Thesis presented October 10, 2023

Abstract:
Phase change memories (PCM) offer faster data processing, cost-effectiveness and bridges the gap between DRAM and NAND-Flash, enabling Storage Class Memory (SCM). Reducing programming currents is the key challenge, and new material structures are being considered. GeTe/Sb₂Te₃ (GST) superlattices (SLs) have shown promising results since 2011. Due to the complex nature of SLs, the precise material configuration in its two resistance states and the exact atomic structure of the crystalline state are not yet fully understood. In this context, the focus of this thesis is the development of a new Python framework for automatic processing of High Angle Annular Dark Field-Scanning Transmission Electron Microscopy (HAADF-STEM) images to bring a quantitative, accurate mapping of the different chemical species with atomic resolution together with the precise measurement of the associated interatomic distances. In addition, we developed a new methodology that involves analyzing the statistical positions of atomic columns, which provides information on the anisotropic displacements of atoms. Firstly, our method is tested and validated on a SiGe/Si multilayer sample. Then it is applied to provide detailed information on the structural properties of Sb₂Te₃ thin films. In addition to being a constituent of chalcogenide SLs and a seed layer for the growth of the SL, Sb₂Te₃ is of great technological interest as it is a topological insulator. Results showed that the presence of one to two Te atomic planes on top of the substrate is crucial for successful growth of these films. This is enabled by a van der Waals gap between the substrate and the first Sb₂Te₃ layer, allowing for van der Waals epitaxy. In addition, we reproduced with an excellent precision the absolute HAADF intensity profile across a Sb₂Te₃ QL providing further insight into the atomic vibration characteristics. Finally, the method is applied to the GeTe/Sb₂Te₃ non periodic pseudo-2D materials SLs; in particular, the Ge/Sb intermixing is analyzed quantitatively.

Keywords:
Electron microscopy, Nanosciences, Memory