Master Theses

Construction of an interatomic potential for the LiTiO battery anodes. Transition metal oxides are known to exhibit variable oxidation states. Therefore, they are preferred as storage materials in lithium ion batteries. We aim to construct a state-of-the-art interatomic potential for the LiTiO system, which will be used in atomistic Monte Carlo/molecular dynamics simulations. The atomistic framework incorporates Coulomb interactions, metallic bonding character via an embedded-atom formalism, as well as charge transfer due to the different oxidation states of the species. Varying oxidation states may induce different crystal structures which we could observe in recent experiments. We want to further explain this from an atomistic point of view. (supervisors: Sebastian Eich)

Machine learning concepts in atom probe tomography. In impressive cases, concepts of machine learning (ML) have been effective tools for materials science. The project tries to develop a computational algorithm for the measurement of particle/precipitate sizes in atom probe tomography data. The preferred machine learning strategy will be the “support vector regression”, which should be trained with data sets that are generated by simulated field evaporation. A simulation tool (TAPSim) is already available for this step. So, practically the project has to develop suitable Python scripts that control the available basic software. After training the KI with simulated data, the ML code should be applied to experimental data sets that were presented in one of our recent publications, (supervisors: Guido Schmitz, Sebastian Eich) 

Artificial intelligence in electron microscopy. A collaborative research initiative of the faculty addresses catalysis in mesoporous materials. In this context, we would like to image Metal-organic and covalent bonded frameworks (MOF, COF) by electron tomography. However, these materials are extremely sensitive to the electron beam, so only a few images can be obtained before the materials get destroyed. Missing information required for the tomographic reconstruction should be replaced by training the algorithms with theoretically expected Molecule structures. The project joins experimental work at the top-end microscope Spectra 300 with the development of Python scripts. Required materials will be provided by the working team of Bettina Lotsch (MPI-FKF) (supervisors: Guido Schmitz, Roham Jeid, Florian Kauffmann)

Lithium diffusion in LiCoO2 and its texture dependence. Diffusion and migration are the fundamental processes controlling the performance of a battery (provided there is enough space). The work will be focused on understanding the diffusion of lithium in thin-film LiCoO2 electrodes. Firstly, the thin films will be prepared using ion-beam sputtering. The deposition parameters will be optimized for reasonable electrochemical performance. Subsequently, with the knowledge gained from a recently concluded master study [1], the optical behaviour of the electrodes will be validated. This optical character will be used to study the diffusion of lithium in thin film LiCoO2 electrodes. [2] Furthermore, the required deposition parameters will be varied to achieve a variation in the texture of the thin film to see the effect of anisotropy in the diffusion for the layered electrode. (supervisor: Yug Joshi)

[1] https://doi.org/10.1149/1945-7111/ac63f6; [2] https://doi.org/10.1002/smtd.202100532

Thin film deposition of NMC 622 (LiNi0.6Mn0.2Co0.2O2) and its optical behavior. The change of optical constants (real and imaginary refractive indices) upon lithium insertion has been reported for LiMn2O4 [1], Li4Ti5O12 [2] and LiCoO2 [3] indicating that the insertion of lithium causes the shift of the band gap due to electrostatic repulsion caused by lithium. Furthermore, the gradient of the change of dielectric constants (obtained for the optical constants) with lithium insertion depends on the microstructure of the thin films and helps in revealing the phase propagation in a battery electrode. From the already developed methodologies, thin films of NMC622 (the current state-of-the-art material used in battery electrodes) will be prepared via ion-beam sputtering, followed by structural, electrochemical and optical characterization. The model description of the optical behaviour will be used to obtain the optical constants. Subsequently, its dependence on lithium insertion will be studied and its correlation to the microstructure of the thin film will be attempted. supervisor: (Yug Joshi)

[1] https://doi.org/10.1002/adom.201701362; [2] https://doi.org/10.1021/acsami.9b19683; [3] https://doi.org/10.1149/1945-7111/ac63f6

Hydrogen transport and embrittlement. Measurement of hydrogen permeation by high-resolution gravimetry. Hydrogen storage, in the form of metallic hydrides, is a promising concept since it enables particularly high volume density in H storage. With this aim, we would like to study the hydrogenation of FeTi alloys, since these alloys are available for quite economical prices. We would like to understand the hydrogen uptake. An innovative method of measurement shall be applied for the first time. We deposit FeTi layers as thin films to oscillating quartz microbalances and measure the gain in weight when absorbing hydrogen in the metallic films. Furthermore, the formation of the hydride phase leads to expansion due to the reaction’s excess volume. Thus, nuclei of the hydride will become visible as surface protrusions, which we will image by white light interferometry to measure their number and size in depending on the hydrogenation treatment. (supervisors: Yug Joshi, Guido Schmitz)

