GW calculation for 1D periodic system

[Zafer Kandemir]

Dear Developers,

I am currently using Yambo version 5.2.0 and I would like to perform GW and BSE calculations for a one-dimensional system with periodicity along the x-direction. Previously, I have only worked on bulk and 2D systems, and I could not find an example for 1D systems in the Yambo wiki.

I am particularly unsure about how to properly set the cutgeo and cutbox parameters for this case. Could you please advise me on the correct setup?

For reference, I am also sharing the system information obtained after running the p2y command.

Code: Select all

____    ____    ___      .___  ___.  .______     ______
\   \  /   /   /   \     |   \/   |  |   _  \   /  __  \
 \   \/   /   /  ^  \    |  \  /  |  |  |_)  | |  |  |  |
  \_    _/   /  /_\  \   |  |\/|  |  |   _  <  |  |  |  |
    |  |    /  _____  \  |  |  |  |  |  |_)  | |  `--"  |
    |__|   /__/     \__\ |__|  |__|  |______/   \______/



 <---> P1: DBs path set to : .
 <---> P1: detected QE data format : qexsd-hdf5
 <---> P1: == PWscf v.6.x generated data (QEXSD-HDF5 fmt) ==
 <---> P1: Header/K-points/Energies...
 <---> P1: Cell data...
 <---> P1: [WARNING] Two or more crystal structures fit the given cell
 <---> P1: Atomic data...
 <---> P1: Symmetries...
 <---> P1: XC functional...Perdew, Burke & Ernzerhof(X)+Perdew, Burke & Ernzerhof(C)
 <---> P1: EXX fraction...:  0.000000
 <---> P1: EXX screening...:  0.000000
 <---> P1: K-points mesh...
 <---> P1: RL vectors...
 <---> P1: IGK arrays...
 <---> P1: Energies...
 <---> P1:  Electrons             :  72.00000
 <---> P1:  Temperature           :  0.950044E-3 [eV]
 <---> P1:  Lattice factors       :  11.25846  37.29556  40.42992 [a.u.]
 <---> P1:  K-points              :  12
 <---> P1:  Bands                 :  300
 <---> P1:  Spinor components     :  1
 <---> P1:  Spin polarizations    :  2
 <---> P1:  Spin orbit coupling    : no
 <---> P1:  Symmetries   [spatial]:  1
 <---> P1:                 [T-rev] : no
 <---> P1:  Max WF components     :  133340
 <---> P1:  RL vectors        (WF):  147889
 <---> P1:  RL vectors    (CHARGE):  1065879
 <---> P1:  XC potential           : Perdew, Burke & Ernzerhof(X)+Perdew, Burke & Ernzerhof(C)
 <---> P1:  Atomic species        :  3
 <---> P1:  Max atoms/species     :   6
 <---> P1: == DB1 (Gvecs and more) ...
 <---> P1: ... Database done
 <---> P1: == DB2 (wavefunctions)  ...
 <---> P1: [PARALLEL distribution for [p2y] WF I/O] Loaded/Total(Percentual):2/12(17%)
 <---> P1: [p2y] WF I/O |                                        | [000%] --(E) --(X)
 <02s> P1: [p2y] WF I/O |########################################| [100%] --(E) --(X)
 <05s> P1: == DB3 (PseudoPotential) ...
 <12s> P1:  == P2Y completed ==

The material under study has antiferromagnetic ordering.

Thank you very much for your support.

Best regards,
Zafer

[Zafer Kandemir]

Dear Developers,

I might have asked my question a bit too early — I think I’ve found the answer

For the 1D system, I prepared the input files as follows:

Code: Select all

CUTGeo= "box YZ"                   # [CUT] Coulomb Cutoff geometry: box/cylinder/sphere/ws/slab X/Y/Z/XY..
% CUTBox
 0.000000 | 37.000000 | 40.000000 |        # [CUT] [au] Box sides
%

Since this structure has an energy gap, I found it more appropriate to use the plasmon-pole approximation (PPA).

Would this be the correct way to generate the GW and BSE input files?

Code: Select all

GW: yambo -coulomb -d p -gw0 p -dyson n -V all  
BSE: yambo -r -X p -o b -k sex -y d -V all 

Best regards,
Zafer

[Daniele Varsano]

Dear Zafer,

you can use the cutoff box, but it is suggested to use:

Code: Select all

CUTGeo= "ws x"
CUTwsGvec= 0.700000
 

You can also raise the cutoff of Gvec to be corrected: CUTwsGvec, but the default value should be enough.
Next, the rest of the variables are the same as the other cases. When using this cutoff check the convergence with kpoints as they affect the quality of the potential.

Best,

Daniele

[Zafer Kandemir]

Dear Daniele,

Thank you very much for your suggestions.
I’ll try using CUTGeo="ws x" as you recommended, and I’ll also check the convergence with respect to the k-point sampling.

Best regards,

Zafer