ANALYSIS OF MULTICOMPONENT FT-PGSE EXPERIMENTS BY MULTIVARIATE STATISTICAL-METHODS APPLIED TO THE COMPLETE BANDSHAPES

It is demonstrated that multivariate statistical techniques (computational procedures NIPALS and Procrustes rotation) successfully handle data evaluation in Fourier-transform pulsed-gradient spin-echo (FT-PGSE) NMR self-diffusion measurements on multicomponent samples in situations where there is extensive peak overlap and closely similar self-diffusion rates. Unlike normal evaluation procedures, based on peak heights - or possibly peak integrals - the techniques described here utilize the complete spectral bandshapes and are capable of extracting individual component contributions to the composite spectral envelope, i.e., a form of "size-resolved" NMR. The same approach should be applicable to other types of NMR relaxation spectroscopy, such as T1 and T2 measurements. © 1993 by Academic Press, Inc.