[QE-users] Inconsistent paw.txt output from AIIDA and default Quantum Espresso

[Wei Bin How]

Thank you for your help Pietro, it is much appreciated!

I understand your point regarding the paw.txt not being used with other density files but I would just like to clarify that the purpose of me mixing them around is to show that the charge density from AIIDA is somewhat reliable while the paw.txt is not. In the end, my goal is to be able to reproduce my original AIIDA results using the charge density and PAW from AIIDA, labelled (4) in the previous message, without the erroneous shift in the 3d orbitals.

I would also like to ask what the paw.txt file contains exactly. I tried searching the documentation but could not find anything about it. Perhaps with some knowledge of what it is, we could figure out a relatively cheap way to obtain the 'correct' paw.txt file.

I have plotted the densities and they seem to agree quite well. For the pseudo charge density (rho_0) the max and mean absolute difference is 1e-4,3e-7. For the all-electron density the max and mean difference is 1e-3 and 3e-07. (Units are based on the standard units output by pp.x)

[https://res.public.onecdn.static.microsoft/assets/mail/file-icon/png/generic_16x16.png]AIIDA_rho_0.cube<https://1drv.ms/u/c/31EAFC1418138016/EeQPKrRI_IdHk8tmiaB4dtYBICNviyPdk_9nrd1xdiZUWw>
[https://res.public.onecdn.static.microsoft/assets/mail/file-icon/png/generic_16x16.png]AIIDA_rho_21.cube<https://1drv.ms/u/c/31EAFC1418138016/EYYIOemTNLNJu5PJyn1JYRMBzrNxf_QoctT1Q88jEVWIhw>
[https://res.public.onecdn.static.microsoft/assets/mail/file-icon/png/generic_16x16.png]pw.in<https://1drv.ms/u/c/31EAFC1418138016/ERPwtkg_xkVDjzWv9oTf5wsBjtqZ7uMg1xm_pTlWaNcZVg>
[https://res.public.onecdn.static.microsoft/assets/mail/file-icon/png/generic_16x16.png]SCF_rho_0.cube<https://1drv.ms/u/c/31EAFC1418138016/ETKtZeRdpTZOuU6sRBxzhTsBuB-mVv_1Y7JMozgkJg2PBw>
[https://res.public.onecdn.static.microsoft/assets/mail/file-icon/png/generic_16x16.png]SCF_rho_21.cube<https://1drv.ms/u/c/31EAFC1418138016/EVYTjtPDECtBi16aOWXLjUcB1CkvTZg1JLbvNArA0DrDKw>


Once again, thanks for your help,
How Wei Bin

________________________________
From: users <users-bounces at lists.quantum-espresso.org> on behalf of Pietro Davide Delugas <pdelugas at sissa.it>
Sent: Wednesday, May 28, 2025 4:23 PM
To: Quantum ESPRESSO users Forum <users at lists.quantum-espresso.org>
Subject: Re: [QE-users] Inconsistent paw.txt output from AIIDA and default Quantum Espresso

Hello
paw.txt is part of the SCF information of pw.x and shouldn't be used with other density files. It may be compatible with .dat files from the same calculation repeated, but there is a possibility that this does not happen.

Could you plot the pseudo charge density and the all-electron density in both cases?
Pietro



From: users <users-bounces at lists.quantum-espresso.org> on behalf of Wei Bin How <howweibin at outlook.com>
Sent: Wednesday, May 28, 2025 14:06
To: users at lists.quantum-espresso.org <users at lists.quantum-espresso.org>
Subject: [QE-users] Inconsistent paw.txt output from AIIDA and default Quantum Espresso

Dear all,

I am trying to perform a NSCF calculation with quantum espresso using the saved charge-density.dat and paw.txt from an AIIDA run, now with a higher number of bands (nbnd). For a number of calculations, the bands from the NSCF are consistent with the original AIIDA SCF, except for the Ga 3d peak. Notably, there was a shift in the 3d peak to higher energy values. For reference, the Ga pseudo is Ga.pbesol-dn-kjpaw_psl.1.0.0.UPF, obtained from SSSP Efficiency 1.2.1. I ran a number of tests and decided that it is likely an issue with the paw.txt

Test:

  1.
Full SCF run with quantum espresso (Eigenvalue matches AIIDA initial run)
  2.
NSCF run using the charge density and paw.txt of (1) (Eigenvalue matches AIIDA initial run)
  3.
NSCF run using the charge density from AIIDA and paw.txt of (1) (Eigenvalue matches AIIDA run)
  4.
NSCF run using the charge density from (1) and paw.txt of AIIDA (Shift in 3d peak to higher energy value)

I have also visually checked the paw.txt from AIIDA and quantum espresso and see that the values are different. I would like to ask if this is a known issue and if there are any suggestions on how to 'correct' the paw.txt file or fix the error?

Thanks,
How Wei Bin
EDMX PhD Student
EPFL
Switzerland

________________________________
From: users <users-bounces at lists.quantum-espresso.org> on behalf of Wei Bin How <howweibin at outlook.com>
Sent: Wednesday, May 28, 2025 14:06
To: users at lists.quantum-espresso.org <users at lists.quantum-espresso.org>
Subject: [QE-users] Inconsistent paw.txt output from AIIDA and default Quantum Espresso

Dear all,

I am trying to perform a NSCF calculation with quantum espresso using the saved charge-density.dat and paw.txt from an AIIDA run, now with a higher number of bands (nbnd). For a number of calculations, the bands from the NSCF are consistent with the original AIIDA SCF, except for the Ga 3d peak. Notably, there was a shift in the 3d peak to higher energy values. For reference, the Ga pseudo is Ga.pbesol-dn-kjpaw_psl.1.0.0.UPF, obtained from SSSP Efficiency 1.2.1. I ran a number of tests and decided that it is likely an issue with the paw.txt

Test:

  1.
Full SCF run with quantum espresso (Eigenvalue matches AIIDA initial run)
  2.
NSCF run using the charge density and paw.txt of (1) (Eigenvalue matches AIIDA initial run)
  3.
NSCF run using the charge density from AIIDA and paw.txt of (1) (Eigenvalue matches AIIDA run)
  4.
NSCF run using the charge density from (1) and paw.txt of AIIDA (Shift in 3d peak to higher energy value)

I have also visually checked the paw.txt from AIIDA and quantum espresso and see that the values are different. I would like to ask if this is a known issue and if there are any suggestions on how to 'correct' the paw.txt file or fix the error?

Thanks,
How Wei Bin
EDMX PhD Student
EPFL
Switzerland
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20250530/9827ec88/attachment.html>