Multivariate curve resolution for X-ray absorption spectroscopy data interpretation: Best practices and limitations

X-ray absorption spectroscopy (XAS) is a powerful method for tracking the time-dependent changes or spatially-dependent variations in the chemical state and structure of heterogeneous material. However, the fact that the collected signal is averaged over all species co-existing in the probed sample complicates significantly the interpretation of XAS data. To address this issue, application of multivariate curve resolution (MCR) techniques becomes increasingly more popular, allowing one to rationalize the trends in the large sets of experimental XAS data. Here we discuss some of the best practices in employing MCR methods for XAS data interpretation. We also emphasize the limitations and the pitfalls associated with this approach, and motivate the need for more transparency in the application and reporting MCR-XAS results. • Multivariate Curve Resolution (MCR) methods is a power tool for the analysis of X-ray absorption spectra from heterogeneous mixtures and/or sample evolving in time • Application of MCR requires caution and proper reporting of all the data processing steps, employed algorithms and the analysis of uncertainties in the obtained solution, due to the intrinsic ambiguities of the method: commonly more than one solution is possible. • Inclusion of additional information and insight from complementary techniques can reduce the ambiguity of the obtained result • The results provided by MCR method should be always carefully validated through comparison with reference spectra, ab-initio modeling, and/or independent complementary experiments