[Wien] Unable to see Li K edges in ELOSS file. Kindly guide.

[Peter Blaha]

I analysed your problem. It has nothing to do with WIEN2k, but is much 
more fundamental.

A Li-K edge is per se not a "real edge". It's energy is so low, that it 
may overlap with many other excitations, they could be from Fe-3p (or 
from other shallow core levels ), but of course also from excitations 
from all other valence states into the conduction band.

You can analyse the different contributions to case.joint nicely, by 
playing with the band-indices. Download the 4 images

http:www.wien2k.at/Depository/LiFePO4.joint-total.png

http:www.wien2k.at/Depository/LiFePO4.joint-without-Fe-3p.png

http:www.wien2k.at/Depository/LiFePO4.joint-without-Fe-3p-zoom.png

http:www.wien2k.at/Depository/Li-K-edge.png from

They show case.joint ( unbroadened eps_xx) with the following band 
ranges in case.injoint (1st line):

     1 9999 9999              : LOWER,UPPER and (optional) UPPER-VAL 
BANDINDEX  (total eps_xx)

     13 9999 9999              : LOWER,UPPER and (optional) UPPER-VAL 
BANDINDEX (eps_xx without Fe-3p statesI

     13 9999 16              : LOWER,UPPER and (optional) UPPER-VAL 
BANDINDEX     (eps_xx only from Li-1s

The Li-1s plot (Li-K-edge) starts as expected only at about 50 eV and 
its shape follows closely the Li-2p DOS (not shown), since the Li-1s 
bands are really flat (around -2.9 Ry) and "JDOS = DOS". Please look at 
the small y-scale !!!!  (Note, such a selective absorption is not 
possible experimentally)

The total eps-xx is very large at low energy, and around 53 eV  a sharp 
peak appears .

However, when you  cut-off the Fe-3p states (first 12 bands) this peak 
is gone, which identifies it as the Fe-3p ("Fe-M--2,3 edge", not the 
usual Fe-L-2,3 edge) contribution. Only a very shallow structure can be 
seen and when you  zoom in, the Li-1s absorption is visible, but only a 
"part of the eps2".


Am 01.03.2024 um 21:50 schrieb Pranjal Nandi:
> Dear Peter,
>
> Thanks for the reply. I would like to clarify some misunderstandings.
>
> There are two files for which I am doing the simulation, they are for LiFePo4 and LiTi2(PO4)3. (Note the addition of Ti in the second sample)
>
> The image I had shared was from Optics. I am discussing both the results below.
>
> For the optics results.
>
> -       For LiFePO4, I do not see any Li edge structure.
> -       For LiTi2(PO4)3, I do not see any Li edge structure. What I see is a new peak compared to LiFePO4 at around 35 eV whose shape is very close to the shape of Ti and not of Li.
>
> --
> --------------------------------------------------------------------------
> Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
> Phone: +43-1-58801-165300
> Email: peter.blaha at tuwien.ac.at    WIEN2k: http://www.wien2k.at
> WWW:   http://www.imc.tuwien.ac.at
> -------------------------------------------------------------------------