Worse PBE0 result of alpha-Ce magnetic moment by v3.9.0.21 compared to v3.9.0.19

[cyxpku]

Details

I calculated the energy and magnetic moment of alpha-Ce (lattice constant ~ 4.7 Angstrom) ,using ABACUS v3.9.0.19 and v3.9.0.21. I used nspin=2, and tried both PBE and PBE0. The two versions reached an agreement with PBE，but give different results of energy and magnetic moment with PBE0.

I know in v3.9.0.20 EXX PW occupation number in nspin=2 was fixed. However, this appears to have inadvertently introduced a regression. The v3.9.0.19 result reproduces the experimentally observed non-magnetic ground state of alpha-Ce, whereas v3.9.0.21 predicts an large magnetic moment which seems to be unphysical.

Could the developers please investigate whether the EXX PW implementation in v3.9.0.21 contains an undiscovered issue affecting spin-polarized hybrid functional calculations? In the following I attach my input files, and outputs logs for both versions.

issue-Ce.tar.gz

Have you read FAQ on the online manual http://abacus.deepmodeling.com/en/latest/community/faq.html

• Yes, I have read the FAQ part on online manual.

Task list for Issue attackers (only for developers)

• Understand the problem or question described by the user.
• Check if the issue is a known problem or has been addressed in the documentation.
• Test the issue or problem on a similar system or environment, if possible.
• Identify the root cause or provide clarification on the user's question.
• Provide a step-by-step guide, including any necessary resources, to resolve the issue or answer the question.
• If the issue is related to documentation, update the documentation to prevent future confusion (optional).
• If the issue is related to code, consider implementing a fix or improvement (optional).
• Review and incorporate any relevant feedback from users or developers.
• Ensure the user's issue is resolved or their question is answered and close the ticket.

[Flying-dragon-boxing]

A method yielding results consistent with experiment doesn't mean that it is implemented correctly. According to this article, $\gamma$ phase is dominant when mixing parameter of EXX is set to 0.25. I suspect that the previous implementation supresses magnetic moment (through misuse of wavefunctions in wrong spin channels).
However, there is still bug with mixing parameter in EXX PW. I've located it and will fix this soon (hopefully in 1-3 days).

[Flying-dragon-boxing]

Previous implementations would give a near-zero magnetic moment even for Fe.

[Flying-dragon-boxing]

Now things does get a bit weird... There ARE bugs in spin-polarized calculations in EXX PW. Current implementation tends to give a FM phase even for AFM settings. So please stay calm for now and wait for a fix. Sorry for the inconvenience.