My Community

I am doing calculations on a surface with a fairly large unit cell (11x11 metal atoms, covered with graphene). To compare with STM measurements, I would like to have the workfunction, as measured by STM, for different locations; ideally resolved by surface atom or on a grid in the x-y plane.

Is it LOCPOT that I need, and how do I get the correct values out of it?

Thanks in advance,

Herbert

LOCPOT contains the entire potential. If you just want to know the work function you can take the energy in eV/atom and divide by two based on the slab calculation for a rough estimate. If you want the entire potential across the surface then you will have to parse the LOCPOT file. Please see http://www.dtc.umn.edu/~duany/vasp/node74.html

You should look how the experiment measured local work function. A common technique to measure local work function through STM is by using the exponential decay of the tunneling current I with tip height z:

I ~ exp{-AV^(1/2)z}, where A is 2Sqrt(2m/hbar^2).

V is the tunneling barrier, which is a measure of the local work function. The local work function V can be obtained by

V = {d(ln(I))/dz}^2/A^2,

which is the logarithmic derivative of the STM tunneling current.

You can simulate the STM current by Tersoff-Harman approach, which is calculating the "band decomposed charge density" in the vasp manual. Then numerically evaluate the logarithmic derivative to get the local work function.

The LOCPOT file is the local part of the self-consistent potential, which is not the local work function.
<span class='smallblacktext'>[ Edited Wed Dec 16 2009, 11:27PM ]</span>

To answer my own question: If the global workfunction is the difference between the potential in the vacuum and the Fermi energy, then I should get a reasonable approximation to a "local workfunction" by subtracting the Fermi energy (a location-independent constant) from the values of the local potential (as printed in LOCPOT) in a plane above the surface. Is this correct, or am I making a mistake?