A phantom-based forward projection approach in support of model-based iterative reconstructions for HAADF-STEM...

Area of Science:

• Materials Science
• Nanotechnology
• Computational Imaging

Background:

• High-angle annular dark-field (HAADF) imaging is vital for characterizing nanomaterials.
• Accurate simulation and reconstruction of complex nanoparticle structures remain challenging.

Purpose of the Study:

• To introduce a validated forward model for HAADF imaging of nano-crystalline spherical particles.
• To assess the capability of model-based iterative reconstruction (MBIR) for mixed-atomic-number nanoparticle assemblies.

Main Methods:

• Development of a forward model for HAADF image computation.
• Image simulations for various nano-sphere materials (Al, Cu, Au) and sizes.
• Application of MBIR to simulated tomographic tilt series of HAADF and bright-field (BF) images.

Main Results:

• The forward model shows strong agreement with experimental observations on aluminum spheres.
• MBIR successfully reconstructs sphere configurations with mixed atomic numbers.
• Reconstructed intensity ratios accurately reflect the atomic number ratios squared.

Conclusions:

• The developed forward model provides reliable HAADF image simulations for nanoparticles.
• MBIR is effective for reconstructing complex, mixed-atomic-number nanoparticle assemblies.
• This work advances the simulation and analysis of nanomaterials using electron microscopy.