Characterization of dissolving pulp by multivariate data analysis of FT-IR and NMR spectra

Several grades of dissolving pulps have been analyzed using FT-IR, solid state C-13 NMR and two dimensional H-1-C-13 HSQC NMR spectroscopy to obtain an extensive data set for further characterization. The selection of the dissolving pulps with high cellulose purity was based on pulping process, wood type and, intrinsic pulp viscosity. Multivariate data analysis was used to investigate how information derived from the spectroscopic data correlate to each of the selection criterion: wood type, process type and viscosity. The spectroscopic methods were also compared with common dissolving pulp analyses to see to what extent spectroscopy can predict pulp analyses. Correlations were found between the spectroscopic data and the pulp characteristics process type and wood type, but not for intrinsic viscosity. A reason for a good correlation to wood type appears to be the hemicelluloses composition, expressed as the xylose:mannose ratio by 2D NMR spectroscopy. For process type, 2D NMR showed the most characteristic property to be the amount of reducing ends in the cellulosic samples, which in turn strongly correlates to lower molecular weight for the sulfite samples as determined by molecular weight distribution. Many common, yet expensive and time consuming, pulp analyses could also be predicted by the achieved models. It can be concluded that investigations of dissolving pulp characteristics, especially concerning different wood and process types, can take advantage of the methods and models presented in this study.