoint_1|kpoint_2|band_1|band_2| #54| #800|
%

% GbndRnge
Gband_1 | Gband_2 | # [GW] G[W] bands range
%

As far as I understand your answer,
band_1 does not need to be smaller than Gband_1
and
band_2 does not need to be larger than Gband_2


Best,
Sincerely,
Dr. Guoyu Yang
Lecturer
Jimei Univ, School of Science, Digital Fujian Big Data Modeling and Intelligent Computing Institute
185 Yinjiang Rd.,
Jimei District, Xiamen,361021
Fujian, China
E-mail: 201961000100@jmu.edu.cn

User avatar
Daniele Varsano
Posts: 3773
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: optical absorption peak of GW+BSE is on lower energy than that of IP and rim_cut band structure

Post by Daniele Varsano » Wed Apr 27, 2022 8:59 am

Dear Guoyu Yang,

yes, Yambo should interpolate the missing qp corrections, of course, the more bands you consider the more the results of the interpolation will be accurate.
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/

Locked