Chemical compositions of hardwood and softwood pulps employing photoacoustic Fourier transform infrared spectroscopy in combination with partial least-squares analysis

In the present study, hardwood and softwood pulps were characterized by employing Fourier transform infrared photoacoustic spectroscopy (Fr-IR-PAS). The pulp samples examined originated from Swedish sulfite and kraft pulp mills, which utilize different cooking processes and modern bleaching technologies. Partial least-squares (PLS) analysis was used to correlate the spectral data obtained with the kappa (K) numbers and carbohydrate compositions of the pulp samples determined by enzymatic hydrolysis and subsequent capillary zone electrophoresis. Using four principal components, the present PLS model based on photoacoustic FT-IR spectra could explain 85% of the variance in the X matrix and 81% of the variance in the Y matrix. The Fr-IR-PAS technique in combination with PLS was found to accurately predict the contents of carbohydrates, i.e., xylose, glucose, mannose, arabinose, galactose, and hexenuronic acid residues, as well as the content of lignin measured in terms of K numbers and corrected K numbers of the pulps. From these predictions, the contents of xylan, glucomannan, and cellulose can also be predicted. The content of 4-O-methylglucuronic acid residues is, however, more difficult to predict accurately, using this approach.