Counting Ions: Statistical Analysis of Time-of-Flight Mass Spectra for Solid Materials

We begin by identifying peaks within the spectrum of a sample of CuSi produced using Atom Probe Tomography. Our algorithm is designed to recognize features within the graphical spectrum in an automated fashion, over several orders of magnitude of peak height. Using the identified peaks, we consider ways to model material composition more accurately by dissociating overlapping peak effects using Bayesian hierarchical regression and MCMC simulation.