We are hiring: PhD students and postdocs welcome
Due to recent successful funding applications within the frameworks of the European Research Council (ERC) and the DFG, our group offers several positions for both PhD projects and at postdoc level. The start date is flexible, our intent is to fill the positions as soon as possible. The scientific field is 4D STEM with focus on the research lines mentioned below briefly. Candidates with experimental and theoretical background within the natural sciences are equally welcome - and needed! All projects act within a broad consortium of interdisciplinary partners, ranging from activities within our DFG-excellence cluster (e-conversion, TUM & LMU) to large-scale European projects (4D-BioSTEM Synergy project) to multiple bilateral international cooperations. This offers a solid and wide network for scientific, social and cultural interaction, exchange, training and development.
For questions and applications, contact Prof. Dr. Knut Müller-Caspary (k.mueller-caspary@cup.lmu.de).
- PhD projects. 4D-STEM data can be evaluated in a plenitude of ways, electron ptychography methods being one of the most promising families currently. This includes direct reconstruction schemes as well as iterative algorithms. They have in common that the goal is to retrieve the complex object transmission function of the specimen, whose phase is proportional to the projected potential created by the atoms. In that respect, different ptychographic methods are suitable for different challenges in nanoscience. In several PhD projects framed below, application of existing ptychography schemes, their optimisation for the investigation of nanostructures in the Physical and Life Sciences, further developments of phase retrieval methodologies and their implementation are the most important research foci.
- Project 1: Structure-property relationships in IT nanostructures. The optical properties of semiconductor heterostructures such as core-shell nanowires fundamentally depend on the local chemistry, and the sharpness of interfaces. Electrical transport and electronic band alignment in 2D material heterostructures are governed by the distribution of vacancies and substitutional atoms. The domain structure in ferroelectric crystals is key to the performance of future high-density data storage devices. In this project, parameters for the data acquisition of 4D-STEM data at the microscope are first optimized in a theoretical study. This is then put into practice by collecting experimental 4D-STEM data of nanowires, 2D materials and ferroelectric crystals at a contemporary electron microscope with an ultrafast detector running at 8000 frames per second. Ptychographic evaluations using existing concepts for so-called "inverse multislice" then yield chemical composition maps in nanowire heterostructures, the local atomic structure of 2D heterobilayers, and the local ionic displacements in ferroelectrics with picometre precision. The results create impact by the high-level structure retrieval with superresolution itself, and especially by correlation with photoluminescence and piezoelectric force microscopy measurements performed by our cooperation partners at TU Munich (Walter Schottky Institute), at LMU Munich (Physics department) and in the US, since they allow unique insights into the structure-property relationships.
Your background should be in physics, physical or theoretical chemistry. Solid-state knowledge and programming skills (basic python, matlab or similar) are appreciated. - Project 2: Low-dose ptychography of organic nanostructures. Covalent and metal organic frameworks (MOFs, COFs), perovskite solar cells, and origami DNA constitute established research lines in energy conversion science. Common to these structures is their fragility under electron-beam irradiation, which breaks the bonds and leads to structural degradations even at very low doses. Therefore, using electron ptychography and optimising 4D-STEM acquisition conditions such that the maximum structural information at the given dose budget is obtained is the kernel of this project. A unique feature is the cooperation with specialists for Cryo-4D-STEM at Forschungszentrum Jülich (Prof. Dr. Carsten Sachse) and EPFL Lausanne (Prof. Dr. Henning Stahlberg). This project includes thus both, on-site research with state-of-the-art microscopy hardware for Physical Sciences, and strong interdisciplinary activities to elucidate the benefits of Life Science hardware for Physical and Materials Sciences. Upon mutual agreement, co-supervision in an interdisciplinary framework is possible.
Your background should be in the natural sciences, such as chemistry, physics or biology (with experience in structural characterisation techniques in the latter case). The project has an experimental focus. - Project 3: Precision and accuracy of ptychographic techniques. Any experimental data is noisy due to several reasons: Electron detection is discrete, and at low doses counting statistics plays an important role to derive the precision of a measured parameter (e.g. the position of an atom). Further sources of noise arise from detector readout, or so-called structural noise in cryo-electron microscopy. In the latter case, the embedment of specimens in amorphous ice leads to noisy images, where the structure of the object of interest is superimposed to a background stemming from the ice. Importantly, structure retrieval by electron ptychography involves highly nonlinear processing of the raw data, which poses a severe challenge for tracking the precision through all steps. Therefore, defining a confidence interval or error margins for the obtained structure of a protein or COF is an outstanding task tackled in this project. You perform simulation studies with known ground truth and feed it into the evaluation pipelines of different ptychography schemes imposing different noise levels and characteristics. A dose corridor for the attainable resolution and structural fidelity of ptychographic reconstructions is to be established in this way. You then use real cryo- and conventional 4D-STEM data obtained by colleagues locally and by partner groups at Forschungszentrum Jülich (Prof. Dr. Carsten Sachse) and EPFL Lausanne (Prof. Dr. Henning Stahlberg) to obtain consistency with experiments. This project provides a key framework to elucidate the theoretical and methodological limitations for ptychography in dependence of the electron dose. It provides the opportunity to work closely with method developers in the field of structural biology so as to make use of Physical Sciences methods in Life Sciences.
