definition of total local potential

[5-903]

Hello.

I want to compute electrostatic potential of a system of interest (organic molecular crystal) on a mesh of points of fairly high density, not necessarily including nuclear positions (if at all). If I understood the instructions this can be done by setting LVTOT = .TRUE. and choosing only electrostatic contributions for the local potential through LVHAR = .TRUE. (I'm using vasp 5.3.X). Then the potential is printed in a cube-like format in the LOCPOT file. Optionally the grid mesh density can be tuned by NGXF, NGYF and NGZF values.

I'm however puzzled by the term "total local potential". How is the electrostatic potential related to it? Is it that if one selects electrostatic contributions only (by LVHAR), the resulting potential is **exactly** what is commonly meant by the "electrostatic potential" term?

Also, the manual states that the potential is specified in electronvolts (eV). In the case of electrostatic potential, shouldn't the units be just in Volts?


Many thanks for clarification.

[cpp6f]

Correct me if I'm wrong, but I think this is the electrostatic part of the pseudopotential, not the all-electron potential, i.e. there are no singularities at the nuclei. Outside the atomic spheres, the two potentials coincide, though. The units are eV per electron, so yes, Volts if you like