Dear Yanny, dear Shelly,

I also came across this issue, with Pt and Ir L3 XAS. For these spectra, deltaE0 is systematically 10-12 eV, compared to K edge XAS spectra, where it is typically +/- a few eV. 

My explanation is as follows; if I am wrong or there are other explanations in the community, I would like to know:

In the K edge case you define the actual edge step as your edge position E0 in Athena, which matches pretty well with the "E0" point that Artemis takes as k=0Å-1 when calculating the paths in FEFF. The difference deltaE0 between what Artemis defines and what you define in Athena is thus small. Note that the pre edge peak (transition to bound states, 1s to 3d is weak) is left on the low energy side and typically ignored when choosing E0, because it is usually significantly smaller than the edge step. 

For L3 XAS, the white line is your transition to bound states, 2p to 3d, and is very intensive compared to the edge step. The actual edge step is then buried about 10 eV higher than where the white line onset is found. So when you define E0 as the white line onset, you "underestimate" the actual edge position by about 10 eV or higher. Artemis then corrects it and everything matches. If you define E0 at the white line peak as Shelly suggested, or maybe even after the white line, you will definitely be closer to theory, but that will also affect the background subtraction. 

Hopefully this is helpful!

Cheers, 

Alexey

-------- Original message --------

From: "Kelly, Shelly Diane" <skelly@anl.gov>

Date: 8/23/23 16:11 (GMT+01:00)

To: XAFS Analysis using Ifeffit <ifeffit@millenia.cars.aps.anl.gov>

Subject: Re: [Ifeffit] Shift in Eo when fitting PtO2 and similar samples

Hello Yanny:

 

I’ve modeled a few Pt EXAFS spectra and have found that using an E0 at the top of the white line usually does a better job of getting the data close to the theory. This seems to be often the case with the L3-edges as they are rather broad compared to the K-edges of transition metals.  If you would like to share your project files with me. I can take a look.  14 eV seems to be too much and maybe an indication that there is a different type of atom in the first shell.

 

B.T.W. Using the first derivative is great for XANES analysis, but often not the best choice for EXAFS.

 

Kind regards,

Shelly Kelly (skelly@anl.gov)

 

From: Ifeffit <ifeffit-bounces@millenia.cars.aps.anl.gov> On Behalf Of Zi Qi Chen
Sent: Tuesday, August 22, 2023 9:53 AM
To: ifeffit@millenia.cars.aps.anl.gov
Subject: [Ifeffit] Shift in Eo when fitting PtO2 and similar samples

 

Dear all,

 

I am looking at EXAFS of PtO₂ powders and Pt particles supported on TiO₂. I used a beta-PtO₂ crystal structure data for the PtO₂ EXAFS and after the first shell fit, I observed a large shift in Eo (delta Eo of ~14 eV) in my fit parameters. When I tried to fit the Pt foil to obtain the So², the energy shift also seemed high (delta Eo of ~7 eV). I am wondering if there could be a problem in the selection of Eo prior to EXAFS analysis? I left the Eo as the energy at the first peak of first derivative (ifeffit default).

All of the data were also aligned using the reference foil in Athena before looking at EXAFS. I am wondering what would have caused this shift? Any help is appreciated, thank you.

 

Best regards,

Yanny

 

_______________________________________________
Ifeffit mailing list
Ifeffit@millenia.cars.aps.anl.gov
http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit
Unsubscribe: http://millenia.cars.aps.anl.gov/mailman/options/ifeffit