Recreating VASP results - CuO PBE+U - Terrible results

[mdsimula...@gmail.com]

Hello,

I have been trying to recreate a VASP calculation of CuO in CP2K 7.1 (I'm working on compiling CP2K 2022.1) with poor results.  I've also seen this calculation using Quantum Espresso with good success too.

The calculation uses PBE+U and I'm using broken symmetry to set up the antiferromagnetic spins. 

In the paper (https://pubs.acs.org/doi/10.1021/acs.jpcc.5b10431), they are able to get pretty good agreement with the experimental geometry (Table 1) even without the +U but regardless of my +U value, the optimized cell is terrible and the band gap is too low.

I have tried different U_MINUS_J values, 0 to 5, and 3 seems to be the "best".  The high values lead to instability. I've also tried without broken symmetry but results are inconclusive (I need to play with this more).

Does anyone have any suggestions on why the calculation might be giving odd/wrong results?

My input, xyz coordinates, and output are attached.

If more information on my end is needed, please don't hesitate to ask.

This has been troubling me for a month now so any help would be greatly appreciated! 

Best regards,

Frank

[mdsimula...@gmail.com]

To summarize what is wrong, the cell A side is too long, the cell B side is too short and the band gap is too small.

I'm using a 3x3x3 multiple unit cell and a U_MINUS_J 3.0

I'm currently running U_MINUS_J 0.0 and a small unit cell with only KPOINTS.

The experimental cell values: 

ABC 4.653 3.410 5.108

ALPHA_BETA_GAMMA 90.00 99.48 90.00

The values I'm getting with:

ABC 4.846 3.208 5.172

ALPHA_BETA_GAMMA 90.00 94.526 90.00

The HOMO-LUMO gap from CP2K (using PRINT/MO/EIGENVALUES) is 0.96 eV and the reported value ~1.3 eV.

Anyone got any ideas???

[mdsimula...@gmail.com]

Using a 3x3x3 supercell and only PBE, the calculated DOS does not match that given in the literature either.  The CP2K calculated DOS is the top image.  My inputs and outputs are attached.