Multipole Correction for Adsorption Energies of Charged Molecules

[oliver_conquest]

Hi all,

I have been doing adsorption energy calculations of charged (using NELECT, so charged super cell) isolated smaller molecules on larger molecules and have found the quadrupole corrections to be exceptionally large (approx. 30 eV) when adding two excess electrons to the system. I have confirmed using charge density difference plots that the charge from the extra electron remains localised.
I realise VASP has the limitation that for charged systems quadrupole corrections are only correct for cubic super cells. Because my system consists of isolated molecules a cubic supercell cannot be used (must be orthorhombic).

Based on the VASP wiki is it best just to set the LDIPOL tag and leave out the IDIPOL and LMONO tags in the INCAR file, as it seems VASP automatically determines if the system is charged?

The alternative might be to turn off the multipole correction/s which seems to provide adsorptions energies closer to what I’d expect but I’m not sure this would be good practice as we want to reduce interactions/contributions between period supercells.

I realise this question has been asked before (see below links). I just want to make sure I’m interpreting the wiki recommendations correctly as the answers provided by the admin just referred onto the manual/wiki. However, the Admin does recommend that multipole corrections be included.
forum/viewtopic.php?f=4&t=1827&p=1872&h ... cule#p1872

Other links to forum posts on corrections for charged systems:
forum/viewtopic.php?f=4&t=17138&p=17753 ... ule#p17753
forum/viewtopic.php?f=4&t=17117&p=17683 ... 539#p17683

Kind regards,
Ollie

[oliver_conquest]

Correction to my previous post, I was meant to say: based in the wiki and previous posts I guess I should just set LMONO = .TRUE. (provides a charge correction to the energy) and leave out the LDIPOL and IDIPOL tags for my charged molecule systems?

Do the moderators or admin/s have any comment on this?

Kind regards,
Ollie

[ferenc_karsai]

I guess you can turn on quadrupole corrections with IDIPOL=1-3. Yes you need to break the symmetry to have more degrees of freedom, but the changes in the lattice constant should be marginal and in practice all lattice constants are the same.
I hope you have read this description on the wiki:
wiki/index.php/Monopole_Dipole_and_Quad ... orrections