Abstract
Principal component analysis (PCA) and multivariate curve resolution (MCR) were applied to TOF-SIMS data of a polyethylene glycol (PEG) and poly(methacrylic acid) (PMA) mixed polymer sample coexisting with Si wafer, contaminants and impurities. PCA and MCR have complementary features in regard to deciding the significant number of components and separating the pure component spectra. As a result, PCA provides useful information for elucidating the existence of characteristic components and estimating the number of components for MCR. On the other hand, MCR is useful for separating the pure component spectra and estimating the origin of each pure component. Combination of these multivariate analysis methods is expected to be applicable to TOF-SIMS data of a wide variety of samples including unknown constituents.
