[QE-users] problem with convergence of MnTe altermagnet hexagonal phase

[Alpin Novianus Tatan]

Dear Ireneusz,

It does
> not converge with scalar-relativistic pseudopotential. However, with
> full-relativistic approach and cg diagonalization the convergence of
> 5D-5 is achievable but
> cannot go lower. The problem with scalar-relativistic calculations
> occurs in the first iteration where scf accuracy skyrockets above the
> total energy.


Maybe you have a pseudopotential issue for one of the elements in your
system. Why not try using different pseudopotentials? there is no need to
stick to PAW.
Try using ultrasoft or norm-conserving pseudopotentials.

Authors of the paper used VASP but it should be possible in Quantum
> Espresso.


This is not always the case, do not assume this.
Pseudopotentials are an important part of the calculation, which are
different sets for VASP and QE.
The calculation and the code implementation are also different.

Best regards,
Alpin N. Tatan
The University of Tokyo


On Mon, Jul 6, 2026 at 4:26 PM <buganski at agh.edu.pl> wrote:

> Dear All,
>
> I have been trying to reproduce the results from this paper:
> https://www.mdpi.com/1996-1944/18/11/2637 . The authors calculated the
> bands structure with and without spin-orbit coupling to show broken
> time-reversal symmetry. The structure under consideration is hexagonal
> MnTe phase with three atoms in a unit cell (cif included)
> I started the scf calculations with nspin=2 using pbe-paw
> pseudopotential by A. Dal Corso (names in the qe input file). It does
> not converge with scalar-relativistic pseudopotential. However, with
> full-relativistic approach and cg diagonalization the convergence of
> 5D-5 is achievable but
> cannot go lower. The problem with scalar-relativistic calculations
> occurs in the first iteration where scf accuracy skyrockets above the
> total energy.
> I tired lowering the ecutrho to 505, using different mixing_beta, from
> 0.1 to 0.7 or using different mixing_mode. I also tried changing
> startingwfc to 'atomic' what makes computation stable for first few
> iterations and than it once again disconverge.
> Authors of the paper used VASP but it should be possible in Quantum
> Espresso.
>
> Can you give me a hint what should I try to solve this issue? The dos
> computed with soc looks to work as it is similar to what is shown in the
> paper (I did not set J parameter because it is not implements in QE for
> PAW pseudo). But I would like to see the time-reversal symmetry being
> broken for lsda calculations.
>
> Input and output files for one of the tests is attached. I use QE v. 7.3
> on the cluster using 8 nodes with 48 cores.
>
> If I missed important information to resolve the issue, please, message
> me.
>
> Best regards,
> Ireneusz Buganski
> AGH University of
> Krakow_______________________________________________________________________________
> The Quantum ESPRESSO Foundation stands in solidarity with all civilians
> worldwide who are victims of terrorism, military aggression, and
> indiscriminate warfare.
>
> --------------------------------------------------------------------------------
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20260706/86b653d2/attachment.html>