Factors Influencing Mean Inner Potentials As Studied Using Electron Holography and Density Functional Theory

In this dissertation, factors affecting electron holographic measurements of the mean inner potential are explored. Electron holography in the transmission electron microscope (TEM) allows for quantitative retrieval of the amplitude and phase of the electron beam. Both the amplitude and the phase reflect properties of the specimen. The phase can yield quantitative measurements of nanoscale electric and magnetic potentials. One such electrostatic potential is called the mean inner potential. The mean inner potential is the average electrostatic potential measured between the bulk of a material and vacuum far from the specimen, and is non-zero for all materials. However, previous mean inner potential measurements have disagreed for the same material measured by different groups. Additionally, experiment and image simulation are known to differ for high-resolution TEM, but not for electron holography, and this difference is known as the Stobbs factor - therefore, a dataset that allows for good comparison between image simulation and experiment might highlight possible improvements in the simulation software.

The factors that affect the mean inner potential are explored through both experiment and simulation. Thickness measurements using different tomographic algorithms (algebraic, geometric, and discrete tomography) and non-tomographic methods are compared on an InAs nanowire. A self-calibrating tilt-series of holograms on the same InAs nanowire is acquired and compared with image simulations to analyze diffraction effects on the amplitude and the phase. There is relatively good comparison between image simulation and experimental data, but the experimental absorption parameter is found to differ between strongly and weakly diffracting conditions. Density functional theory simulations of the mean inner potential are carried out using the GPAW program, allowing for exploration of the surface dependence of the mean inner potential. Factors including surface facet, structure optimization (atomic position relaxation), adsorbates, and fringing fields at corners are all examined. Finally, surface modification is attempted on an InAs/InP nanowire, but leads to nanowire dissolution instead - this dissolution is briefly characterized for GaAs nanowires.