Atomic-Scale Structural Modelling of Functional Nanomaterials using Hybrid Experimental/Machine Learning Techniques

Supervisor: Dr Nick Bedford


The properties of all materials are a direct result of the organization of atoms within that material.  As such, the development of structure/function relationships is critical for understanding why a material has its characteristic properties and then using this knowledge to make a new material with improved properties.  Nanoscale materials, in particular, are inherently limited in structure and thus difficult to characterize by traditional means.  One promising route for understanding local structure of nanoscale materials involves the using of synchrotron X-ray characterization methods.  Data collected from the synchrotron can then be used to generate nanoparticle structure models to help established structure/function relationships.  This project aims to couple various synchrotron datasets together in conjunction with machine learning principles to build structural models for nanoparticles with interesting properties.  These include multimetallic nanoparticles and hyperdoped semiconductor nanoparticles.

Prospective students for this project should have a decent background in fundamental thermodynamics and physics and keen computer programming capabilities, as much modelling software is operated in Python.  Interested students should contact Dr. Nicholas Bedford ( for more information.

Suitable for: Chemical Engineers and Food Scientist

Level of difficulty: Challenging