EDS tomographic reconstruction regularized by total nuclear variation joined with HAADF-STEM tomography

Area of Science:

• Materials Science
• Nanotechnology
• Analytical Chemistry

Background:

• Energy-dispersive X-ray spectroscopy (EDS) tomography provides 3D compositional data for nanostructures.
• Current EDS tomography is limited by poor image quality due to low signal-to-noise ratios and insufficient X-ray counts.
• Limited tilt series in EDS tomography hinder accurate reconstruction of nanostructure composition.

Purpose of the Study:

• To develop a method for accurate EDS tomography reconstructions from low-quality data.
• To improve the 3D compositional characterization of nanostructures.
• To overcome the limitations of insufficient X-ray counts and limited tilts in EDS tomography.

Main Methods:

• Augmenting EDS tomography with high-angle annular dark-field STEM (HAADF-STEM) tomography.
• Utilizing total nuclear variation (TNV) regularization for joint reconstruction.
• Encouraging sparse gradients, common edges, and parallel gradients in joint EDS and HAADF reconstructions.

Main Results:

• Achieved more accurate EDS reconstructions compared to non-regularized and total variation methods.
• Demonstrated improved performance even with a small number of tilts.
• Showcased enhanced reconstruction quality with low X-ray counts.

Conclusions:

• The proposed TNV-regularized joint reconstruction significantly enhances EDS tomography accuracy.
• This method effectively addresses the challenges of low signal-to-noise ratios and limited tilt data.
• The technique offers a pathway to more reliable 3D compositional analysis of nanostructures.