Advanced Analysis of Molybdenum Oxide Peak Shapes

A recent article from Jonas Baltrusaiti (Lehigh University) et. al. [1] offers some excellent insights into the complexities involved in the analysis of the Mo 3d spectra of the three main oxides, MoO2, Mo2O5, and MoO3.

From the Abstract:
Unlike traditional XPS spectra fitting procedures using purely synthetic spectral components, here we develop and present an XPS data processing method based on vector analysis that allows creating XPS spectral components by incorporating key information, obtained experimentally. XPS spectral data, obtained from series of molybdenum oxide samples with varying oxidation states and degree of crystallinity, were processed using this method and the corresponding oxidation states present, as well as their relative distribution was elucidated. It was shown that monitoring the evolution of the chemistry and crystal structure of a molybdenum oxide sample due to an invasive X-ray probe could be used to infer solutions to complex spectral envelopes.

Figure 4B from [1]. The synthetic LF components were summed to form a single complex line shape for MoO2 (green) once a consistent model with the experimental data emerged.*
Figure 9 from [1]. Two XPS fitting models for a Mo 3d spectrum of an amorphous molybdenum oxide sample: (A) Informed Amorphous Sample Model, and (B) Purely Synthetic Model.*
Some key points from this paper to consider for analysis of molybdenum oxides.
1) MoO3 degrades over time under X-ray exposure. The Mo 3d peak shape for MoO3 is a simple spin-orbit doublet.
2) Pure MoO2 has a complex Mo 3d5/2 peak shape showing a two component (see Figure 4) structure. The sharper, slightly asymmetric main peak at 229.3 eV and broader higher binding energy peak at 231.0 eV are ascribed to screened and unscreened final states [2].
3) This complex structure for MoO2 seems to be lost for mixed oxide samples (see Figure 9). The Mo(IV) peak also moves to a higher binding energy. This is something we've also seen in our studies here. An explanation of why this is so is still to be found.
4) Mo2O5 can only be present when Mo(VI) species are also present. Peak shapes for Mo(V) from this work also suggest a two (or more) component Mo 3d5/2 structure. See black line for Mo(V) in Figure 4B.

Some very speculative comments on these results - read at your own risk...
For point 3 - Are we losing the screened (conductive) portion of the Mo(IV) in unconductive mixed oxide samples?
For point 4 - Is the peak shape obtained for Mo(V) using a multivariate approach showing us a multiplet split peak shape (with or without shakeups etc.) or is it showing us the Mo(VI) component that always must be present when Mo(V) is present? This necessary Mo(VI) component will vary with the peak area of the Mo(V) component (and thus will be elucidated as part of the Mo(V) peakshape). Perhaps some theoretical peak modelling may be useful here.

References:
[1] J. Baltrusaitis, B. Mendoza-Sanchez, V. Fernandez, R. Veenstra, N. Dukstiene, A. Roberts, N. Fairley, Generalized molybdenum oxide surface chemical state XPS determination via informed amorphous sample model, Appl. Surf. Sci. 326 (2015) 151-161.
[2] D.O. Scanlon, G.W. Watson, D.J. Payne, G.R. Atkinson, R.G. Egdell, D.S.L. Law, Theoretical and experimental study of the electronic structures of MoO3 and MoO2, J. Phys. Chem. C, 114 (2010), 4636–4645.

* Figures from [1] reproduced with permission from J. Baltrusaitis.