KERNEL_TRUNCATION for 2D charged system

[oliver_conquest]

Dear VASP,

I am calculating to effect of Efield for a charged 2D system and was wondering if it is possible to use KERNEL_TRUNCATION to remove interactions along the non-periodic (z-axis) direction?
The VASP 6.5.0 change log states: "Coulomb kernel truncation: open boundary conditions to compute the properties of dipolar and charged molecules, 2D materials, and surfaces.". This suggest KERNEL_TRUNCATION could be used in this capacity, but when I did a test calculation (see attached CKT_test.zip) the SCF cycle begins to diverge after a couple of ionic relaxation steps.

Some notes about the calculation:
- Want to determine how the magnetic moment changes on the Co site as we increase Efield. For CKT_test we just set EFIELD = 0 to start.
- Slab system is asymmetric along the z-direction with the top and bottom layers fixed in the bulk geometry.
- Only interested in how the Co and O positions change along the z-direction with increasing Efield so x,y positions remain fixed.

Kind regards,
Oliver

[ahampel]

Dear Oliver,

thank you for reaching out to us on the official forum.

I fear this might be a similar issue than https://vasp.at/forum/viewtopic.php?p=30167 . The user also reports that the after a few ionic relaxation steps the SCF calculation diverges. Apparently there is a small bug requiring a rather large vacuum layer (more than twice the slab thickness). I will try this over the weekend but since your system is rather large it will take some time to test. Feel free to also try this.

As a backup you can use the old flags: DIPOL, IDIPOL, and LDIPOL to correct electrostatics.

Best,
Alex