Quantitative atomic resolution transmission electron microscopy for heterogeneous nanomaterials
20 December 2017
Campus Groenenborger, U0.24 - Groenenborgerlaan 171 - 2020 Antwerpen (route: UAntwerpen, Campus Groenenborger
Organization / co-organization:
Department of Physics
Karel H.W. van den Bos
Sandra Van Aert
PhD defence Karel H.W. van den Bos - Faculty of Science, Department of Physics
Nanotechnology forms an inescapable part of our modern everyday life with applications ranging from protective sunscreens to electronic chips in smartphones and computers. In this field, particles are created with sizes of the order of about 1 to 100 nanometres, equal to a billionth of a metre. These sizes are just above the size of individual atoms, which are around 0.1 nanometres. Since the special properties of matter are different at this nanoscale, it enables the creation of new materials with enhanced properties. The basic idea behind the research in this field is that materials properties are determined by their three-dimensional (3D) atomic structure. Therefore, for developing new materials with outstanding properties one fully needs to understand the relation between properties and structure.
In this process, characterisation techniques are required that can retrieve the structure of a material down to the atomic level. One appropriate technique is transmission electron microscopy (TEM), since it provides atomically resolved images which are sensitive to the local 3D structure of the sample under investigation. Since these images are two-dimensional (2D) projections of the materials under study, quantitative methods are required to extract structural information. Ultimately, these methods only use a single 2D projected image to extract 3D structural information. For nanomaterials containing only one chemical element, current techniques can successfully retrieve the 3D atomic structure, while applications for other materials are still limited. Here, 2D TEM images are quantified in a model-based manner by using statistical parameter estimation theory. In this thesis, an open source program called StatSTEM is developed which contains an efficient implementation of statistical parameter estimation theory. Next, state-of-the-art methods have been used to quantitatively compare different TEM techniques. Furthermore, new methods have been developed to retrieve 3D information at the atomic scale of complex nanomaterials containing different chemical elements.