Technische Universität Dresden
Quantitative off-axis Electron Holography and (multi-)ferroic interfaces
Abstract
dc:description.abstractA particularly interesting class of modern materials is ferroic ceramics. Their characteristic order parameter is a result of quantum chemistry taking place on a sub-Å length scale and long-range couplings, e.g. mediated by electrostatic or stress fields. Furthermore, the particular subclass of multiferroics possesses more than one order parameter and exhibits an intriguing coupling between them, which is interesting both from the fundamental physics point of view as well as from a technological vantage point. While on a more fundamental level it is desirable to elucidate the physical details of the coupling mechanism, this knowledge could subsequently lead to new and technologically interesting multiferroic materials, which overcome their current drawback that only one of the multiple order parameters is appreciably large while the others stay small. Due to the short and long range nature of the driving forces, one challenge for thoroughly understanding ferroic ceramics is the characterization of material properties within a large interval of length scales from several tens of µm to sub-Å. To that end, it is useful to exploit that all order parameters can be described as macroscopic fields, e.g. electric polarization or strain, which, in turn, can be either directly or indirectly probed with an electron beam such as used in Transmission Electron Microscopy (TEM). Consequently, TEM is excellently suited for investigating ferroic materials, i.e., state-of-the-art instruments facilitate aberration corrected imaging within a large magnification interval covering the length scales of interest, in particular the atomic regime. A general drawback of conventional TEM techniques is the loss of phase information originally contained in the scattered electron wave introduced by recording only the electron density. Electron Holography is an advanced TEM technique that facilitates the complete evaluation of the complex electron wave, which, in combination with the manifold possibilities of TEM, provides rather straightforward access to static electromagnetic fields within the ceramic. Nevertheless, quantification of order parameters such as the electric polarization or minute details in electromagnetic fields still require to correlate the experimentally gained observations to physical models, which combine the details of the microscopic imaging process, the electron-specimen scattering, and solid state physics of the specimen. The goal of this work is to investigate and advance the limits of Electron Holography as a truly quantitative TEM technique and apply the findings in, e.g., the investigation of ferroic ceramics. In the light of the previously mentioned difficulties, the problem has to be tackled from different directions: Firstly, the whole holographic imaging process is reviewed and extended, if necessary, in order to provide quantitative measures for systematic and statistical errors inherent to reconstructed waves. In the course of that process, two previously not recognized holography-specific aberrations are identified, firstly, a resolution limiting spatial envelope and secondly, a spatial distortion to the reconstructed wave. Furthermore, several correction strategies have been developed, in order to correct the aforementioned two and other well-known disturbances, e.g. Fresnel fringes from the biprism filament. The previous holographic noise model has been extended to incorporate the important contribution from the detector and consequently to provide realistic statistic error bars of the holographically reconstructed amplitude and phase. Secondly, an investigation of the electron-specimen scattering process itself is conducted, leading to a density matrix description of the holographic measurement. The general laws of quantum electrodynamics provide the framework of that description. Relativistic phenomena such as retardation of electromagnetic fields exchanged between beam electron and specimen and spin-orbit coupling of the beam electron are quantified, where the latter is found to be negligible within TEM. The decoherence of the electron wave by statistical coupling to the thermally moving crystal lattice of ceramics is treated by a newly developed algorithm facilitating in particular the accurate quantification of elastic scattering on heavy elements. Inelastic excitations in the ceramic, e.g. bulk plasmons or core electrons, are treated in combination with elastic scattering to identify their role in the holographic reconstruction process and to develop methods for an accurate calculation. A new scattering algorithm combining elastic and inelastic scattering is developed and applied to predict peculiar scattering contrasts of dipole transitions and to discuss the long-standing problem of contrast mismatch between scattering simulations and conventional imaging. To provide a user-friendly and continuing use of the findings, a software package SEMI (Simulation of Electron Microscopy Imaging) has been written, which facilitates the simulation of elastic and inelastic scattering processes and the subsequent imaging within different approximations, incorporating the newly developed algorithms. Thirdly, Density Function Theory (DFT) solid state calculations have been employed to identify and quantify structural modifications and characteristic electromagnetic fields, such as occurring at domain boundaries, within typical ferroic ceramics like BaTiO3 or BiFeO3, and concomitantly provide models correlating observables of the (holographic) experiment to characteristics of the materials, e.g. the order parameters. This is particularly important when static electromagnetic fields provide no direct information about the order parameter, e.g. the electric polarization, i.e., it is possible to correlate the measurable atomic positions to the electric polarization within linear response theory. A software package ATA (AuTomated Atomic contrast fitting) has been developed facilitating an automated fitting of atomic positions and a subsequent determination of local polarization. In a fourth step, electron holographic experiments analyzed with the help of the revised imaging process in combination with the knowledge gained from scattering theory are used as an input to the models established from solid state physics to yield quantitative information about bulk ferroelectric materials such as BaTiO3 and PbTiO3 and more complicated configurations such as domain walls in BiFeO3 and KnbO3. It is found that particular atomic shifts characteristic for ferroelectrics provide the most reliable quantitative information about the polarization down to nm length scales, whereas minute wave modification due to characteristic electron distributions within the ceramic are currently insufficiently quantitatively interpretable within Electron Holography. The linear response program, correlating atomic positions to ferroelectric polarization with the help of ab-initio calculated Born effective charges, has been successfully applied to determine finite size effects, screening layer widths and polarization charges in non-ferroelectric/ferroelectric layered systems. Finally, a special section considers the evaluation of 3D electromagnetic fields by Electron Holographic Tomography, which provides the means to characterize even more complex 3D domain wall configurations. As the capabilities of the technique are still limited by holographic reconstruction errors and particular tomographic issues such as incomplete projection data, the main focus of that section is put on the characterization and improvement of the tomographic reconstruction process. A Singular Value based reconstruction method is developed, which facilitates a quantification and control of the tomographic reconstruction error. Furthermore, vector field reconstruction is extended in order to treat magnetic vector fields leaking out from the reconstruction volume.
Degree
thesis:*- Level thesis:degree_level
- thesis.doctoral
- Grantor dc:publisher
- Technische Universität Dresden
- Year
- 2010
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Lubk, Axel
- Contributors dc:contributor
-
- Lichte, Hannes
- Kohl, Helmut