Detection of hydrogen by atom probe tomography. Future hydrogen technology needs new construction materials for the 700 to 1000 bar hydrogen pressure range. Hydrogen embrittlement is a phenomenon feared when using steel constructions in hydrogen atmosphere. In order to progress the understanding of this phenomenon, one needs to measure tiny amounts of hydrogen in high spatial resolution. We would like to explore atom probe tomography for this. Cryo-preparation and direct transfer into the atom probe should prevent hydrogen from escaping before the measurement. We will first test and benchmark the method at NiTi multilayers, where hydrogen should preferentially segregate to the Ti layers. Afterwards, new high-entropy alloys should be analyzed. Are they single- or multiphase? In which phase or structural defects does hydrogen prefer to segregate? (supervisors: Patrick Stender, Guido Schmitz)

Reactive wetting on quantitatively evaluated rough surfaces. In a previous PhD work, we have developed a new thermodynamic model to understand the spreading of liquid solder metals on rough substrates when downscaling the size of the solder connections. We plan to critically check this theory. Nano-structured rough surfaces are produced by electron beam lithography in our FIB-microscope, for which we exactly know the amplitude (RMS) and periodicity of the roughness. On these, the spreading of solder droplets with a size downward to about 10 micrometers is studied by in-situ SEM investigation. (Supervision: Guido Schmitz)

Pump-probe experiments in atom probe tomography. In Atom Probe Tomography (APT), very fine samples are evaporated as tiny molecular fragments by superposition of high electric field and laser pulses. Especially with organic matter, the size and the charge state of the evaporated fragments have not been understood, yet. Likewise, oxygen and nitrogen can be lost due to insufficient ionization. The aim of the project is to investigate whether a second, subsequent laser pulse, that hits the desorbed species right after leaving the sample surface, can positively influence the level of ionization and the fragmentation through multi-photon ionization. This requires laser powers > 10²⁰ W/cm², which we plan to obtain with ultra-short (40 fs), highly focused laser pulses. (supervisors: Patrick Stender, Guido Schmitz)

Development of cryo-APT preparation techniques. Samples for Atom Probe Tomography must be in a needle-like shape with a radius of curvature of 50 nm at the apex. Preparation of conventional materials is available by using Focused Ion Beam techniques. Currently, liquids are of great interest, with the goal of studying processes at biological interfaces like cell membranes, viruses or bacteria in atomic resolution. Aqueous solutions are cryo-frozen and transferred into the FIB for tip shaping and finally to the atom probe. The sample must be permanently cooled to preserve them in solid state. To avoid artifacts, a crystallization has to be prevented. Through innovative ink-jet printing we try to avoid a crystallization of ice crystals. The goal of the study is the measurement of a virus by atom probe tomography. (supervisor: Patrick Stender)

High-resolution TEM analysis of the grain boundaries in LiFePO₄ cathodes. In previous studies, we collected indicators that the grain boundaries (GB) of LiFePO₄ cathodes must be of special character. Probably, they represent purely ion-conductive pathways in the mixed electronic-ionic conducting matrix. We would like to elucidate this issue by the analysis of the GBs in the highest possible resolution. The cutting-edge, aberration-corrected electron microscope Spectra 300 will be applied. It allows imaging in world-leading resolution and chemical analysis of local oxidation states via electron energy loss spectroscopy. The cathode material will be produced by thin film deposition. Electron transparent slices are prepared by focused-ion-beams in the lift-out technique. The project is especially suited for those who like to work with top instrumentation tools. (supervisors: Guido Schmitz, Yug Joshi)

Analysis of diffusional transport in mesoporous matrices filled by organic liquids. A research collaboration of the faculty addresses catalysis in confined geometry of cylindrical pores, a few nanometers in diameter. The project shall produce such pores in thin layers of silica by spin coating and sol/gel methods (in collaboration with J. Bruckner of the IPC) fill with organic liquids (Bromo- and Chlorobenzol, Dodekan), cryo-freeze after defined diffusion times and measure composition profiles of the diffusing species by high resolution scanning electron microscopy in our dual beam/FIB microscope at 140K. (Supervisor: G. Schmitz)