about monopole correction

[crazyweed]

Dear VASP experts:

As we know the first order energy correction for charged supercell by VASP is referred to the Makov-Payne correction[reference 51 in the related topic in VASP manual] that in the original paper is only for cubic system by the formula:

e^2*q^2*alpha/L/epsilon

Does VASP generalize this first order correction to be suitable for orthorhombic supercell? This point is not stated explicitly in the VASP manual, so I want to make sure.

I tried to compere the correction calculated by VASP for charged supercells with only the length in z direction changes, such as 3x3x2 to 3x3x3 to 3x3x4. I find the absolute value of first order energy correction in OUTCAR is decreasing, which makes me think whether VASP is using a more general method than Makov-Payne correction for the first order correction that can be used to tetragonal or orthorhombic supercell.

Thanks.

[scerwin]

Yes, VASP's Makov-Payne corrections work for any cell whose lattice vectors are orthogonal - not just cubic systems, despite what the guide says.

See http://cst-www.nrl.navy.mil/~erwin/vasptips/#dip_quad

[crazyweed]

hi, scerwin,

Thanks for you reply. But I would like to know how VASP generalizes it to be suitable for orthorhombic lattice system, and what the references are, as the original Makov_Payne reference listed on the vasp manual is only for cubic.