Deconvolution algorithms for electron spectroscopy

Abstract

In the article different techniques for processing of spectrometric data to be fixed under executing diagnostic methods of electron spectroscopy are discussed. Data processing includes a preliminary analysis of signal, a filtration and a smoothing, distinguishing useful signal and estimating its parameters, and finally, interpretation of the result. First stages are based on methods of digital signal processing, some of them are considered in the article. Interpretation of the results being measured by a detection system is assumed an analysis of an instrument function of electron spectrometer. It is founded that the instrumental function is one depending on a ratio of particle energy to device adjustment energy. The main object of data interpretation is to distinguish a source energy spectrum from a current registered by detector. The last problem is a strict integral equation of convolution type. Several deconvolution algorithms are applied for a testing problem, and their results are compared each other. It is shown that some of deconvolution algorithms are preferable as data processing gives opportunity to improve energy resolution of the energy spectrometer without changing construction of the system.