External Electric Fields

VASP has the ability to add external electic fields to, e.g., a surface slab calculation. For this purpose, a sawtooth potential is added to the cell. The kink of the sawtooth is of course unphysical, therefore, enough vacuum should be provided along the z-axis (or, more general, the c-axis) such that the wavefunction will effectively not affected by it. Usually, the kink occurs in the middle of the cell; thus, the surface slab should be located in the lower or upper edge of it.

Electric fields can essentially be added to all types of VASP calculations, e.g., to frequency calculations to obtain Stark shifts.

Build the usual input for the calculation of choice and then add the following commands:

• LDIPOL = .TRUE. Switches on dipole corrections to the potential and forces in VASP
• IDIPOL = 3 The dipole correction is applied along the c-axis (usual choice for slab systems)
• EFIELD = [value] Applies the electric field along the axis specified by IDIPOL with the given strength. Here, an electric field of 0.4 eV/Ang. is applied along the c-axis.

One interesting aspect might be to know how large the actual local field change due to the external field is in certain region of the system. In order to analyze this, first add the LVTOT = .TRUE. command to the INCAR file of the calculation (should be single point now).

The calculation then writes a LOCPOT file (same format as the CHGCAR file), which contains the local electrostatic potential in each integration grid point in the simulation cell in eV. In order to determine the change of effective electrostatic potential at a certain region of space, a second LOCPOT file must be obtained from a calculation of the system without applied electric field.

The effective potential change can then be obtained by subtracting both LOCPOT files from each other on each grid point, respectively. This can be done efficiently with the our program local_efield ([see here](add link)).