In addition to Xaviers comments: Yes, if you have eg. 2 oxygen atoms in your structure, the differences in the core energies should reflect pretty well the experimentally measurable core-level shifts.
If you want to calculate the core energies better you can use a) Slater's transition state b) delta-scf calculations a) remove HALF a core electron from case.inc (add it to case.inm) do an scf calculation. Now the (eigenvalue-EF) difference should be closer to experiment (expect errors of a few eV instead of 30) b) You can also use a full core hole and compare E-tot of groundstate and excited state. Eventually this should be done in a supercell and only on one of the atoms the hole should be added to avoid hole-hole interactions. Rocquefelte schrieb: > Dear James, > > If you want to compare valence XPS spectrum with the DOS deduced from > your Wien2k calculation, you will obtain a qualitative agreement. > Concerning the core states, you may have access to the energy of these > states in the case.scfc file. These values are relative to the fermi > level given in the case.scf file. > Thus you should modify these energies in such a way to have a fermi > level comparable to your XPS experiment (the top of the valence). > > Here is an example based on TiO2 rutile : > > - Energy of Ti 2p1/2 (2p* in scfc file) = -442.5 eV > - Energy of Ti 2p3/2 (2p in scfc file) = -436.7 eV > > - Energy of O 1s = -504.2 eV > > These values should be compared to the following experimental XPS > binding energies (BE): > > - BE Ti 2p1/2 = 464.4 eV > - BE Ti 2p3/2 = 458.6 eV > > - BE O 1s = 530.1 eV > > So, as you can see, a strong deviation (of about 30 eV) is observed > between experimental and calculated data. > However, I would believed that the deviation is smaller if you compare > different BE in a same structure. > > Hope that this email will help you > > Regards > > > > James Enterkin a ?crit : >> Dear All, >> >> I was wondering if anyone would know how to go about calculating the >> energy of an XPS peak? >> I have a converged structure, and would like to be able to calculate >> the binding energy of an electron on a specific atom in the structure, >> to see if it matches up with the shift in the experimental XPS of that >> same structure. >> If anyone has previously done this or similar, or knows what method >> would be best to go about this, your suggestions would be most helpful. >> >> Thank you, >> Jim >> >> -- >> James A. Enterkin >> j-enterkin at nortwestern.edu <mailto:j-enterkin at nortwestern.edu> >> Ph.D. Candidate >> K.R. Poeppelmeier and L.D. Marks Groups >> Department of Chemistry, Northwestern University >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> Wien mailing list >> Wien at zeus.theochem.tuwien.ac.at >> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien >> > > _______________________________________________ > Wien mailing list > Wien at zeus.theochem.tuwien.ac.at > http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien