Optimization of analysis of soft X-ray spectromicroscopy at the Ca 2p edge

Absorption saturation frequently occurs in X-ray absorption spectroscopy whenever transmission detection is used and the sample is too thick. This can strongly distort X-ray absorption spectra, particularly for transitions which have very high intensity and narrow line width relative to the instrumental resolution. This problem is particularly severe at the strong 2p -> 3d resonances at the Ca 2p edge, such that there are detectible absorption saturation effects for mineral samples as thin as 30 nm. We describe methods to identify such spectral distortion in scanning transmission X-ray microscopy (STXM) and avoid its effects when generating reference spectra by selecting spatial regions with undistorted spectra. We also show that taking absorption saturation into account is critical when STXM is applied to systems with thick regions. We outline an approach which can provide accurate thickness maps even in the regions where the thickness is such the absorption is saturated at the resonance peaks. Environmental samples such as biofilms, soil samples, sediments, and precipitates are typically heterogeneous in both composition and thickness. They often contain regions thicker than the saturation threshold, yet one would still like to analyse such samples correctly and quantitatively. By excluding the peak energies where absorption saturation distortion occurs, we show it is possible to achieve a meaningful quantitative analysis. We demonstrate the advantages of this approach in the context of mapping two different CaCO3 species biomineralized within a natural river biofilm. (C) 2009 Elsevier B.V. All rights reserved.