Advanced detection in Lorentz microscopy: pixelated detection in differential phase contrast scanning transmission electron microscopy

Krajnak, Matus (2017) Advanced detection in Lorentz microscopy: pixelated detection in differential phase contrast scanning transmission electron microscopy. PhD thesis, University of Glasgow.

Abstract

Modern devices require fundamental length scales to be analysed in a maximum detail to enable research of new types of phenomena and design new materials. In this thesis, an advancement in Lorentz microscopy will be presented where the focus was placed not only onto resolution in spatial space but also onto resolution in reciprocal space. This allows greater sensitivity to measurements of the integrated magnetic induction within thin samples. This was achieved by a novel approach to the data acquisition, where instead of a segmented (annular) detector, a pixelated detector was used to measure the deflection of the scanning transmission microscopy (STEM) probe due to the in-plane integrated magnetic induction.

Computer vision algorithms were researched to find an efficient, noise-robust way to register the deflection of the STEM probe. This enabled a novel approach to data analysis, where a scatter of the 2D integrated induction (a bivariate histogram) is used to show the distribution of the magnetic induction vector. The experimental results are supported by simulations, where a model of a thin polycrystalline sample causes a shift of the simulated beam due to phase modulations. The results of the detection in both the simulation and experiment showed that cross-correlation based processing can efficiently separate the low spatial frequencies (from the in-plane magnetic induction), and high spatial frequencies (from the structure of the polycrystalline sample).

This work will enable quantitative analysis of a greater number of thin magnetic samples, for which the current methods are hampered by the diffraction contrast. This will be particularly helpful for the study low moment, out of plane, magnetised thin films. Currently such systems are of great interest due to the tunability of their magnetic properties and the novel magnetic structures present within them. This work also provides an important step for computational methods in transmission electron microscopy, as this is one of the first examples of 4D data acquisition of processing in STEM (where two dimensions represent the spatial scanning dimensions and other two the reciprocal space).

Imaging methods developed in this thesis were applied to the topic of skyrmions in a thin layer of a FeGe cubic helimagnet, where the very fine detail of the structure of their in-plane integrated magnetic induction was shown to contain a distorted modulations of its profile. This was compared to a simple three harmonic frequency model, which was altered to fit some characteristics of the imaged magnetic skyrmions.

In this work, for the first time, a direct comparison of differential phase contrast and electron holography will be shown for a simple experiment in which the integrated electric field between two needles was measured in free space in the same microscope. Although it was concluded that both methods are equivalent, some small discrepancies of measured values were present due to a long range electric field in electron holography and/or drift of the beam in between scans in STEM.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Keywords: Lorentz microscopy, magnetism, electron microscopy, scanning transmission electron microscopy, pixelated detectors, computer vision, thin films, helical magnets, electric fields, Lorentz microscopy, virtual detectors, GPU accelerated processing.
• Subjects: Q Science > QC Physics
• Colleges/Schools: College of Science and Engineering > School of Physics and Astronomy
• Supervisor's Name: McVitie, Dr. Stephen and Stamps, Prof. Robert
• Date of Award: 2017
• Depositing User: Dr Matus Krajnak
• Unique ID: glathesis:2017-7906
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 03 Feb 2017 13:26
• Last Modified: 22 Feb 2017 14:43
• URI: https://theses.gla.ac.uk/id/eprint/7906