Live 3D imaging with sub-nanometer resolution

Transmission electron microscopy has had a tremendous impact on various scientific disciplines, including biology and materials science, as it allows to study various materials in great detail. Here, fast electrons shine through a thin sample imaging it at a very high resolution. The fact that the images obtained in this way only represent a projection of 3D objects is a major drawback of this technique. Therefore, tomography was developed to overcome this problem. Here, many images from different projections are collected in complicated experiments and combined with algorithms to form a 3D model. Tomography is time consuming and not suitable for direct live imaging of materials. Therefore, the goal of this project is to tilt the electron beam automatically very fast, and to transform the obtained images directly into a stereo image. The 3D-image can be directly observed by the spectator using suitable glasses, just like in a 3D cinema. Transmission electron microscopes use specialized sample holders, where miniaturized samples can be deformed, to directly observe the underlying damage processes in materials; for example, the movement of dislocations (a form of defect) in metals. The goal is to perform these experiments in 3D. This would allow to better understand the deformation properties of more complex materials. Despite great advances in data analysis, we still interact with the transmission electron microscope through watching images. 3D vision could be just a first step. The concept of augmented reality, in which, for example, data analysis is directly overlaid live with the image, would enable more efficient investigation of complex nanomaterials, accelerating the discovery of new materials.

Metals and alloys are used in a wide range of different applications. Decades of research have enabled to develop materials with improved mechanical properties. Strengthening and increasing the reliability of metallic parts is essential for lowering our carbon footprint, for example by enabling lighter cars, more efficient turbines, and longerlasting components. Although metallic materials consist of atoms arranged in a regular crystal lattice, it is the imperfections in this lattice that largely determine their properties. Plastic deformation, for instance, is controlled by tiny line like defects called dislocations, which move through the crystal when a material is stressed. To study these processes, researchers rely on transmission electron microscopy, which uses electrons instead of light to achieve very high resolution. Special holders allow to deform a very thin metal strip in the microscope and thus image individual dislocations as they move. In the present project, we use nanodiffraction mapping (also called 4D-STEM) to gain deeper insight into the driving forces and internal structure of dislocations during deformation. While traditional microscopy just gives you a simple movie of the dislocation, 4D-STEM yields a large dataset consisting of diffraction patterns, which directly reflect changes in atomic positions that can be interpreted as local lattice stresses. Traditionally such stress measurements require to orient the material such that it shows a well-defined atomic arrangement. Here we successfully applied this method to samples with differently oriented grains, significantly broadening its applicability. A further challenge is that dislocations are moving in three dimensions, whereas standard microscopy shows only a single projection. Although tilting the sample allows to reconstruct their 3D shape, this requires stopping the deformation experiment. Therefore, in the present project we have explored the use of tilting the electron beam. Our first results show that this is a feasible approach yielding important information on dislocation behavior. These new methodological developments open the door to a more complete understanding of dislocation motion in complex high-performance alloys. Ultimately, this knowledge can accelerate feedback between experiment, materials processing and simulation, helping to speed up the development of next-generation materials.