High Resolution X-Ray Spectra for Chemical Speciation in the SEM

A major focus of ongoing work will be to seamlessly integrate the data acquisition with a nearly simultaneous co-adding of spectra, energy calibration, and spatial mapping of the spectrum associated with each pixel obtained by the SEM.A machine learning effort towards an accurate and efficient energy calibration is a near-term goal of the project. We plan to implement a machine learning engine utilizing pattern recognition for key spectral features. The engine will identify major lines based on spacing and absorption edges for energy calibration in real time. This will permit an efficient and automated map of chemical composition for every pixel in the acquired sample image.