[Wien] XPS core binding energy

[Alexey Kartsev]

Dear WIEN2K users.

My project involves XPS study of adsorbed molecules on the surface of
graphene.
As a starting point, I have tried to calculate binding energy for a 1S core
state of pristine graphene but I did not succeed. The value 296.8 eV I have
got is too far from the experimental results 284.5 eV.
Currently, I am using 16.1 version of WIEN2K.
In my calculations, I have used R*Kmax=7 and 6x6x1 k-points mesh, so it is
should be a fairly good numbers. I have tried to increase another parameter
Gmax from the 12 to 14 but the result is more or less the same.

Basically, from the WIEN2K forum and presentation slides, I understood that
delta-SCF method should be quite precise. It claimed that by using WIEN2K
it is possible to reproduce experimental values for TiC, Ti4C4 and TiN
within error less than 1%.

The procedure I have followed in order to calculate core binding energy is
following:
1) Create large enough supercell in order to avoid hole-hole interaction.
2) To do an SCF calculation for a ground state of that supercell where I
can grab a total energy Eg for the ground state.
3) Choose one single atom in the supercell to be perturbed (ionised) which
correspond to the element of interest.
4) Take away one electron from the core state you want to study in that
atom by changing an occupation number in the case.inc. In the pristine
graphene case, it is the 1S state in one of the supercell atoms. So for
graphene, you have to change occupation of 1S orbital from the 2 to 1.
5) Then in order to enforce charge neutrality of the supercell, you should
put BG charge equal -1 in the case.inm file.
6) Do an SCF calculation for an excited system which will give you a value
of total energy for the ionised supercell - Eexc.

So, in principle, the difference Eb=Eg-Eexc should give me the XPS shift
for the chosen core state. And if your supercell is big enough the results
should be close to the experimental value.

Here are some results for few different parameters and supercells:
4x4 supercell R*Kmax=7 Gmax=14
k-points 4x4x1 dE=296.787287946 eV
k-points 6x6x1 dE=296.87888205 eV

3x3 supercell R*Kmax=7 Gmax=12
k-points 6x6x1 dE=298.147365363 eV
R*Kmax=6 k-points 3x3x1 dE=297.998728554
R*Kmax=5 k-points 3x3x1 dE=298.51212406

5x5 supercell R*Kmax=7 Gmax=12
k-points 4x4x1 dE=297.039874874

Talking about Slater transition method - it gives me an even worse number.
So, I took 3x3 super cell with a 1/2 core-hole.
The 1s level energy of perturbed C atom is  -25.266283216 Ryd. Fermi energy
is -0.5810770100 Ryd.
So Eb= -24.6852062060=335.859462333 eV - very bad.

Am I doing something wrong or misunderstanding something?
Could it be a problem with WIEN2K compilation?
I have tried to calculate DOS and charge density for a pristine graphene
and they look reasonable.

==============================================
*With best regards, *
*   Kartsev Alexey.*

Postdoc Fellow
Atomistic Simulation Centre,
School of Mathematics and Physics,
Queen's University Belfast,
Belfast BT7 1NN, Northern Ireland,
United Kingdom
E-mail: *A.Kartsev at qub.ac.uk <A.Kartsev at qub.ac.uk>*







<https://mailtrack.io/> Sent with Mailtrack
<https://mailtrack.io/install?source=signature&lang=en&referral=karec1@gmail.com&idSignature=22>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20170727/709d3210/attachment.html>