Developing quantitative structure-property relationships for predicting properties of self-assembled materials

MacDonald, Connor Rees Murray (2026) Developing quantitative structure-property relationships for predicting properties of self-assembled materials. PhD thesis, University of Glasgow.

Full text available as:
[thumbnail of 2026MacDonaldPhD.pdf] PDF
Download (62MB)

Abstract

Smart materials are increasingly utilised across many innovative technologies. As these applications become increasingly complex, so too must the materials that facilitate them. Supramolecular materials are of particular interest due to their functional, tuneable nature and their derivability from sustainable feedstocks. While this tunability affords significant opportunity for highly tailored material design from a small library of molecules, the variability in resulting bulk properties makes predictability an ongoing challenge. In this Thesis, we report the development of quantitative structure-property relationships (QSPRs) designed to bridge the gap in predictive ability, enabling a more targeted investigation of material candidates.

First, we introduce the concepts central to developing our quantitative structure-property relationship framework and show how it complies with the Organisation for Economic Cooperation and Development (OECD) guidelines for effective QSPR development. By benchmarking the predictive performance of a model trained on molecular structure descriptors and an IC50 dataset, we show how our system can target a defined endpoint with an unambiguous algorithm, as assessed using appropriate metrics. Predictions can be made by identifying those with sufficient chemical similarity to afford reliable predictions. Lastly, we show how the model can be probed to identify the key chemistries that govern the models it yields.

We go on to demonstrate how our QSPR framework can be adapted to investigate the solubility of amino acid functionalised naphthalene and rylene derivatives in organic solvents. We show how investigating physicochemical and structural descriptors alone is insufficient to elucidate the complexity governing solubility. We demonstrate that QSPRs are an effective strategy for establishing useful relationships between molecular descriptors and this molecular-scale property. Notably, we can highlight the limitations of the current model iteration, providing insight into how future models could be significantly improved.

We then demonstrate how such systems can provide insight into more complex interactions at the nano- and microscale. Using a synchrotron small angle X-ray scattering instrument, a high-quality dataset of structural information was obtained. We discuss the complexity of data fitting and instead show how a simpler quantification of scattering quality can provide insight into the presence of well-defined micellar structures in each solution. The resulting QSPR models demonstrate the ability of QSPR modelling to utilise descriptors of both the solute and the solvent to afford a system capable of incorporating descriptors of chemical constitution and material preparation.

We highlight how the systems developed across this project can be used in tandem to reduce the pool of all candidate materials to those most likely to afford useful aggregated structures.

We conclude with a summary of the achievements of this thesis, limitations that need to be addressed in future systems, and current limitations of literature reporting in the supramolecular community, discussing the need for increased standardisation across the community if more insightful QSPRs are to be developed.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Subjects: Q Science > QD Chemistry
Q Science > QH Natural history > QH345 Biochemistry
Colleges/Schools: College of Science and Engineering > School of Chemistry
Funder's Name: Engineering and Physical Sciences Research Council (EPSRC)
Supervisor's Name: Draper, Professor Emily
Date of Award: 2026
Depositing User: Theses Team
Unique ID: glathesis:2026-86347
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 13 Nov 2026 11:25
Last Modified: 13 Nov 2026 11:36
Thesis DOI: 10.5525/gla.thesis.86347
URI: https://theses.gla.ac.uk/id/eprint/86347
Related URLs:

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year