Hi all, I was following this article about Li2TiS3 https://urldefense.us/v3/__https://pubs.acs.org/doi/full/10.1021/acs.inorgch... and found this Ti edge XANES figure [image: image.png] The authors claim The peak position for LTS-213 is approximately 4970 eV, which agrees well
with that for rutile TiO2 with a valence of titanium species in the 4+ oxidation state.
Should we use the pre-edge to determine the oxidation state? Is it reliable, or should we use E₀ and compare it with references of well-known oxidation states?
The second conclusion is even more unclear. Would you help me see if I am missing something in my understanding of oxidation states using XANES? Best, B.
Hi B. Have you had any lessons on XAFS? i.e. attended an XAFS course offered by many synchrotrons and related institutions. Do you know the difference between pre-edge features and absorption edge? In the plot you provided, the edge positions are clearly different. Formal oxidation state is a convenient way to count electrons. Edge position is governed by how the screening of the nucleus of the absorber is affected by the type and arrangement of near neighbours - orbitals involved in bonding and extent of charge transfer between absorber and ligands...hence the usefulness of XAFS as a technique for characterising local environments. Still confused? Start here: https://urldefense.us/v3/__https://xafs.xrayabsorption.org/tutorials.html__;... -R. On 2025-02-26 11:31 p.m., Benito Melas via Ifeffit wrote:
Hi R., I understand that the pre-edge corresponds to a transition that is normally forbidden by symmetry, while the absorption edge represents the transition from the core level to the continuum. In Fundamentals of EXAFS, Figure 7.3 shows that the absorption edge shifts when iron changes from its metallic state to +2 and +3 oxidation states. The caption states: "Fe K-edge XANES of Fe metal and several Fe oxides, showing a clear
relationship between edge position and formal valence state."
As the edge position is unclear for me in the LST compound, should it be the point of maximum slope? The LST peak at 4975 eV could correspond to the B peak reported by Rossi et al. (DOI: 10.1103/PhysRevB.100.245207), which is an aspect I am not entirely familiar with. I reviewed the paper by Farges (DOI: 10.1103/PhysRevB.56.1809) and did not observe anything similar. This is the first time I have seen the B peak so intense; previously, I have encountered lower-energy peaks with higher intensity, likely due to distortions in the local symmetry. I understand that the pre-edge is more related to local symmetry than to the oxidation state, based on the manuscripts I mentioned. Could you please clarify this further? I am confused by this sentence:
I will check the tutorials, I always catch something new. Best regards, On Fri, Feb 28, 2025 at 2:04 AM Robert Gordon via Ifeffit < ifeffit@millenia.cars.aps.anl.gov> wrote:
Hi B. A correlation between formal oxidation and edge shift from metal (metal = pure covalent) to oxides can certainly be observed. If you consider Fe bonded to O in a variety of compounds (i.e. not just Fe-oxides), you will see that for different formal valence, the edge positions cluster. They are not all the same for a given formal valence. This occurs due to changes in the Fe-O interaction when the O also interacts with something else. The pre-edge can also show this. A particularly beautiful paper illustrating this is by Petit, Farges, Wilke and Sole (J. Synchrotron Rad. (2001). 8, 952-954) for an assortment of Fe minerals. For the LI2TiS3 compound in the plot you provided, I would take the edge position (XAFS Eo) to be about 4975. I would want to look at the derivative and look at the first peak in the derivative of the main rising edge (i.e., yes, max slope). Similarly for r-TiO2, I would consider the edge position to be about 4978 eV. The difference in edge position between two Ti species with formal oxidation of +4 would be attributed to the differing bonding environments. With Ti-S more covalent than Ti-O, one wold expect the Ti-S compound to have a smaller edge shift from the metal than the Ti-O compound. "I am confused by this sentence: "The peak position for LTS-213 is approximately 4970 eV, which agrees well with that for rutile TiO₂ with a valence of titanium species in the 4+ oxidation state." The authors observe peaks in close proximity for two compounds with differing bonding environments (Ti-O vs Ti-S) and note that one of these compounds has Ti with a formal valence of +4. What the authors have not done is a systematic study of correlations between pre-edge peak positions for Ti-containing compounds where Ti is bonded to sulphur or oxygen. Any conclusions based on one comparison are likely highly influenced by observer bias. They made a compound called Li2TiS3, and they want to say S-2, Li+1 ergo Ti+4. They are being somewhat carefree in their interpretation of the limited data. If they had at least compared to TiS2, it would be a more valid comparison. Formal oxidation state is convenient for electron counting. For XAFS edge position, what influences that is governed by the extent of charge transfer to/from the local environment about the absorber. Correlations with formal oxidation state can be established for similar environments (i.e. ligand type), but even that will show subtleties for variations in environment. When analyzing and interpreting data, don't overreach. -R. On 2025-02-28 6:10 a.m., Benito Melas wrote:
participants (4)
-
Benito Melas -
matt.newville＠gmail.com -
matthew marcus -
Robert Gordon