[gpaw-users] gpaw-users Digest, Vol 62, Issue 23

[Toma Susi]

Hi Eric,

The Slater transition state gives in some cases better agreement in the spectral shape for e.g. XAS, but as far as I know, for total energy differences the full core hole gives better energies/shifts.

But if you have literature that shows otherwise (specifically XPS!), I would be very interested!

As for spin-orbit in VASP: reading the documentation seems to indicate that the core excited state won't be affected by this flag, but I may be wrong (http://cms.mpi.univie.ac.at/vasp/vasp/ICORELEVEL_tag_core_level_shifts.html).

Best,
Toma

> On 16.3.2015, at 18.26, Eric Hermes <ehermes at chem.wisc.edu> wrote:
> 
> Toma,
> 
> I understand that both VASP and GPAW are unable to explicitly create a 3d 5/2 core-hole state, but I think that's fine as I'm merely attempting to calculate shifts in core-level binding energies, rather than absolute binding energies.
> 
> To touch on the code snippets, would it be possible to create a core-hole setup with corehole=(3, 2, 0.5) and run the calculation with charge=-0.5? This is what is done in Slater-Janak transition state theory, and my understanding from reading the literature is that it is more directly comparable to experimental values than the final state approximation (which is what you have written).
> 
> Also, to respond to your previous email, VASP is capable of including spin-orbit coupling through the LSORBIT=.TRUE. flag. I am not sure if this is exactly what you were referring to, and I'm not sure it would affect my calculations as they have been thus far non-spin-polarized.
> 
> Thanks,
> Eric
> 
>> On 3/16/2015 11:45 AM, Toma Susi wrote:
>> Hi again,
>> 
>> Ah, if the system is that poorly defined then perhaps you are right. It is also experimentally possible to refer the binding energies to the vacuum level, but in that case you need to measure the sample work function, typically by valence band photoemission (UPS).
>> 
>> Perhaps GPAW would give you some advantage in such a calculation, but the description of a 5d core-hole may be rather poor… You can generate a Pd 3d core-hole setup in your script with:
>> 
>>  gen('Pd', name='Pdfch3d', xcname='PBE', corehole=(3, 2, 1.0))
>> 
>> and then apply it to a target atom number # by specifying in the calculator
>> 
>>  setups={#:'Pdfch3d’},  # Applying the generated setup to atom number #
>>  charge=-1     # For a neutral cell
>> 
>> However, this will be neither a 5/2 nor a 3/2 core-hole, but something centro-symmetric and thus not really physical. You might get something sensible for chemical shifts, but I wouldn’t guarantee it.
>> 
>> Although I doubt the error is bigger than what you get from disregarding the spin-orbit coupling, you might get more accurate results by specifying in the core-hole setup generation to add extra projectors to match the GPAW setup for palladium (https://wiki.fysik.dtu.dk/gpaw/setups/Pd.html).
>> 
>> Best,
>> Toma
>> 
>> 
>>> On 16 Mar 2015, at 17:23, gpaw-users-request at listserv.fysik.dtu.dk wrote:
>>> 
>>> Mathias,
>>> 
>>> It sounds to me like GPAW's method for generating the core-hole setup is 
>>> equivalent to VASP's, so I think their results should be comparable.
>>> 
>>> As far as XPS being taken relative to the Fermi level, I've read that in 
>>> the literature but I do not think it is actually true, at least for our 
>>> case. The experimental spectra that I am comparing to are taken for 
>>> amorphous carbon-supported nanoparticles that are a mixture of metallic 
>>> Pd, PdOx, and Pd-Te intermetallics, and we are not even certain about 
>>> the stoichiometry of the intermetallics. Because the system is 
>>> ill-defined, I don't see how the binding energy could be relative to the 
>>> Fermi level (because it is not well-defined). In addition, my advisor 
>>> has spoken to another experimental collaborator for a different project, 
>>> and he has indicated to us that the XPS spectra they take are with 
>>> respect to the vacuum energy. I am not at all an expert in this matter, 
>>> so please correct me if I am wrong.
>>> 
>>> Thanks,
>>> Eric
>> 
> 
> -- 
> Eric Hermes
> J.R. Schmidt Group
> Chemistry Department
> University of Wisconsin - Madison
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.fysik.dtu.dk/pipermail/gpaw-users/attachments/20150316/104da697/attachment-0001.html>