[QE-users] Advice on HP for a PBA material

[Krystian Gierczak]

Dear QE users,

I hope this email finds you well.

I am currently trying to compute the Hubbard U parameters for the Mn[Fe(CN)6] material with the HP code.

I am following the standard workflow from the HP examples as well as the procedure from the Phys. Chem. Chem. Phys., 2023,  25, 9061-9072 article by Dr Timrov.

My procedure is as follows:

  1.
vc-relax,
  2.
fake-metal SCF,
  3.
insulator SCF,
  4.
hp.x

But then, I get a pair of U values that seem chemically counterintuitive, i.e., chemically I would expect my high-spin Mn(III) to have a larger U value than my low-spin Fe(III). Interestingly, this is not at all what I get from hp.x.

Please find the inputs and outputs attached. I have omitted the vc-relax output as it is quite a heavy file.

Please, would you be able to advise on whether I am doing something wrong?

There are also some more theoretical questions relating to the HP code. For example, while I appreciate that the entire DFPT reformulation in terms of monochromatic perturbations allows us to avoid the need for supercells of sizes matching the perturbation wavelengths, I do not quite understand why the aforementioned paper uses a supercell composed of 8 formula units for LiMn2O4? Is it generally desirable to use cells with orthogonal lattice vectors with HP?


Many thanks in advance for your help.

Best wishes,

Krystian

Krystian A. Gierczak
PhD Student in the VLD Group,
Department of Chemistry, University of Oxford
Web: https://deringer.chem.ox.ac.uk

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20260522/1fce702b/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: outputs.zip
Type: application/zip
Size: 225738 bytes
Desc: outputs.zip
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20260522/1fce702b/attachment.zip>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: inputs.zip
Type: application/zip
Size: 4972 bytes
Desc: inputs.zip
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20260522/1fce702b/attachment-0001.zip>