Balder beamline is dedicated to hard X-ray spectroscopy at the MAX IV Laboratory with capacity to record sub-second X-ray absorption spectra. “Rapid hyperspectral mapping” is tailored to researchers with cm sized samples with chemical heterogeneity of 50-100 µm, for example, growth rings of a tree, and particles in bottom ash. The method is also efficient to get good spectral quality of low concentration element unevenly dispersed in a sample. The new development combining XANES (chemical state) and XRF (elemental) mapping will give the local chemical structure related to the geometric structure for the non-homogeneous sample. Using the energy dispersive fluorescence detector, we are able to capture both the XANES of target element(s) and the concentrations and distribution of other excited elements on a 2D map. We are exploiting the fast monochromator system and hardware orchestrated energy scanning at Balder by taking the XANES spectra for each x-y pixel of the sample before moving to the next spatial position. Resulting in a multidimensional map array based on high energy resolved XANES in each spatial position and a full XRF map in each energy point within the XANES scan at each pixel. To explore the chemical and elemental correlations of the 2D map, we have developed a fast software (XHyView) that allow the fast manipulation of the data and can project a map over any 4D florescence-energy-xy group divided by any other such group (e.g a normalization group), and produce plots of any integral on any such 4D group. So that the relative density of two elements or the relative XANES to an element can be viewed, or the density of an element relative to a low energy group and high energy group can be seen, which shows if that element is primarily on the surface. Additionally we have added standard features like PCA and data extraction for live and offline analysis. The mapping strategy exceedingly reduces scanning time and radiation dose while providing better spectral resolution compared to traditional stacking strategies in XANES mapping.
Kajsa Sigfridsson Clauss (Mon,) studied this question.