Prediction of the Crystallinity of White Pine Using near Infrared Spectroscopy

Abstract

The crystallinity of wood is an important property of wood materials, it has an important effect on the physical, mechanical and chemical properties of cellulose fibers such as MOR, density, hardness increase, alpha-cellulose content, dimensional stability, moisture regain and dye sorption, chemical reactivity etc. The aims of this study were to investigate the ability of near infrared spectroscopy (NIR) to predict the crystallinity of white pine wood and the effect of spectra pretreatment on the prediction of crystallinity using NIR. Spectra were collected from wood powder a slowly rotating turntable and the crystallinity of wood was determined by X-ray diffractmeter (XRD) in this experiment. The results showed that NIR coupled with partial least square (PLS) method could be correlated with the crystallinity of white pine wood, and the ability of NIR prediction based on first derivative spectra was better than based on raw spectra or second derivative pretreated spectra. There was a significant correlation between NIR spectra and XRD determined crystallinity. The correlation coefficient for calibration (RC) was 0.932; the mean square error of calibration (RMSEC) was 0.022; the correlation coefficient for validation (RV) was 0.911; the mean square error of calibration (RMSEV) was 0.023. It was proved that NIR can rapidly and accurately predict white pine wood crystallinity.