Atomic Imaging With Highly Convergent Electron Beams
Atoms and their arrangement in materials have become a central focus in the era of nanoscience. Materials and devices are now designed at the atomic level-exploiting unique quantum effects that manifest at confined scales. Observing the atomic world is often accomplished through the use of high-energy electrons. The recent advancement of aberration-corrected electromagnetic lenses have enabled highly convergent electron beams confined to lateral dimensions less than one angstrom. Local atomic structure can be probed by scattering sub-angstrom beams through a specimen, with atomic resolution images formed by measuring the scattered electrons as the beam is scanned. In this dissertation, we explore the imaging capabilities and limitations of highly convergent electron beams used in aberration-corrected scanning transmission electron microscopes (STEM). This work provides theoretical & computational approaches as well as experimental work on real systems-including semiconductor devices, nanoparticles used to catalyze hydrogen fuel cells, polymer scaffolds, and 2D membranes. The first two chapters provide a refreshed review of the imaging theory for elastically scattered electron beams. Chapter I motivates atomic characterization by high-energy electrons with their exceptional resolving power. Chapter II takes a more in-depth discussion of elastic scattering in STEM and how beam propagation can be described analytically and computationally. The limitations and detectability of single atoms are explored in Chapter III. The advent of clean monolayer membranes-like that of graphene-provide a playground for exploring the approaches to single atom imaging. Here, detection efficiency and interpretability of common detector geometries are optimized to improve signal-to-noise and open the possibility to single atom imaging of dose limited specimens. In Chapter IV, the thesis extends electron scattering to 3D crystals. When viewing a crystal down a principle zone axis, as is done to obtain atomic images, the complexity of the problem is reduced by mapping the propagating beam to the time evolution of a nonstationary state of a 2D-columnar "molecule". The excitation of the resulting 2D molecular orbitals have distinct characteristic signatures in the images that we are able to observe experimentally, and can drastically and predictably change the apparent location of atoms in samples currently used as resolution tests. Chapter V presents the problematic, small depth-of-field that accompanies highlyconvergent electron beams-causing regions in an extended object to appear blurred and poorly defined. This is overcome by implementing extended depth of field techniques that extract the in-focus regions from a through-focal image stack. Finally, in Chapter VI we discuss new possibilities for 3D imaging with highconvergence angles. To date, high-resolution (< 1 nm) imaging of extended objects in three-dimensions (3D) has not been possible. With current approaches, one is forced to choose between high resolution and large field of view. Instead, by combining throughfocal depth sectioning and traditional tilt-series tomography, we are able to reconstruct over an entire extended object, with improved resolution, in all three dimensions.
Electron Microscopy; Physics; Atomic Imaging
Muller, David Anthony
Silcox, John; Schlom, Darrell
Ph.D. of Applied Physics
Doctor of Philosophy
dissertation or thesis