Kernel Truncation method limitations

[oliver_conquest]

Dear VASP,

Not sure this is the right place to ask, I have a couple of questions with regards to the Kernel Truncation method paper (Phys. Rev. B 112, 045409) and the limitation of the method.
1. In Figure 4 of Phys. Rev. B 112, 045409 was the unit cell used to calculate MoS2/MoS2(+1) or a NxNx1 super cell? Also confirming for MoS2(+1) a whole valence electron is removed?
2. For large electron doping of a 2D system like MoS2 is it possible there might be some charge spillage into the vacuum region? I have attached a calculation (GaSe_KT) I performed for GaSe which has -0.25e/f.u. electron doping with the Kernel Truncation method. I find charge spillage/accumulation at the boundary edges of the cell which is in the vacuum region this can be observed in the averaged xy local potential and the charge density. I am wondering if there is anything I can do to avoid this charge spillage (e.g. reduce vacuum size) or is it unavoidable for such large doping?

Many thanks for your help!
Kind regards,
Oliver

[christopher_sheldon1]

Dear Oliver,

Thank you for your question. I'll try to clarify.

1) Yes, a unit cell was used, and a whole valence electron was removed.
2) If the xy-averaged potential goes below the Fermi energy, this creates an artificial minimum. I repeated your calculation and found this to be the case. This is the expected solution, though it is non-physical (cf. Phys. Rev. B 85, 045121. In order to fix this, I recommend reducing the amount of negative charge until the potential stays above the Fermi energy.

Does this answer your question?

Best wishes,

Chris