Wrong curvature for fractionally charged system compared to QE for both nspin=1,2

[kluonju]

Describe the bug

I ran the case of LiH with fractional charges and compare the the energy curves against quantum espresso (QE). Note in our case ABACUS uses LCAO basis, and QE uses Planewaves. Both cases uses the same pseudopotential (attached) and orb files and input files for both cases

reprod.zip

are also included. Pay special attention when nspin=2, I set tot_magnetization=0.0 to avoid crash.

The energies from QE with both nspin settings align well with each other. However, the energies from ABACUS do not agree and diverge more when delta is larger. Also the curvature does not agree with QE's behavior especially for the electron abundant side (completely wrong).

Expected behavior

No response

To Reproduce

No response

Environment

No response

Additional Context

No response

Task list for Issue attackers (only for developers)

• Verify the issue is not a duplicate.
• Describe the bug.
• Steps to reproduce.
• Expected behavior.
• Error message.
• Environment details.
• Additional context.
• Assign a priority level (low, medium, high, urgent).
• Assign the issue to a team member.
• Label the issue with relevant tags.
• Identify possible related issues.
• Create a unit test or automated test to reproduce the bug (if applicable).
• Fix the bug.
• Test the fix.
• Update documentation (if necessary).
• Close the issue and inform the reporter (if applicable).