[QE-users] Fwd: Problem with electric polarization calculation in quantum espresso

[Paolo Giannozzi]

(large attachment removed, the pseudopotential file is 
Bi.pbe-dn-rrkjus_psl.0.2.2.UPF)
-------- Forwarded Message --------
Subject: 	Problem with electric polarization calculation in quantum espresso
Date: 	Thu, 14 Mar 2024 01:05:32 +0000
From: 	Akshay Mahajan <akshaymahaja at iisc.ac.in>
To: 	users at lists.quantum-espresso.org <users at lists.quantum-espresso.org>


Dear QE Users,
I have been trying to calculate electric polarization for a 2D Bi 
monolayer (along the in-plane x-direction). I am using QE version 
6.4.1.There are two issues that I am facing,

  1.
     The magnitude of the polarization changes significantly when the k_y
     values change (I am keeping k_z to 1 only since it is a 2D system).
     I expected it to reach some converged value but rather it keeps
     increasing on increasing the k_y value and then drops and then again
     starts increasing. The bash file is included in the email for this
     calculation. In the bash file (polarization_conv.sh) I am changing
     both k_x and k_y in the scf calculation and changing nppstr to 2*k_x
     and keeping k_x =nppstr in the nscf calculation. The result from
     running this bash script run is also included(pol_conv_Bi_mono). One
     can modify the script and check that only changing the k_y is
     causing this huge change in the polarization values.
  2.
     As per this paper
     (https://onlinelibrary.wiley.com/doi/full/10.1002/adfm.201707383
     <https://onlinelibrary.wiley.com/doi/full/10.1002/adfm.201707383>)
     the positive delta_h (refer to the paper for definition) should
     result in a positive polarization. While trying to calculate the
     variation of electric polarization with delta_h, I am getting
     negative polarization values for positive delta_h near the
     centrosymmetric structure (delta_h=0) and positive values for
     negative delta_h (again near the centrosymmetric structure
     (delta_h=0)). The bash script (polarization.sh) for this calculation
     and the result(pol_data_Bi_mono) are also attached.

I am also attaching the pseudopotential file for reference. Please let 
me know if anyone can determine the issue behind these incorrect results.
Thank you.
Regards,
*Akshay Mahajan*
*Ph.D. Student *
*Prime Minister Research Fellow (PMRF) *
*Solid State and Structural Chemistry Unit (SSCU) *
*Indian Institute of Science (IISc)*
*Bengaluru, India *
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pol_conv_Bi_mono
Type: application/octet-stream
Size: 1061 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20240316/5ee7cfed/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pol_data_Bi_mono
Type: application/octet-stream
Size: 865 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20240316/5ee7cfed/attachment-0001.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: polarization.sh
Type: text/x-sh
Size: 7360 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20240316/5ee7cfed/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: polarization_conv.sh
Type: text/x-sh
Size: 3385 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20240316/5ee7cfed/attachment-0001.bin>