SUPR
Simulations for vibrational STEM-EELS
Dnr:

NAISS 2024/22-772

Type:

NAISS Small Compute

Principal Investigator:

Paul Zeiger

Affiliation:

Uppsala universitet

Start Date:

2024-09-06

End Date:

2025-10-01

Primary Classification:

10304: Condensed Matter Physics

Webpage:

Allocation

Abstract

Vibrational Electron Energy Loss Spectroscopy (EELS) has developed into a vibrant field since its inception around 2014, when a new generation of monochromators enabled to study vibrations in the Scanning Transmission Electron Microscope (STEM). This technique promises to enhance our fundamental understanding of the flow of heat at the atomic scale with potential applications to for example heat management of microchips, design of thermoelectric materials, and quantum computers. The interaction between the electron beam employed as a probe in STEM-EELS and the sample is, however, very complex and explicit simulations of the beam-sample interaction are required to interpret experiments. In this project, we will perform simulations of vibrational STEM-EELS measurements based on the Frequency-Resolved Frozen Phonon Multislice (FRFPMS) method, a versatile and computationally efficient method for systems with large unit cells and compare such simulations with experimental measurements performed by collaborators. Specifically, we will study a system containing a defect and relate the EELS signal to structural and vibrational properties. Furthermore we will perform simulations on diamond, which will also be compared with experiments and other simulations based on a Bloch wave method. Our calculations follow the FRFPMS method first reported in PRL 124, 025501 (2020). The FRFPMS method can be thought of as a four step process: (1) Molecular Dynamics simulations are used to sample trajectories of the atoms of the vibrating sample. (2) The trajectories are post-processed using band-pass filtering to obtain frequency-dependent snapshots of the structure. (3) These snapshots are then fed into multislice simulations and the vibrational EELS is extracted in a post-processing step by averaging the multislice exit wave functions (4).