Modelling high-precision TEM phase imaging with Density Functional Theory

Modern electron microscopes are now capable of routinely achieving sub-Angstrom resolution, due to recent technological advancements such as aberration correction.
Electron ptychography within the scanning transmission electron microscopy (STEM) provides means to measure the phase shift of an electron wave transmitted through a material. The highly convergent nature of the illuminating beam leads to overlap and interference of waves scattered at different angles and the interference can be used to solve the phase problem allowing inversion to an image. Ptychography therefore bridges diffraction and imaging techniques. One application of ptychography is overcoming the diffraction resolution limit imposed by the numerical aperture of the objective lens. We have recently shown that the measurement is accurate and sufficiently precise to provide a local, real-space measurement of charge redistribution due to bonding.

The quantitative interpretation of such experiment relies heavily on comparison with simulation. This project will develop a novel first-principles quantum mechanical approach for the prediction and interpretation of phase sensitive images, such as those provided by ptychography. The approach will use density functional theory (DFT) within the CASTEP materials modelling code. The initial project aims are the assessment of density functionals, the inclusion of thermal effects, as well as developing the capability to model images in thick samples i.e. materials of more than one atomic layer, in which the weak phase approximation is not valid. Once this is complete the second aim of the project will be to apply the techniques to interpret experimental data including simulations of defects in hexagonal boron nitride, interfaces in strontium titanite, and transition metal oxides. Experimental data will also be provided by Johnson Matthey, who will also provide the student with a 3 month internship. The outcome of this project will provide the electron microscopy community with a valuable tool for the interpretation of phase sensitive images - and hence a powerful approach for the characterization of materials at the atomic scale. However, it will have significant benefits for the materials modelling community as the approach will provide a unique method for evaluating the performance of exchange-correlations functionals, the most significant approximation made in modern electronic structure modelling. In turn this may lead to the development of more accurate modelling approaches.

This project falls within the EPSRC Physical sciences, Energy and Manufacturing the Future research areas.