Phase imaging in the Transmission Electron Microscope (TEM) has a long history, from the implementation of off-axis holography in TEM to Differential Phase Contrast (DPC) on the Scanning Transmission Electron Microscopy (STEM). The advent of modern computing has enabled the development of iterative algorithms which attempt to recover a phase image of a specimen from measurements of the way it diffracts an incident electron beam. One of the most successful of these iterative methods is focused probe ptychography, which relies on far field diffraction pattern measurements recorded as the incident beam is scanned through a grid of locations across the specimen. Focused probe ptychography implemented in the STEM has provided the highest resolution images available to date, allows for lens-less setups avoiding the aberrations typical in older STEMs and allows for simultaneous reconstruction of the illumination and specimen. Ptychography is computationally flexible (highly constrained), allowing for additional unknowns other than the phase of the specimen to be recovered, for example positions can be refined during reconstruction.
Near field ptychography is a recent variation on ptychography that replaces the far-field diffraction data with diffraction patterns recorded in the near field, or Fresnel, region. It promises to obtain a much larger field of view with fewer diffraction patterns than focused probe ptychography. The main contribution of this thesis is the implementation of a new form of near field ptychography on the Transmission Electron Microscope (TEM), using an etched silicon nitride window to structure the electron beam. Proof-of-concept results show the method quantitatively recovers megapixel phase images from as few as 9 recorded diffraction patterns, compared to many hundreds of diffraction patterns required for focused probe ptychography. Additional sets of results show how near-field ptychography can recover extremely large fields of view, deal effectively with inelastic scattering, and accommodate several sources of uncertainty in the experimental process.
Further contributions in the thesis include: experiments and results from visible-light versions of near field ptychography, which explain its limitations and practical application; a description and code for analysis tools that are used to assess phase imaging performance; DigitalMicrograph (DM) code and a data collection workflow to realise TEM-based near-field ptychography; details of the design, realisation and performance of the etched silicon nitride windows; and simulation studies aimed at furthering understanding of the frequency response of the technique. Future work is outlined, focusing on potential applications in a wide range of real-world specimens and improved TEM setups to implement near field ptychography.
