Illumination diversity in multi-wavelength extreme ultraviolet ptychography

With the development of high harmonic generation (HHG), lensless extreme-ultraviolet (XUV) imaging at nanoscale resolution has become possible with table-top systems. Specifically, ptychographic phase retrieval using monochromatic XUV illumination exhibits extraordinary robustness and accuracy to computationally reconstruct both the object and the illumination beam profile. In ptychography, using structured illumination has been shown to improve reconstruction robustness and image resolution by enhancing high-spatial-frequency diffraction. However, broadband imaging has remained challenging, as the required multi-wavelength algorithms become increasingly demanding. One major aspect is the ability to separate the available information into different physically meaningful states, such as different spectral components. Here we show that introducing spatial diversity between spectral components of a HHG beam can significantly improve multi-wavelength XUV ptychography. We quantify the diversity in the polychromatic illumination by analyzing the diffraction patterns using established geometry- and information theory-based dissimilarity metrics. We experimentally verify the major influence of diversity by comparing ptychography measurements using HHG beams with Gaussian and binary structured profiles, as well as with beams carrying wavelength-dependent orbital angular momentum. Our results demonstrate how structured illumination acts in a twofold way, by both separating the spectral information in a single diffraction pattern while providing maximized added information with every new scan position. We anticipate our work to be a starting point for high-fidelity polychromatic imaging of next-generation nanostructured devices at XUV and soft-X-ray wavelengths.