Automatized Diffraction Pattern Recognition for Scanning Surface X-Ray Crystallography of Polycrystalline Materials

Supervisors: Prof. Andreas Stierle (DESY/UHH), Dr. Vedran Vonk (DESY), Dr. Nastasia Mukharamova (DESY), Prof. Peer Stelldinger (HAW)

Understanding surface structure-property relationships in polycrystalline (structural) materials requires state-of-the-art characterization resolving atomic (near) surface structure locally on the sub-grain level. It is largely unexplored how the crystallographic grain surface structure influences processes like oxidation, corrosion, carbide formation or friction.
As showcases, we will investigate polycrystalline Nb surfaces employed in superconducting radio frequency cavities for high gradient field linear accelerators and the surfaces of Fe-based tooling alloys. Varying grain orientation and grain boundaries are key elements influencing the materials' performance. We propose to develop a novel x-ray scanning technique allowing to study the surfaces of polycrystalline samples. It makes use of sub-micron x-ray beams, to locally record surface sensitive diffraction patterns. The computational challenge of the proposed method lies in the fact that each locally obtained diffraction pattern needs to be automatically analyzed and indexed according to the grain orientation. A typical data set contains about 500 diffraction patterns of about 100 grains, whereby each diffraction pattern consists of about 1800 2D detector images. Here we propose to develop a machine learning algorithm to extract the surface structure information from the large data set. In addition, it will be necessary to develop adequate data treatment pipelines from the synchrotron beamline.