This PhD offers the chance to develop advanced machine learning methods for one of the most widely used surface analysis techniques in materials science: X-ray Photoelectron Spectroscopy (XPS). XPS is central to research in semiconductors, batteries, catalysis, and nanomaterials, but spectral interpretation remains a major bottleneck. Complex, overlapping features, charging effects, and background subtraction demand expert judgement, making analysis slow and difficult to scale.
In this project, you will design and implement machine learning architectures capable of robust peak deconvolution, chemical state recognition, and background modelling. You will assemble large experimental datasets using the EPSRC National Research Facility in XPS, supplemented by simulated spectra for training and benchmarking. Machine Learning approaches will include supervised and unsupervised learning, model interpretability techniques, and human-in-the-loop strategies where expert feedback improves model accuracy. You will receive training in Python, data curation, neural network design, and software deployment, alongside advanced knowledge of XPS and surface chemistry. The project therefore provides an interdisciplinary skill set bridging physical sciences and AI, highly valued in both academia and industry.
The ideal candidate will have a background in chemistry, physics, materials science, or computer science, with strong analytical skills and enthusiasm for applying machine learning to complex experimental data.
For more information and to apply visit https://ucl-epsrc-dtp.github.io/2026-27-project-catalogue/projects/2531bd1640.html
