Hello,
Dear colleagues,
I will be grateful if anyone helps me in determining which of these XPS results is more accurate. All of these results have been published previously in high impact journals.
Co 2p:
XPS_1 Cobalt Sulfide Nanoflakes Grown on Graphite Foam via solvothermal route
XPS_2 Ball mill assisted synthesis of cobalt–iron sulfide/N-doped carbon
XPS_3 graphene nanosheets anchored with homogeneous cobalt sulfide nanoparticles via sol–gel route
Thanks
Dear Tohid Mahmoudi
I can understand your point. There seems to be some "happy" dispersion in "happy" assignments reported in literature. In my opinion there are sometimes evidences of a rush for giving numbers a label while a critical examination of the experimental results is more or less guilty neglected.
I also understand that examining systems like Co sulfide is not easy, just because the almost always present oxidation may play a crucial role.
The crucial points should be:
1. be sure of the BE axis alignment (usually based on a carbon 1s peak)
2. be sure that no concurrent charging effects would cause peak shifting and smearing
3. remember that core levels are not that much affected by a compound companion when a valence state is fixed
4. remember that the core lelels spin-orbit splitting is usually not modified by the valence state
As a matter of facts, what I would assume is that:
- the Co(II) BE has to be around 781 eV and this can be the ONLY contribution from a Co-S bond (of course, also CoO may give a contribution in the same range)
- the Co(III) peak should be around 2.5 eV lower BEs than the Co(II) one
- the spin-rbit splitting between 2p1/2 and 2p3/2 must be close to 15 eV
- unless spurious additional contributions, the relative wieight of Co(II)-to-Co(III) intensities should be similar in the 2p3/2 and 2p1/2 ranges.
as a consequence, it seems to me that:
A) spectrum XPS_1 wrongly assign the 778.8 eV to Co-S ... I would say it is a Co(III) contribution, for example from a Co2O3 phase. It is clear that the said peak is the lowest BE one in the spectrum! Instead, the 781.2 Ev peak can be assigned to Co(II), which includes CoS2 but may involve also CuO of course. The stellite at 785.6 (with an obviously large position uncertainty due to the large FWHM and low intensity) is roughly where expected with my assignment.
B) Apart from the absolute BE assignments, the results shown in XPS_2 in the 2p3/2 range semms properly assigned. But the 2p1/2 range assignments are, in my opinion, wrong. I think that what the authors identified as a 3+ peak is instead the 2+ one (the true 3+ contribution is very low and not detectable, at about 793 eV).
C) XPS_3 seems to be tyhe most reliable assignment, but the spin-orbit splitting is wrong. It depends on the fact that (too) often one forgets that a fitting is just.. a fitting, and to get more reliable results one should be able to wisely use fitting constraints stemming from "physical" (and not just "numeric") expectations. If you looks to that spectrum in more details you will notice that a 15 eV spin-orbit splitting would be also reproduced if a better analysis of fit uncertainties would have been carried out.
Hope having given you some useful ideas
Dear Giuseppe Curro,
Thanks for your careful examination of the figures and helpful comments.
I don't access to the XPS instrument and we just send the material for the analysis and take the results and then we investigate them.
I know the theoretical basics of the XPS technique and the complexity of the Cobalt sulfide system. But what wondering me is that why some scientific journals publish such roughly examined data, which causes more confusion for the researchers!
In my mind, it's better to consider the quality of published works besides the number of them!
Hello Tahid, have a look at the attached figure (DOI: 10.1016/j.apcatb.2014.04.013). It shows the evolution of Co2p spectrum of oxidized Co(2+) species in a catalyst under sulfidation. The lower spectrum in the figure is typical for Co(2+) species in CoO oxide (in this particular case it may be a more complex oxide with Co(2+) species because of the presence of other metals in the catalysts). The upper spectrum is typical for Co sulfide. Look at the difference between the spectra of oxidized and sulfidized Co, though the formal oxidation state is the same.
In my opinion, all the spectra that you attached to your question are misinterpreted, at least partly. There are also several misleading points in the Giuseppe's answer. Be carefull in analyzing Co2p spectra. This is always tricky.
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publisher