Correlation Analysis in X-ray Photoemission Spectroscopy

Robert Palgrave,Prajna Bhatt,Yuhan Liu,Mark Isaacs

DOI: https://doi.org/10.26434/chemrxiv-2024-m32rt

2024-04-01

Abstract:X-ray photoelectron spectroscopy (XPS) is a powerful technique for surface analysis, but such analysis can be hindered by uncertainty in modelling spectra. Often, many spectral models have a similar goodness of fit, and distinguishing between them can be impossible without additional information. A further challenge is found in interpreting spectra from samples consisting of multiple chemical compounds. We show here how correlation analysis can be used to interpret large XPS datasets. Correlations in atomic concentrations and binding energies of core lines can be interpreted within a framework of an underlying chemical model and this can yield additional information compared with analysis of each spectrum individually. We give examples of the usage of this analysis on some simple systems, and discuss the potential and limitations of the technique.

Chemistry

What problem does this paper attempt to address?

The paper primarily aims to address several key issues in the analysis of X-ray Photoelectron Spectroscopy (XPS) data:

1. **Uncertainty in Model Selection**: When analyzing XPS data, there are often multiple possible spectral fitting models that may have similar goodness of fit, but represent vastly different chemical environments. Therefore, it is difficult to distinguish between these models without additional information.

2. **Interpretation of Complex Samples**: For samples composed of multiple chemical compounds, interpreting their XPS spectra becomes more challenging. This requires the analyst to not only focus on the information from individual spectra but also to consider the relationships between multiple spectra.

To address these issues, the authors propose a correlation analysis-based method to interpret large XPS datasets. This method utilizes the correlation between atomic concentrations and core-level binding energies, placing them within a potential chemical model framework for interpretation. This provides additional information for the analysis of each individual spectrum.

Specifically, the authors first define three criteria that a good XPS peak model needs to meet:
- The model needs to fit the experimental data well.
- The model must be consistent with spectral theory.
- The model must be consistent with chemical theory.

Next, the authors introduce a simple method to parse large-scale XPS datasets of samples containing multiple compounds, surface contamination, or oxide layers. By analyzing the correlation of atomic percentages and binding energies, the relationships between different phases in the sample can be revealed, helping to determine a more reasonable chemical model.

The paper illustrates this through two examples:
1. A detailed analysis of the natural oxide layer on the surface of aluminum foil, determining which model best fits the chemical theoretical expectations by comparing correlation plots under different models.
2. Analysis of XPS data from a mixture of two tin oxides (SnO and SnO2), similarly using correlation analysis to validate the model's effectiveness.

In summary, the goal of this paper is to provide a new methodology to improve the reliability and accuracy of XPS data analysis, especially when dealing with complex multiphase samples.