Dear All,
I am going to do a GW calculation about a 1D ribbon structure along y direction, how can I set the variable of "CutGeo"? "CutGeo=none" or "CutGeo=box xz" or "CutGeo=box"?
Many thanks in advance!!!
Coulomn Interaction
Moderators: Davide Sangalli, andrea.ferretti, myrta gruning, andrea marini, Daniele Varsano
-
- Posts: 53
- Joined: Tue Nov 01, 2011 11:04 am
Coulomn Interaction
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
WEI Wei
Postdoctor
Institute of Electrochemistry
Ulm University
Germany
E-mail: dftwei@163.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
WEI Wei
Postdoctor
Institute of Electrochemistry
Ulm University
Germany
E-mail: dftwei@163.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- Daniele Varsano
- Posts: 3816
- Joined: Tue Mar 17, 2009 2:23 pm
- Contact:
Re: Coulomn Interaction
Dea Wei,
Find here an example of input file.
Kepp in mind that the given Lx, the interacting box goes
from -Lx/2 to Lx/2, so Lx/2 should be large enough to include
all your system and the size of supercell big enough that two replica
do not interact in a distance of Lx/2. (see the picture below).
Moreover in order to build up the box Coulomb potential, you need to use the RIM technique
(Random integration method (see the Yambo paper (Comp. Phys. Comm. 180, 1392 (2009)),
as integral of Coulomb potential over the Brillouin zones are needed (the number in this input files
are totally indicative, and may be you need much less RandGvec).
Cheers,
Daniele
Find here an example of input file.
Kepp in mind that the given Lx, the interacting box goes
from -Lx/2 to Lx/2, so Lx/2 should be large enough to include
all your system and the size of supercell big enough that two replica
do not interact in a distance of Lx/2. (see the picture below).
Code: Select all
xxxxxx )( xxxxxxxxx )
<--------|--------->
-Lx/2 Lx/2
(Random integration method (see the Yambo paper (Comp. Phys. Comm. 180, 1392 (2009)),
as integral of Coulomb potential over the Brillouin zones are needed (the number in this input files
are totally indicative, and may be you need much less RandGvec).
Code: Select all
RandQpts= 1000000 # [RIM] Number of random q-points in the BZ
RandGvec= 123 RL # [RIM] Coulomb interaction RS components
CUTGeo= "box xz" # [CUT] Coulomb Cutoff geometry: box/cylinder/sphere
% CUTBox
Lx| 0.0000 | Lz | # [CUT] [au] Box sides
%
Cheers,
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/
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/
-
- Posts: 53
- Joined: Tue Nov 01, 2011 11:04 am
Re: Coulomn Interaction
Dear Daniele,
Great! You always help me a lot!! Thank you so much!!!
Great! You always help me a lot!! Thank you so much!!!
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
WEI Wei
Postdoctor
Institute of Electrochemistry
Ulm University
Germany
E-mail: dftwei@163.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
WEI Wei
Postdoctor
Institute of Electrochemistry
Ulm University
Germany
E-mail: dftwei@163.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~