CHEMICAL INFORMATION OBTAINED FROM AUGER DEPTH PROFILES BY MEANS OF ADVANCED FACTOR-ANALYSIS (MLCFA)

The advanced multivariate statistical technique ''maximum likelihood common factor analysis (MLCFA)'' is shown to be superior to ''principal component analysis (PCA)'' for decomposing overlapping peaks into their individual component spectra of which neither the number of components nor the peak shape of the component spectra is known. An examination of the maximum resolving power of both techniques, MLCFA and PCA, by means of artificially created series of multicomponent spectra confirms this finding unambiguously. Substantial progress in the use of AES as a chemical-analysis technique is accomplished through the implementation of MLCFA. Chemical information from Auger depth profiles is extracted by investigating the variation of the line shape of the Auger signal as a function of the changing chemical state of the element. In particular, MLCFA combined with Auger depth profiling has been applied to problems related to steelcord-rubber tyre adhesion. MLCFA allows one to elucidate the precise nature of the interfacial layer of reaction products between natural rubber vulcanized on a thin brass layer. This study reveals many interesting chemical aspects of the oxi-sulfidation of brass undetectable with classical AES.