Applying automation and machine learning to scanning transmission electron microscopy

Pattison, Alexander J. ORCID: 0000-0001-7869-8910 (2021). Applying automation and machine learning to scanning transmission electron microscopy. University of Birmingham. Ph.D.

[img]
Preview
Pattison2021PhD.pdf
Text
Available under License All rights reserved.

Download (71MB) | Preview

Abstract

This work studies how the benefits of automation and machine learning can be applied to the creation, imaging and image analysis of scanning transmission electron microscopy (STEM) samples. Recrystallised tungsten tips are produced using a semi-automated multi-stage process for use as sample platforms in atomic electron tomography (AET). Two coating techniques are tested to see whether they may be viable methods of reducing sample oxidation. An automated microscope control software framework is presented and demonstrated in three different scenarios: the high-throughput acquisition of CdSe/CdS core-shell nanoparticles, the acquisition of CBED patterns of chiral tellurium nanoparticles and the search for candidate particles for alpha tomography. Finally, machine learning is used to classify the handedness of simulated chiral particles using stereopairs of simulated STEM projections. A 'weak labelling' approach is also demonstrated that takes advantage of the intrinsic nature of chirality to remove the need for manually labelling training datasets.

Type of Work: Thesis (Doctorates > Ph.D.)
Award Type: Doctorates > Ph.D.
Supervisor(s):
Supervisor(s)EmailORCID
Theis, WolfgangUNSPECIFIEDUNSPECIFIED
Kysela, BorisUNSPECIFIEDUNSPECIFIED
Li, ZiUNSPECIFIEDUNSPECIFIED
Licence: All rights reserved
College/Faculty: Colleges (2008 onwards) > College of Engineering & Physical Sciences
School or Department: School of Physics and Astronomy
Funders: Engineering and Physical Sciences Research Council
Subjects: Q Science > QC Physics
URI: http://etheses.bham.ac.uk/id/eprint/11334

Actions

Request a Correction Request a Correction
View Item View Item

Downloads

Downloads per month over past year