RAPID PREDICTION OF STICKIES CONTENT IN DEINKED PULP BY NIR SPECTROSCOPY

A rapid method to predict the stickies content of deinked pulp by near infrared reflectance (NIR) spectroscopy was developed. The content of stickies extracted by tetrahydrofuran from deinked pulp was quantified by Gel Permeation Chromatography with Evaporative Light Scattering Detector (GPC-ELSED). After vector normalization pretreatment of the original spectra in the wave-number range of 7502-6800cm(-1) and 5450-4246cm(-1), the correlation between stickies content and NIR spectrum of deinked pulp was established by partial least squares. The calibration model and prediction model of the stickies content were characterized by the correlation coefficient square (R-2) of 0.90 and 0.93 and standard error of 0.10 and 0.11, respectively. The relative standard deviation of the repeatability of prediction model of the stickies content in deinked pulp was less than 8% when the stickies amount was higher than 0.2 mg/g dry pulp. Therefore, It is feasible that NIR spectroscopy could be used for stickies rapid determination in industry.