Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

Computational methods for quantifying surface structure and dynamics with Low-Energy Electron Microscopy

Ivanov, Matyo 2023. Computational methods for quantifying surface structure and dynamics with Low-Energy Electron Microscopy. PhD Thesis, Cardiff University.
Item availability restricted.

[thumbnail of 2023IvanovMPhD.pdf]
PDF - Accepted Post-Print Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (44MB) | Preview
[thumbnail of Cardiff University Electronic Publication Form] PDF (Cardiff University Electronic Publication Form) - Supplemental Material
Restricted to Repository staff only

Download (304kB)


This thesis focuses on the development of computational approaches to quantify the observations of semiconductor surfaces with Low-Energy Electron Microscopy (LEEM). The real-time surface imaging with LEEM gives us dynamical observations with a suitable temporal and spacial resolution that is difficult to achieve with other microscopy techniques. It allows us to track surface structural dynamics on a nanoscopic scale and characterise them using Computer Vision methods. We monitor the surface of GaAs (001), where we observe a stable coexistence between the (6 x 6) and c(8 x 2) surface phases, as well as a metastable coexistence of those phases during growth. Using the LEEM imaging of the surface dynamics, we are then able to extract previously inaccessible parameters for the two phases. Through Computer Vision methods and computational algorithms, we develop data pipelines that enable the required accuracy and throughput in the analysis of our imaging data. With that, we extract fundamental surface parameters, such as the entropy and stress difference, as well as the step edge energy between the (6 x 6) and c(8 x 2) phases. We use these insights to clarify the long-standing stability question of the (6 x 6) phase and to explain the observed behaviour of the two phases. Furthermore, we use Machine Learning and Deep Learning techniques to develop an approach that streamlines the analysis of the complex and abstract imaging data from the Convergent Beam Low-Energy Electron Diffraction (CBLEED) technique. We show the high accuracy and performance of the developed models in finding surface structural parameters with sub-angstrom accuracy based on CBLEED images.

Item Type: Thesis (PhD)
Date Type: Completion
Status: Unpublished
Schools: Physics and Astronomy
Subjects: Q Science > QC Physics
Uncontrolled Keywords: Low-Energy Electron Microscopy surface imaging phase coexistence entropy difference Gallium Arsenide machine learning
Funders: Engineering & Physical Sciences Research Council (EPSRC)
Date of First Compliant Deposit: 19 October 2023
Last Modified: 20 Oct 2023 10:24

Actions (repository staff only)

Edit Item Edit Item


Downloads per month over past year

View more statistics