Monitoring the kappa number of bleached pulps based on FT-Raman spectroscopy

A new strategy was proposed to determine the Kappa number of bleached Eucalyptus globulus kraft pulps based on FT-Raman spectroscopy. Before modeling, smoothing (Savitzky-Golay), baseline correction (air-PLS), and Gaussian fitting were executed for obtaining data, including peak intensities and areas, from the Raman spectra. Two different Raman peaks, namely 1096 cm⁻¹ and 2895 cm⁻¹, were preset to normalize the lignin-related peak of 1600 cm⁻¹. Therefore, two data types and two internal standard peaks were utilized to calculate ratios of selected peak intensities and peak areas. Then, they were interpreted as independent variables in the establishment of predicting model via Regression analysis. Among four different models, comparison of evaluations suggested that the value of peak intensity was superior to the value of area as an inputting value and the peak located at 2895 cm⁻¹ performed marginally better than 1096 cm⁻¹ as an internal standard. Statistical evaluation of the optimal predictive model, which presented as R is 0.91, RMSE is 0.46 and MAE is 0.36, indicated that the model is reliable. Therefore, the model has the high potential of accurately monitoring the Kappa number of bleached pulp and then could be helpful to reducing the extra chemical and energy consumptions in the pulp and paper industry.