Practical guide on chemometrics/informatics in x-ray photoelectron spectroscopy (XPS). I. Introduction to methods useful for large or complex datasets
Tahereh G. Avval,Neal Gallagher,David Morgan,Pascal Bargiela,Neal Fairley,Vincent Fernandez,Matthew R. Linford
DOI: https://doi.org/10.1116/6.0002082
2022-12-02
Journal of Vacuum Science and Technology A: Vacuum, Surfaces and Films
Abstract:Chemometrics/informatics and data analysis, in general, are increasingly important topics in x-ray photoelectron spectroscopy (XPS) because of the large amount of information (data/spectra) that are often collected in degradation, depth profiling, operando , and imaging studies. In this guide, we discuss vital, theoretical aspects and considerations for chemometrics/informatics analyses of XPS data with a focus on exploratory data analysis tools that can be used to probe XPS datasets. These tools include a summary statistic [pattern recognition entropy (PRE)], principal component analysis (PCA), multivariate curve resolution (MCR), and cluster analysis. The use of these tools is explained through the following steps: (A) Gather/use all the available information about one's samples, (B) examine (plot) the raw data, (C) developing a general strategy for the chemometrics/informatics analysis, (D) preprocess the data, (E) where to start a chemometrics/informatics analysis, including identifying outliers or unexpected features in datasets, (F) determine the number of abstract factors to keep in a model, (G) return to the original data after a chemometrics/informatics analysis to confirm findings, (H) perform MCR, (I) peak fit the MCR factors, (J) identify intermediates in MCR analyses, (K) perform cluster analysis, and (L) how to start doing chemometrics/informatics in one's work. This guide has Paper II [Avval et al ., J. Vac. Sci. Technol. A 40 , 063205 (2022)] that illustrates these steps/principles by applying them to two fairly large XPS datasets. In these papers, special emphasis is placed on MCR. Indeed, in this paper and Paper II, we believe that, for the first time, it is suggested and shown that (1) MCR components/factors can be peak fit as though they were XPS narrow scans and (2) MCR can reveal intermediates in the degradation of a material. The other chemometrics/informatics methods are also useful in demonstrating the presence of outliers, a break (irregularity) in one of the datasets, and the general trajectory/evolution of the datasets. Cluster analysis generated a series of average spectra that describe the evolution of one of the datasets.