AVS 66th International Symposium & Exhibition
    Surface Science Division Tuesday Sessions
       Session SS-TuP

Paper SS-TuP17
Common Errors in XPS Peak Fitting

Tuesday, October 22, 2019, 6:30 pm, Room Union Station B

Session: Surface Science Poster Session
Presenter: George H. Major, Brigham Young University
Authors: G.H. Major, Brigham Young University
C. Easton, CSIRO Manufacturing
W. Skinner, Future Industries Institute
D.R. Baer, Pacific Northwest National Laboratory
M.R. Linford, Brigham Young University
Correspondent: Click to Email
X-ray photoelectron spectroscopy (XPS) is the most popular method for chemically analyzing surfaces, being used in many areas of research and technology. XPS spectra have layers of information that can be extracted with proper analysis. Information ranges from a basic understanding of components (elements) present in a material to advanced peak fitting and background analysis that reveal chemical states and sample morphologies. There are many examples of good XPS peak fitting in the scientific literature. However, the process of peak fitting XPS spectra is still a mixture of art and science and in many cases have no absolutely correct fits. The peak fitting process can be affected by the instrument design and components, experimental settings, and the sample. Here, we discuss a series of common errors that regularly appear in XPS peak fitting in the literature and how to avoid them. These include: (i) not plotting the data according to the international convention with binding energy increasing to the left, (ii) presenting and interpreting data that are far too noisy to be interpretable, (iii) labeling noise as chemical components, (iv) not showing the original data -- only showing the synthetic (fit) peaks and their sum, (v) not showing any background in a fit, (vi) not providing the sum of the fit components, which makes it difficult for the future reader to determine the quality of a fit, (vii) having widely varying peak widths in a fit, e.g., using extremely broad and extremely narrow peaks when there is no chemical reason for doing so, (viii) having a baseline completely miss the noise/background on either side of the peak, (ix) not collecting data over a wide enough energy window to see a reasonable amount of baseline on both sides of the peak envelope, i.e., truncating the peak, (x) in a fit to a C 1s spectrum, reversing the labeling of the C-O and C=O fit components, and other mislabeling of the components in this envelope (fitting the C 1s peak envelope is well understood so these types of errors should not be made), (xi) for the most part, higher oxidation states of elements correlate with higher binding energies; unfortunately, fit components at higher binding energies are sometimes incorrectly labeled as coming from lower oxidation states, (xii) not taking spin-orbit splitting into account when it is necessary, and/or using inappropriate ratios for these pairs of peaks, and (xiii) in a comparison of related spectra, employing widely different peak widths and positions for components that are supposed to represent the same chemical state and/or using different background types or obviously different types of synthetic peaks in these spectra.