Positive matrix factorization (PMF) is a least squares approach to factor a
nalysis which was originally developed for environmental data analysis and
has been applied to several problems in resolving sources of environmental
pollutants. PMF has been used as both a two-way and three-way data analysis
tool. In this investigation, three-way data arrays were used to explore th
e ability of PMF in curve resolution. Pulsed gradient spin echo (PGSE) nucl
ear magnetic resonance (NMR) data were measured for spectral mixtures where
the concentrations of the compounds decay exponentially. Three-way data ar
rays were constructed by packing different parts of the data from single ex
periments and were analyzed with three-way PMF to obtain the NMR spectra, d
ecay profiles and the self-diffusion coefficients of constituents. (C) 1998
John Whey & Sons, Ltd.