NMR processing techniques based on multivariate data analysis and orthogonal signal correction. 13C CP/MAS NMR spectroscopic characterization of softwood kraft pulp

Abstract

AbstractThis paper presents a novel way of extracting information from a series of severely overlapped NMR spectra using multivariate data analysis techniques. A number of softwood pulps were prepared from wood chips that were subjected to kraft cooking conditions in laboratory digesters. In addition to measurements of traditional physical parameters, the pulps were characterized using standard 13C CP/MAS NMR spectroscopy. The relationship between the kappa number and both the NMR time domain and frequency domain data was modeled using multivariate data analysis techniques. The variation in the NMR spectra that was not correlated with the kappa number was removed using a new preprocessing tool, orthogonal signal correction (OSC). The resulting OSC‐treated NMR spectra were used as descriptors to generate partial least‐squares projections to latent structures (PLS) models for the variation of the kappa number. PLS weights were used to generate NMR ‘sub‐spectra’ which correspond to changes in the pulps that occur as the pulping process proceeds from high to low values of the kappa number. The sub‐spectra were used to gain insight into the changes in the pulps occurring at the molecular level. Concomitant changes in cellulose crystallinity and the amounts of hemicellulose and lignin were observed in these sub‐spectra. Copyright © 2001 John Wiley & Sons, Ltd.