[QE-users] Problem with computed bandgap

[AISWARYA CHANDRAN]

Dear Quantum ESPRESSO developers and users,

I am working on Ce₂O₃ and trying to reproduce the experimentally reported
band gap (~2.4–2.7 eV) using LDA+U in combination with spin–orbit coupling
(SOC) in Quantum ESPRESSO. However, despite using U = 6 eV and including
SOC via fully relativistic pseudopotentials, the computed band gap is
significantly underestimated (around 1.1 eV).

Here are the key points of my setup:

- Code version: QE 7.X
- Pseudopotential: Ce.pz-spdfn-rrkjus_psl.1.0.0.UPF (fully relativistic)
- Functional: PZ-LDA
- Hubbard_U applied: U = 6.0 eV on Ce 4f
- Calculation type: SCF → bands

- Structure: 5-atom primitive cell from literature (a = 3.87 Å, c/a = 1.522)
- `nbnd = 41`
- Gap is computed using `bands.x` and visual inspection

Despite these, the gap remains ~1.1 eV. Literature reports higher values
for similar setups. I also observe that relaxing the cell via `vc-relax`
makes the lattice smaller, which further reduces the gap. Using
experimental lattice constants does help, but not enough to reach the
expected values.

Could you please advise on:
1. Whether this pseudopotential is appropriate for SOC + LDA+U calculations
on Ce?
2. Any suggestions on correcting the band gap underestimation?
3. Whether a better suited pseudopotential or approach (e.g., using GGA+U,
HSE, etc.) is recommended? I am attaching the input file
4. I tried the same with Ce.pz-spdn-rrkjus_psl.1.0.0.UPF it gives band gap
of 3.7 eV. shall i proceed with spdn?

I would greatly appreciate your insights.


-- 
*Thanks and with Regards, *
*Aiswarya Chandran*

*Research Scholar*

*Laboratory for Advanced Materials and Applied Research*
*Department of Chemistry*
*NIT Calicut, Kerala, India, 673 601*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20250629/9339ccfc/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: scf-ce2o3-new.in
Type: application/octet-stream
Size: 1512 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20250629/9339ccfc/attachment.obj>