Characterisation of dissolving pulp using designed process variables, NIR and NMR spectroscopy, and multivariate data analysis

Two processability variables, filter clogging and alkali resistance, were measured in a series of laboratorycooked viscose pulps. The pulps were also characterised by ¹³CCP/MAS solid state nuclear magnetic resonance (NMR) and near infrared (NIR) spectroscopy. Partial least squares (PLS) regression was used to investigate the information provided by the spectroscopic methods with respect to the processability measurements. The study showed that the alkali resistance, R18, of the pulp and the filter clogging value, Kw, of the laboratoryprepared viscose can be modelled by NIR and multivariate data analysis (MVA). The alkali resistance, R18, of the pulp can also be calculated by NMR and partial least squares (PLS) regression. Analysis of the loading values in the PLS model showed that pulps with high alkali resistance have higher crystallinity than pulps with low alkali resistance. Analysis of variables of the cooking conditions showed that the chemical charge (Na₂O) and pH should be kept low to give high alkali resistance, whereas high Na₂O gives low filter clogging values.