Development of interferometric X-ray Imaging Techniques for the Assessment of Functional Layers for future clean energy production on industrial scale

INTER-EXCELLENCE II

Abstrakt

In this project, we will develop an instrument that can be used to assess the microstructure of functional layers in membrane electrolyte assemblies, which are key components in clean energy systems (fuel cells, electrolysis systems) based on hydrogen. These functional layers exhibit micro-structures whose integrity is essential for efficiency and longevity. With our instrument, we will be able to assess the quality of these layers and detect defects in a non-destructive way. It will be designed in a way that even future large area samples of the functional layers can be monitored for quality. To achieve this goal, we will combine interferometric X-ray imaging with a scanning technique and with the cutting-edge photon counting pixel detectors of Timepix3 and Timepix4 technology, both being able to provide energy information about each detected photon passing the interferometers. Machine learning will help us to derive distributions of micro-structure scales of the object from this energy information with improved sensitivity and image contrast and to detect defects which might influence the energy efficiency or lifetime of future hydrogen-based clean energy systems.

Původní abstrakt

The proposed research project aims at imaging of functional layers in the membrane electrolyte assembly (MEA) of fuel cells and electrolysis systems for hydrogen use and production, which are important

elements in the transition to a decarbonized society. The optimized imaging technique shall be usable in both, R&D in the laboratory to develop manufacturing processes and for quality assurance in future large scale production processes to increase efficiency and longevity of produced clean energy systems. While currently used detectors can determine the photon energy spectra in a limited number of bins, the cutting-edge hybrid pixel detectors proposed will provide a continuous photon spectrum within each of the pixels. By this, we expect higher sensitivity of the imaging modality and improved categorization of the defect or crack density in the produced images in particular when combined with state-of-the art machine learning techniques. After determination and quantification of the performance using reference samples, the operational parameters will be refined for the defect detection in the functional layers of the membrane electrolyte assembly (MEA) providing valuable feedback for optimizing the production process.