3D Spectrum Imaging and Visualisation of Chemical Microstructure

It is essential to understand the structure of materials, as it is the link between structure and properties that determines potential applications. This PhD studentship will make a key contribution to the development of a methodology to produce 3D chemical maps of a multiphase material (or the interface between two different materials) from the visible to the atomic scale by utilising the state of the art electron microscopy facilities in the Leeds Electron Microscopy and Spectroscopy centre (LEMAS).

This will be achieved by developing novel approaches to linking analyses, such as energy dispersive X-ray (EDX) and electron energy loss (EEL) spectroscopies, across length scales, and the development of methods to handle the large data sets that will be generated. The processing will start by using recognized image cleaning techniques. Next machine learning clustering techniques will be applied to the data so that regions with similar spectra and therefore chemical states are identified. This will also identify small regions that could be as small as 1 or 2 data points where any reaction site is and where there are likely to be unusual and/or unstable chemical states. A spectrum that represents each region will be fitted initially manually and over time using AI and deep learning so that chemical states can be visually presented and followed in high resolution over time. These challenges are timely, fitting well with the vision of the Bragg Centre to create an integrated interdisciplinary environment for world-class materials research.

Several materials systems have been identified within the remit of the Bragg Centre and provide an opportunity to develop methods initially using simple model systems (metal alloys and ceramics). Three-dimensional spectroscopic data sets will be collected using techniques such as electron tomography and serial sectioning, with a workflow comprising data collection, analysis which includes machine learning (and eventually AI), visualisation and quantification developed. The versatility of the workflow will then be tested through application to more complicated systems such as 1) composites, 2) catalysts and 3) additive manufacturing samples, with high resolution chemical analysis across a range of length scales visualised and analysed. The final workflow will be of vital importance in the development and understanding of materials and its development is not trivial.

Applicants will have a background in science, engineering or computing; with a willingness to learn to use state of the art equipment and advanced software.