Challenges to quantitative multivariate statistical analysis of atomic-resolution X-ray spectral.

Area of Science:

• Materials Science
• Analytical Chemistry
• Solid-State Physics

Background:

• Aberration-corrected scanning transmission electron microscopy (STEM) offers atomic resolution.
• Energy-dispersive X-ray spectroscopy (EDS) is crucial for elemental analysis.
• Noise and overlapping signals challenge atomic-resolution elemental mapping.

Purpose of the Study:

• To develop and demonstrate a method for atomic-resolution elemental quantification using STEM-EDS.
• To apply multivariate statistical analysis for noise reduction and signal extraction.
• To investigate the elemental distribution in Y₂Ti₂O₇ pyrochlore at the atomic level.

Main Methods:

• Acquisition of spectral image data using an aberration-corrected STEM with Si-drift EDS detectors.
• Application of multivariate statistical analysis for noise filtering and removal of non-sample-specific X-ray signals.
• Development of a quantification method involving subtraction of non-column-specific signals and internal k-factor derivation.

Main Results:

• Successful acquisition of atomic-resolution spectral image data.
• Demonstration of noise filtering and extraction of relevant correlated X-ray signals (K and L lines).
• Identification of pure Y, pure Ti, and mixed Y-Ti atomic columns in Y₂Ti₂O₇ at the [011] projection.

Conclusions:

• The proposed method shows potential for atomic-resolution quantification in materials.
• Subtraction of non-column-specific signals is a key step in the quantification process.
• A theoretical basis for predicting non-column-specific signals is needed for general applicability.