Fast STEM image simulation in low-energy transmission electron microscopy by the accurate Chen-van-Dyck multislice...

Area of Science:

• Materials Science
• Physics
• Computational Science

Background:

• The Chen-van-Dyck (CVD) formulation is a rigorous numerical solution to the Schrödinger equation.
• It is the only accurate multislice method for low-energy transmission electron microscopy (TEM) diffraction and imaging.
• CVD accounts for both forward and backscattering effects.

Purpose of the Study:

• Investigate the influence of cutoff value and slice thickness on STEM image simulation accuracy and efficiency using the CVD formulation.
• Assess the significance of including backscattering effects in STEM image simulation.
• Propose methods to enhance computational speed.

Main Methods:

• Numerical simulation using the Chen-van-Dyck (CVD) formulation.
• Analysis of cutoff value and slice thickness effects on accuracy and efficiency.
• Development and implementation of a hybrid computation model with multiple graphic process units (GPUs).

Main Results:

• A small cutoff value is necessary for accuracy in low-energy regimes, particularly for thick specimens.
• An equation for predicting optimal slice thickness was developed.
• A hybrid GPU-based model accelerated STEM image simulation by up to 17 times.

Conclusions:

• The CVD formulation is crucial for accurate low-energy TEM and STEM simulations, especially when including backscattering.
• Optimizing simulation parameters and employing GPU acceleration significantly enhance efficiency and accuracy.
• This work is vital for advancing quantitative STEM imaging in low-energy TEM.