@phdthesis{Bayer2024,
  author    = {Bayer, Florian},
  title     = {Investigating electromagnetic properties of topological surface states in mercury telluride},
  doi       = {10.25972/OPUS-35212},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-352127},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2024},
  abstract  = {This doctoral thesis investigates magneto-optical properties of mercury telluride layers grown tensile strained on cadmium telluride substrates. Here, layer thicknesses start above the usual quantum well thickness of about 20 nm and have a upper boundary around 100 nm due to lattice relaxation effects. This kind of layer system has been attributed to the material class of three-dimensional topological insulators in numerous publications. This class stands out due to intrinsic boundary states which cross the energetic band gap of the layer's bulk. In order to investigate the band structure properties in a narrow region around the Fermi edge, including possible boundary states, the method of highly precise time-domain Terahertz polarimetry is used. In the beginning, the state of the art of Teraherz technology at the start of this project is discussed, moving on to a detailed description and characterization of the self-built measurement setup. Typical standard deviation of a polarization rotation or ellipticity measurement are on the order of 10 to 100 millidegrees, according to the transmission strength through investigated samples. A range of polarization spectra, depending on external magnetic fields up to 10 Tesla, can be extracted from the time-domain signal via Fourier transformation. The identification of the actual band structure is done by modeling possible band structures by means of the envelope function approximation within the framework of the k·p method. First the bands are calculated based on well-established model parameters and from them the possible optical transitions and expected ellipticity spectra, all depending on external magnetic fields and the layer's charge carrier concentration. By comparing expected with measured spectra, the validity of k·p models with varying depths of detail is analyzed throughout this thesis. The rich information encoded in the ellipitcity spectra delivers key information for the attribution of single optical transitions, which are not part of pure absorption spectroscopy. For example, the sign of the ellipticity signals is linked to the mix of Landau levels which contribute to an optical transition, which shows direct evidence for bulk inversion asymmetry effects in the measured spectra. Throughout the thesis, the results are compared repeatedly with existing publications on the topic. It is shown that the models used there are often insufficient or, in worst case, plainly incorrect. Wherever meaningful and possible without greater detours, the differences to the conclusions that can be drawn from the k·p model are discussed. The analysis ends with a detailed look on remaining differences between model and measurement. It contains the quality of model parameters as well as different approaches to integrate electrostatic potentials that exist in the structures into the model. An outlook on possible future developments of the mercury cadmium telluride layer systems, as well as the application of the methods shown here onto further research questions concludes the thesis.},
  subject      = {Quecksilbertellurid},
  language  = {en}
}
@phdthesis{Hauschild2015,
  author    = {Hauschild, Dirk},
  title     = {Electron and soft x-ray spectroscopy of indium sulfide buffer layers and the interfaces in Cu(In,Ga)(S,Se)2-based thin-film solar cells},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-126766},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2015},
  abstract  = {In this thesis, thin-film solar cells on the basis of Cu(In,Ga)(S,Se)2 (CIGSSe) were investigated. Until today, most high efficient CIGSSe-based solar cells use a toxic and wetchemical deposited CdS buffer layer, which doesn't allow a dry inline production. However, a promising and well-performing alternative buffer layer, namely indium sulfide, has been found which doesn't comprise these disadvantages. In order to shed light on these well-performing devices, the surfaces and in particular the interfaces which play a major role for the charge carrier transport are investigated in the framework of this thesis. Both, the chemical and electronic properties of the solar cells' interfaces were characterized. In case of the physical vapor deposition of an InxSy-based buffer layer, the cleaning step of the CdS chemical-bath deposition is not present and thus changes of the absorber surface have to be taken into account. Therefore, adsorbate formation, oxidation, and segregation of absorber elements in dependence of the storing temperature and the humidity are investigated in the first part of this thesis. The efficiencies of CIGSSe-based solar cells with an InxSy buffer layer depend on the nominal indium concentration x and display a maximum for x = 42 \%. In this thesis, InxSy samples with a nominal indium concentration of 40.2\% ≤ x ≤ 43.2\% were investigated by surface-sensitive and surface-near bulk-sensitive techniques, namely with photoemission spectroscopy (PES) and x-ray emission spectroscopy (XES). The surfaces of the films were found to be sulfur-poor and indium-rich in comparison with stoichiometric In2S3. Moreover, a direct determination of the band alignment at the InxSy/CISSe interface in dependence of the nominal indium concentration x was conducted with the help of PES and inverse PES (IPES) and a flat band alignment was found for x = 42 \%. In order to study the impact of a heat treatment as it occurs during subsequent cell process steps, the indium sulfide-buffered absorbers were annealed for 30 minutes under UHV conditions at 200 °C after the initial data set was taken. Besides a reported enhanced solar cell performance, a significant copper diffusion from the absorber into the buffer layer takes place due to the thermal treatment. Accordingly, the impact of the copper diffusion on the hidden InxSy/CISSe interface was discussed and for x = 40.2\% a significant cliff (downwards step in the conduction band) is observed. For increasing x, the alignment in the conduction band turns into a small upwards step (spike) for the region 41\% ≤ x ≤ 43.2\%. This explains the optimal solar cell performance for this indium contents. In a further step, the sodium-doped indium sulfide buffer which leads to significantly higher efficient solar cells was investigated. It was demonstrated by PES/IPES that the enhanced performance can be ascribed to a significant larger surface band gap in comparison with undoped InxSy. The occurring spike in the Na:InxSy/CISSe band alignment gets reduced due to a Se diffusion induced by the thermal treatment. Furthermore, after the thermal treatment the sodium doped indium sulfide layer experiences a copper diffusion which is reduced by more than a factor of two compared to pure InxSy. Next, the interface between the Na:InxSy buffer layer and the i-ZnO (i = intrinsic, non-deliberately doped), as a part of the transparent front contact was analyzed. The i-ZnO/Na:InxSy interface shows significant interdiffusion, leading to the formation of, e.g., ZnS and hence to a reduction of the nominal cliff in the conduction band alignment. In the last part of this thesis, the well-established surface-sensitive reflective electron energy loss spectroscopy (REELS) was utilized to study the CIGSSe absorber, the InxSy buffer, and annealed InxSy buffer surfaces. By fitting the characteristic inelastic scattering cross sections λK(E) with Drude-Lindhard oscillators the dielectric function was identified. The determined dielectric functions are in good agreement with values from bulk-sensitive optical measurements on indium sulfide layers. In contrast, for the chalcopyrite-based absorber significant differences appear. In particular, a substantial larger surface band gap of the CIGSSe surface of E^Ex_Gap = (1.4±0.2) eV in comparison with bulk values is determined. This provides for the first time an independent verification of earlier PES/IPES results. Finally, the electrons' inelastic mean free paths l for the three investigated surfaces are compared for different primary energies with theoretical values and the universal curve.},
  subject      = {Photoelektronenspektroskopie},
  language  = {en}
}