EELS calculation on MgO

[tomohisa_ogawa]

Dear VASP developers
I'm interested in VASP-EELS calculation, and have a question.

I followed XANES in Diamond, https://www.vasp.at/wiki/index.php/XANES_in_Diamond

To test VASP-EELS for two-element system, I calculated O-K edge and Mg-K edge on MgO sinceI found several references for those edges.

the belows are references I compared with
O-K https://www.edge.iis.u-tokyo.ac.jp/CAST ... nglish.htm
O-K https://www.jeol.com/words/emterms/2012 ... #gsc.tab=0
Mg-K https://muller.research.engineering.cor ... oxide-mgo/
Mg-K https://www.advancesoft.jp/products/nano/advance-phase/

But, the spectra I got (attached as VASP-EELS_MgO.png) are a little bit different, especially the initial intensity of Mg-K edge I got is stronger

Are those spectra I got is the expected ones as VASP-EELS? or Do I mistake any settings for the VASP-EELS calculation?
I also attached my code I used for the calculation.

Sorry for VASP-biginner questions, but if you could give me some hints or point out mistakes, it would be very helpful.
Best regards,

[michael_wolloch]

Dear Tomohisa Ogawa,

first of all, I want to apologize for the late response and thank you for your patience and the care you took constructing your post.

Your question is slightly out of scope for this forum, so I moved the topic into the from users for users section. It is unfortunately not possible for the VASP developers manning this forum to dive into comparisons with experiments and other codes due to time reasons. Maybe someone else has some experience with your scientific questions and would like to comment on your issue.

If you have a more specific technical question (about input, output, algorithm, etc.), please make a new topic on the Using VASP forum, which we will be happy to answer.

To get the discussion started, I have some initial points that might be helpful for you:

• There is more material on the wiki regarding super-cell core-hole calculations, which actually has MgO as an example and which you might not have found yet.

• It is recommended to use the _GW variants of the POTCAR files, since your calculation requires a large number of empty bands.

• You can also look at BSE / RPA calculations, especially our tutorial on optical adsorption of LiF. The RPA approximation has shortcomings, i.e., a lack of excitons. However, it accounts for the interaction with the plasmon excitations and often provides a good description of the EELS spectrum.

I hope this helps,
Michael

[tomohisa_ogawa]

Dear Michael
thank you for the helpfull comments, I'll try to follow your comments, first.
If I encounter some trobles, I'll ask some qestions, here.

(actually, I also have a few technical questions on VASP-EELS, I'll post it later as a new topic)
Best regards,
Tomohisa