Your background should be in physical chemistry, physics or theoretical chemistry. Experience in python and/or matlab programming is welcome, as well as knowledge about methods for structural characterisation.
- Project 1: Structure-property relationships in IT nanostructures. The optical properties of semiconductor heterostructures such as core-shell nanowires fundamentally depend on the local chemistry, and the sharpness of interfaces. Electrical transport and electronic band alignment in 2D material heterostructures are governed by the distribution of vacancies and substitutional atoms. The domain structure in ferroelectric crystals is key to the performance of future high-density data storage devices. In this project, parameters for the data acquisition of 4D-STEM data at the microscope are first optimized in a theoretical study. This is then put into practice by collecting experimental 4D-STEM data of nanowires, 2D materials and ferroelectric crystals at a contemporary electron microscope with an ultrafast detector running at 8000 frames per second. Ptychographic evaluations using existing concepts for so-called "inverse multislice" then yield chemical composition maps in nanowire heterostructures, the local atomic structure of 2D heterobilayers, and the local ionic displacements in ferroelectrics with picometre precision. The results create impact by the high-level structure retrieval with superresolution itself, and especially by correlation with photoluminescence and piezoelectric force microscopy measurements performed by our cooperation partners at TU Munich (Walter Schottky Institute), at LMU Munich (Physics department) and in the US, since they allow unique insights into the structure-property relationships.
- Postdoc position (TV-L 13 100%). Especially the reconstruction of 3D volumes by inverse multislice is highly topical with respect to both applications and methodologically. This refers not only to nanoparticles and -wires in materials science, but also to cell tissue in the life sciences. 3D imaging from a single projection acquired with a 4D-STEM setup could either be used instead of tomography, or to stabilise tomographic reconstructions. The convergence of ptychographic methods to yield 3D information employing different regularisation schemes is to be studied systematically here. The positions offers tight connection to the PhD projects above, and involvement in the supervision and topical steering upon agreement.
You bring experience in TEM, ideally 4D-STEM, tomography or ptychographic techniques. Programming skills are welcome (basic-advanced python, matlab, C++).
Background: 4D-STEM
The innovative character of the technique "4D-Scanning TEM" (4D-STEM) arises from the simultaneous availability of data in real and diffraction space. The technique has been put into practice by the introduction of ultrafast cameras to TEM delivering many thousands of images per second, partly pioneered by our group. The data itself is interesting in many respects being at the heart of fundamental physics, nanoscience, organic chemistry and structural biology. It especially enables
- Super-resolution. The resolution of microscopes is usually fundamentally limited by its optics, i.e. the quality of the lenses and apertures used. A famous resolution limit for coherent imaging is the Abbe limit λ/sin(θ) with the electron wavelength λ and the maximum scattering angle used for image formation θ. Incoherent image formation reaches a factor of 2 better resolution, even. However, although electron wavelengths are as small as 2 pm, the direct resolution of electron microcopes is limited to ≈50 pm, because θ needs to be kept small due to unavoidable lens aberrations. By applying so-called ptychographic methods to 4D-STEM data, the spatial resolution is henceforth only limited by the thermal vibration amplitude of the atoms (Chen et al., Science 372, 826-831).
- Versatility. Access to the full diffraction pattern allows for ultimate flexibility to form different types of images to visualise different types of specimen information. By defining "virtual" detectors, imaging is nowadays performed via software in the computer rather than directly at the microscope. This includes Z-contrast due to varying chemical composition by evaluating high scattering angles, strain contrast due to crystal lattice distortions by using medium angles, or even the mapping of electric fields inside atoms via advanced schemes, that is, the centre of mass of the diffraction pattern.
- Dose efficiency. As luck would have it, the 4D data sets are not only capable of mapping chemistry, structure, strain, light elements and electric/magnetic fields - they additionally require only a fraction of the dose during data acquisition. This opens a completely new field for scanning TEM: The investigation of materials and biological matter being very sensitive to electron irradiation. Among these are metal- and covalent organic frameworks (MOFs/COFs), origami DNA, proteins, viruses and cells studied in our group and with cooperation partners from the Cryo-TEM community.
- Inverse Problem. Recording an image or diffraction pattern is described by the squared modulus of the quantum mechanical electron wave function in the detector plane. Therefore, the phase information is lost entirely in each single recording. Retrieval of the phase information, which carries crucial and in many cases the most interesting information about the specimen, is subject to the so-called inverse problem as one of the most fundamental problems in physics and mathematics. The combination of diffraction and real space information in 4D-STEM transforms phase retrieval from an ill-posed task in a single recording to a well-posed challenge. Suitable algorithms, partly developed in our group, are able to recover the phase even in the presence of multiple scattering, delivering 3D information about thick specimens.
Last update: December 2023