non-physical potential jump at the end of the simulation cell. VASPsol

[anwin-john]

I drew the work-potential vs. position graphs using the LOCPOT file for different cell heights. The system is Cu(111) surface with *CO surface with 2H2O molecule and one H3O. As you have answered in the previous section, there will be a bound charge in the Cu metal because there is a transference of charge from H3O to the surface.
(H3O to H3O+, the charge on H atom was found to be +0.64 using bader charge analysis)
But, as you can see, there is a peak in all the figures, and as the simulation size increases, there are two observations,
(i) The E,bulk potential reaches near 0 as the cell size increases.
(ii) The height of the nonphysical potential jump also increases near the simulation cell's end.
(iii) As I increase the simulation cell E-fermi printed in the OUTCAR changes and E_bulk as you can see in the above image changes, such that E_bulk - E_fermi ~ constant. But the "Vacuum-Level (eV): " printed in the output of VASPkit is that of the nonphysical potential jump.

I am just trying to find here the work potential, Φ=E_bulk - E_fermi. So that I can find the system's potential with respect to SHE as U=Φ-Φ,SHE/e.

Questions:

1. What is the explanation for the jump?

[anwin-john]

I have solved the issue of potential jump. But I am not able explain the reason.

As you can see from the paper, to calculate the effect of cavitation we need a higher cut off around 1000 eV. But we are only using 450 eV. The cavitation contribution is given as,

So, I changed the default value of Tau from the default value of TAU= 0.000525 to TAU = 0. Now, comparing the old and new potential graph we can see,

[ericfonseca95]

Hello,

The potential jump is non-physical. It is due to having TAU > 0. For surfaces, the cavitation energy is small compared to the other solvation effects, therefore for reliable results, we recommend using TAU=0.000 . This should eliminate the jump you are observing.

Eric