Key parameters affecting quantitative analysis of STEM-EDS spectrum images

Area of Science:

• Materials Science
• Analytical Chemistry
• Microscopy

Background:

• Scanning transmission electron microscopy-energy dispersive X-ray spectroscopy (STEM-EDS) is a powerful technique for elemental analysis.
• Quantitative analysis of STEM-EDS spectrum images is essential for accurate material characterization.
• Principal component analysis (PCA) is commonly used for processing complex spectral data.

Purpose of the Study:

• To investigate the impact of operator-controllable parameters on quantitative STEM-EDS analysis.
• To determine the influence of signal level, detector resolution, and PCA factor selection on analytical precision and accuracy.

Main Methods:

• Utilized both simulated and experimental STEM-EDS data.
• Applied principal component analysis (PCA) for spectral image processing.
• Varied signal level, detector resolution, and the number of PCA factors for analysis.

Main Results:

• Increased signal level and detector resolution enhance the precision of quantitative STEM-EDS analysis.
• Signal level was found to be more critical than detector resolution for improving precision.
• Incorrect selection of the PCA solution rank (number of factors) can lead to inaccurate data fitting and reduced accuracy.
• Including too many factors in the PCA model degrades analytical precision.

Conclusions:

• Optimizing signal level and detector resolution is vital for precise quantitative STEM-EDS analysis.
• Careful selection of the principal component analysis (PCA) rank is critical to avoid data misinterpretation and ensure accurate results.
• Overfitting by including excessive factors in PCA negatively impacts the reliability of quantitative STEM-EDS analyses.