New capabilities in x-ray microscopy have been enabled by rapid advances in synchrotronlight sources and in high performance computing. We present here conceptual and computational advances in two areas of x-ray microscopy. One involves ptychography, where computation is used to obtain images from diffraction patterns as a finite coherent illumination beam is scanned. Because this method involves significant amounts of data, we demonstrate two methods that can be used for on-detector data compression, with compression factors ranging from about 1.5 to possibly even factors of a thousand. The other involves the reconstruction of the 3D distribution of particular elements via x-ray fluorescence tomography, in samples that are large enough that the experience some self-absorption of the fluorescence signal. We show that the method of automatic differentiation in numerical optimization can be used to solve both problems, enabling advances in x-ray microscopy.

Creator

• Huang, Panpan

DOI

• https://doi.org/10.21985/n2-777w-3k70

Subject

• Physics

Language

• en

Alternate Identifier

• etdadmin_upload_910750
• http://dissertations.umi.com/northwestern:16123

Keyword

• Ptychography
• Fluorescence tomography
• Data compression
• X-ray Imaging
• X-ray microscopy

Date created

• 2022-01-01

Resource type

• Dissertation

Rights statement

• In Copyright