Clustering analysis strategies for electron energy loss spectroscopy (EELS)

Area of Science:

• Materials Science
• Data Science
• Spectroscopy

Background:

• Electron energy loss spectroscopy (EELS) generates large datasets.
• Analyzing EELS data for compositional and oxidation state information can be complex.
• Big data techniques offer potential for efficient EELS data analysis.

Purpose of the Study:

• To explore the application of cluster analysis algorithms for EELS data exploration.
• To evaluate different clustering approaches for analyzing spectral data.
• To assess the ability of clustering to extract material properties from EELS.

Main Methods:

• Direct data clustering of acquired EELS spectra.
• Principal component analysis (PCA) for spectral variance analysis within clusters.
• Data clustering applied to PCA score maps.

Main Results:

• Cluster analysis successfully recovered compositional information from simulated and experimental EELS data.
• Oxidation state information was accurately determined using clustering methods.
• Three distinct clustering approaches were evaluated for their effectiveness and requirements.

Conclusions:

• Cluster analysis provides a powerful tool for analyzing EELS data.
• These methods require minimal user input for extracting valuable material insights.
• Clustering algorithms show significant promise for advancing EEL spectroscopy applications.