X-ray photoelectron spectroscopy (XPS) is the most widely used surface analysis technique for chemically probing surfaces. Its popularity stems from the large amount of information that can be gathered about the electronic states of the atoms it probes, including core shell information and valence electron information. Simple qualitative analysis (peak identification) can often be performed, but quantitative analysis is a much more complicated process. Although XPS usage has increased dramatically, so has the amount of erroneous analysis observed in the literature. In my thesis, I first present a perspective on how to improve the quality of surface and material data analysis. This chapter focuses on responsible groups, using population biology models and the Prisoner's Dilemma to describe the situation and the potential changes that must be made to counteract error propagation. I quantify errors in XPS data analysis to provide perspective on the gravity of the situation. Over 400 publications in three journals were analyzed. Additionally, another 900 journals were surveyed to determine the quantity of information in the analysis. The parameters include experimental parameters, e.g., the pass energy, peak fitting parameters, the spot size, X-ray source, and the type of spectrometer. I found that over 40% of the publications had significant errors that could potentially change the conclusions of the publication. About 35% of all papers neglected to note the type of spectrometer used, and 85% did not mention the type of software used for analysis. The latter half of this work focuses on XPS peak fitting. I present a broad overview of peak fitting, including how to determine the appropriate background and peak shapes to use, how to quantify XPS data, and how to account for other phenomena associated with photoemission. The line shape chosen for peak fitting is critical, as it is the synthetic shape that is used to model observed physical phenomena. A detailed review on typical line shapes, including the Voigt and pseudo-Voigt functions is presented, along with how to apply them in peak fitting. How and why asymmetric peak shapes are required is also discussed, including which effects cause asymmetry, and if it is inherent to the material or the method of analysis. Finally, a discussion on using constraints to properly model known effects is presented. These efforts were guided by the findings in the former half of this work. The trends presented here are not unique to XPS. Other fields and techniques have similar reproducibility problems. This work discusses possible solutions and what efforts as a community need to be taken to remedy the reproducibility crisis. Additionally, this work includes guides that have original research to improve approaches to XPS analysis, including peak fitting, constraint parameters, and the appropriate use of line shapes.
| Identifer | oai:union.ndltd.org:BGMYU2/oai:scholarsarchive.byu.edu:etd-10931 |
| Date | 18 April 2023 |
| Creators | Major, George Hobbs |
| Publisher | BYU ScholarsArchive |
| Source Sets | Brigham Young University |
| Detected Language | English |
| Type | text |
| Format | application/pdf |
| Source | Theses and Dissertations |
| Rights | https://lib.byu.edu/about/copyright/ |
Page generated in 0.0026 seconds