Efficient and versatile model for vibrational STEM-EELS

Paul M. Zeiger,Ján Rusz

DOI: https://doi.org/10.1103/PhysRevLett.124.025501

2019-09-10

Abstract:We introduce a novel method for the simulation of the impact scattering in vibrational scanning transmission electron microscopy electron energy loss spectroscopy (STEM-EELS) simulations. The phonon-loss process is modeled by a combination of molecular dynamics and elastic multislice calculations within a modified frozen phonon approximation. The key idea is thereby to use a so-called $\delta$-thermostat in the classical molecular dynamics simulation to generate frequency dependent configurations of the vibrating specimen's atomic structure. The method includes correlated motion of atoms and provides vibrational spectrum images at the cost comparable to standard frozen phonon calculations. We demonstrate good agreement of our method with simulations and experiments for a 15nm flake of hexagonal boron-nitride (hBN).

Materials Science,Computational Physics

What problem does this paper attempt to address?

This paper aims to solve the problem of how to efficiently and accurately simulate the influence of atomic vibrations on electron scattering when performing vibrational scanning transmission electron microscopy - electron energy loss spectroscopy (vibrational STEM - EELS) simulations at high spatial resolutions. Specifically, the authors attempt to develop a new method that can simulate vibrational electron energy loss spectra while considering dynamic diffraction effects, and this method needs to have the following characteristics:

1. **Efficiency**: The computational cost should be comparable to that of the traditional "frozen - phonon" approximation, but be able to provide more accurate results.
2. **Generality**: Applicable to any system, especially complex systems including interfaces, impurities or defects.
3. **Accuracy**: Able to generate vibrational spectra images consistent with experimental results.

### Method Overview

To achieve the above - mentioned goals, the authors propose a method that combines molecular dynamics (MD) and elastic multislice calculations. The key innovation is the use of the so - called δ - thermostat, which can generate frequency - dependent vibrational structure configurations in classical molecular dynamics simulations. This method not only takes into account the correlated motion between atoms, but also can provide vibrational spectra images with a computational cost comparable to that of standard frozen - phonon calculations.

### Specific Steps

1. **Generate Snapshots**: Heat specific vibrational modes in different frequency ranges through the δ - thermostat to generate a series of "snapshots", that is, atomic structure configurations at different time steps.
2. **Multislice Calculation**: Use these snapshots to perform elastic multislice calculations to simulate the propagation of electron wave functions in crystals.
3. **Separate Vibrational Scattering Intensity**: Extract the vibrational scattering intensity \(I_{\text{vib}}\) by calculating the difference between the coherent and incoherent averages:
\[
I_{\text{vib}}(k_\perp, r_b, \omega)=\frac{1}{N}\sum_{n = 1}^N|\Psi(k_\perp, r_b, \tau_n(\omega))|^2-\left|\frac{1}{N}\sum_{n = 1}^N\Psi(k_\perp, r_b, \tau_n(\omega))\right|^2
\]
4. **Integrate to Generate Images**: Generate vibrational spectra images by integrating over the detector collection angles, and further form the final images by integrating over the energy window.

### Verification and Application

The authors verified the effectiveness of this method by simulating a 15 - nm - thick flake of hexagonal boron nitride (hBN). The results show that this method can well reproduce the experimental data and can produce atomic - scale images with high contrast under both dark - field and bright - field conditions.

In conclusion, this paper proposes a novel and efficient method for vibrational STEM - EELS simulations, providing a powerful tool for vibrational spectra imaging at high spatial resolutions.