Hard x-ray nano-XANES and implementation deep learning tools for multi-modal chemical imaging

Abstract

Spectromicroscopy techniques allow the study of local chemical states along with morphology information. At the hard X-ray nanoprobe (HXN) beamline at NSLS-II, we developed nanoscale chemical imaging with high chemical state sensitivity and micron-scale penetration depth. In addition to the chemical images, XRF and phase-contrast images collected simultaneously offer multi-modal, correlative image analysis. We also developed a highly interactive, python-based graphical user interface (NSLS-II MIDAS) that allows multi-modal analysis of nano-XANES and XRF images. Advanced supervised and unsupervised learning algorithms enable users to explore the traditional XANES analysis along with standard machine-learning tools