Chemical characterization from carbon Auger spectra by application of pattern recognition and factor analysis

Abstract

The feasibility of using pattern recognition and factor analysis to extract chemical state information from Auger line shapes was explored. These data analysis techniques were applied to x-ray-excited carbon Auger spectra. It was found that Auger line shapes could be described by using the amplitudes of selected frequencies from the Fourier-transformed Auger spectrum as variables during the computations. Factors analysis was found to predict the length of the alkyl chain in a series of poly(alkyl methacrylate) spectra and also indicated a chemical state difference in carbonaceous residues on sheet steels exhibiting good corrosion resistance and poor corrosion resistance.