Multislice forward modeling of coherent surface scattering imaging on surface and interfacial structures

Area of Science:

• Surface science
• Materials science
• X-ray scattering physics

Background:

• Surface-sensitive scattering techniques like grazing incident small angle X-ray scattering provide averaged structural information.
• Grazing incidence geometry with coherent beams can reveal absolute 3D surface morphology.
• Coherent Surface Scattering Imaging (CSSI) is a non-invasive technique for surface nanostructure analysis.

Purpose of the Study:

• To address the limitations of conventional Coherent X-ray Diffractive Imaging (CDI) reconstruction for CSSI.
• To develop a forward model capable of simulating dynamical scattering in CSSI geometry.
• To enable accurate 3D reconstruction of surface nanostructures from CSSI data.

Main Methods:

• Development of a multislice forward model to simulate dynamical scattering.
• Application of the model to CSSI geometry, considering substrate effects.
• Utilizing CUDA-assisted PyTorch optimization with automatic differentiation for reconstruction.

Main Results:

• The multislice forward model successfully simulates dynamical scattering from surface structures and substrates.
• The model overcomes limitations of standard CDI reconstruction for CSSI.
• An elongated 3D pattern was reconstructed from a single CSSI image.

Conclusions:

• The developed multislice forward model is effective for CSSI.
• This method allows for accurate 3D nanostructure imaging on substrates.
• The technique offers a powerful tool for surface morphology characterization.