Core level shifts

[mhyman]

Hello,

I am attempting to perform some calculations to compare with binding energies obtained by XPS. Earlier posts have indicated that the code used to determine core level shifts is not officially supported, which explains why there is no information in the vasp manual.

To start out with, I'd like to calculate the binding energy of the 4f electrons of Pt(111)

The initial state, I add to the INCAR

ICORELEVEL = 1
CLNT = 1
CLN = 4
CLL = 3
CLZ = 1

Atom 1 in the POSCAR corresponds to a 'different' Pt atom from the others in my slab, such that only that atom will be excited.
TOTEN is identical to that without these added lines, which I would have expected since the electron isn't being excited.

To calculate the final state, I add

ICORELEVEL = 2
CLNT = 1
CLN = 4
CLL = 3
CLZ = 1

TOTEN in this case is more negative by -121.727 eV, which confuses me. It would seem to me that

TOTEN(ICORELEVEL=2) - TOTEN(ICORELEVEL=1)

should equal the binding energy of the core electron. My numbers indicate that exciting the core electron is energetically favorable, thus the binding energy is negative!! For reference, the Pt4f7/2 and Pt4f5/2 binding energies are approximately 71 eV and 75 eV, respectively.

So, I am either inputting the information incorrectly or I am not understanding the results. If the admin or someone else could explain what is going on and what I need to do to perform the desired calculations, it would be much appreciated.

Much thanks,
Matt