Abstract
Stationary phases for solid-phase extraction (SPE) were analysed by ESCA and SIMS in combination with multivariate analysis (MVA). The SPEs were endcapped and non-endcapped octadodecyl-bonded silica spheres of various carbon loadings (CL). The SIMS survey spectra and ESCA survey, detail and valence band spectra were recorded. In order to explore the potential of using MVA in combination with surface analysis, and to improve and simplify extraction of information from the acquired spectra, principal components analysis (PCA) and partial least squares (PLS) were applied. Various methods for normalization of spectra and variables were tested to improve separation between groups in the PCA score plots. With appropriate normalization, the different types of SPEs formed distinguishable clusters in the score plots from PCA on all types of spectra recorded. The clusters were ordered in correlation to CL. From the PCA performed on the SIMS spectra it was possible to extract information regarding both CL and endcapping status of the samples. The endcapping status could be predicted successfully using the PLS model built from the SIMS spectra. The weights of the PLS model could be used to identify mass fragments typical for the sample surfaces, and the fragments identified correlated well to the different surface structures of the samples. © 1997 by John Wiley & Sons, Ltd.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have