ABSTRACT
ABSTRACT: Two techniques are described for the reduction and characterisation of electron energy-loss spectrum-image datasets. Firstly, an automated edge identification algorithm, applied previously to single spectra only, is shown to be an efficient tool for the detection and identification of core-loss edges within both energy-filtered and scanning EELS spectrum-images. Secondly, multivariate histogram analysis is demonstrated as a convenient method for establishing intensity correlations between multiple chemical distribution maps, enabling the identification and extraction of chemical phases present. This approach is also shown to facilitate rapid and accurate quantification of chemical phases within spectrum-image datasets via a phase-specific spectroscopy approach.
