The Hard X-ray Nanoprobe (HXN) at NSLS-II provides a nanoscale 3D multi-modality imaging capability, useful for investigating diverse material systems. The multi-modality scanning-probe imaging utilizes a variety of imaging contrasts such as fluorescence, transmission, scattering, and diffraction. Images taken simultaneously using different contrast mechanisms can provide 3D visualization of a sample, producing complementary information about the sample. Such comprehensive 3D characterizations are extremely useful in studying materials with multiple phases or complex internal structures. An important scientific problem is to investigate phase or grain boundaries of multi-component materials during or after material processing such as sintering, since re-organization of these boundaries due to annealing or phase-separation often result in profound impact on material property or functionality. However, accurate quantification of 3D elemental concentration is hampered by a well-known self-absorption problem, particularly severe for the low energy fluorescence x-rays. Correcting absorption is non-trivial and requires an iterative and three-dimensional solution. In this presentation, we will describe our approach using experimental data taken from mixed ionic ceramic membrane samples and elaborate on how accurate absorption correction led to discovery of a new material phase in this material system.
We developed a python-based fluorescence analysis package (PyXRF) at the National Synchrotron Light Source II (NSLS-II) for the X-ray fluorescence-microscopy beamlines, including Hard X-ray Nanoprobe (HXN), and Submicron Resolution X-ray Spectroscopy (SRX). This package contains a high-level fitting engine, a comprehensive commandline/ GUI design, rigorous physics calculations, and a visualization interface. PyXRF offers a method of automatically finding elements, so that users do not need to spend extra time selecting elements manually. Moreover, PyXRF provides a convenient and interactive way of adjusting fitting parameters with physical constraints. This will help us perform quantitative analysis, and find an appropriate initial guess for fitting. Furthermore, we also create an advanced mode for expert users to construct their own fitting strategies with a full control of each fitting parameter. PyXRF runs single-pixel fitting at a fast speed, which opens up the possibilities of viewing the results of fitting in real time during experiments. A convenient I/O interface was designed to obtain data directly from NSLS-II’s experimental database. PyXRF is under open-source development and designed to be an integral part of NSLS-II’s scientific computation library.