[QE-users] Compatibility of lelfield = .true. with PAW Pseudopotentials

[Minsung An]

Dear users,

I am trying to apply a finite electric field to a 3D bulk material using
the `lelfield` tag, and I have a question about its compatibility with PAW
pseudopotentials.

After going through the source code (bp_c_phase.f90), I see that support
for Berry phase calculations is mentioned only for norm-conserving and
ultrasoft pseudopotentials. However, I was able to run atomic position
relaxation (with cell volume fixed) using PAW-LDA without any errors, and
the final relaxed positions were quite close to those obtained using the
ONCV pseudopotential.

My question is:
Is the use of PAW pseudopotentials with Berry phase finite electric field
calculations (`lelfield = .true.`) officially supported or verified for
reliability in QE?

While the calculation runs and seems to yield physical results, I’d like to
know whether the successful run indicates compatibility or whether it's an
unsupported use that somehow just happens to proceed without error.

Thank you!

Best regards,
Minsung An
PhD student
Carnegie Mellon University
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20250422/00676ba6/attachment.html>