Trusting the LDIPOL correction

[david6477]

Hello All,

To me, this seemed like a technical question about VASP, so I am placing it here.

I am doing charged point defect calculations in a crystalline solid. The supercell I am using is cubic. I have set NELECT, LDIPOL=T, and DIPOL= 0.5 0.5 0.5 in the INCAR . The current calculation is on a vacancy that is located at the origin (0 0 0)(NELECT is one less than the nuetral supercell that has the same vacancy). All corrections in the OUTCAR, however, are 0:

DIPCOR: dipole corrections for dipol
direction 0 min pos 0,
dipolmoment 0.000000 0.000000 0.000000 electrons x Angstroem
Tr[quadrupol] 0.000000

energy correction for charged system 0.000000 eV
dipol+quadrupol energy correction 0.000000 eV
added-field ion interaction 0.000000 eV (added to PSCEN)

I would have expected that at least the first term (energy correction for charged system) should be non-zero. Can I trust these numbers from VASP?

Thank you for time and help!

[admin]

please check if you have also set IDIPOL in INCAR.

[david6477]

Thank you for the reply!

IDIPOL is not set in the INCAR. The manual says that 1-3 is for slab calculations, which I am not doing, and 4 is for isolated molecules, which I am also not doing. Searching through this forum suggested to me that I only had to set LDIPOL and DIPOL for crystalline solids (symmetrical in all three lattice vectors).

If I need to set IDIPOL, what should it be set to, as none of the possible values outlined in the VASP manual seem appropriate for a crystalline solid that is symmetrical in all directions (I would imagine 4 is the correct value, but the manual says to use it for isolated molecules, and some posts on this forum say that it is for ONLY isolated molecules)?

Thank you again for your time and help!