Comparison of near infrared and mid infrared spectroscopy for the prediction of Pinus radiata bark chemical properties relevant to a biorefinery

Abstract

Forestry bark residues are an underutilised resource that could be exploited as a biorefinery feedstock to produce sustainable platform chemicals. However, variability in bark quality will significantly impact biorefinery processes, yields, and product quality. Therefore, simple, quick, and cost-effective analytical approaches for characterising incoming bark are required. A range of Pinus radiata bark samples were analysed to determine hot water extract yield, gross calorific value, ash content, and properties of the hot water extract (total phenolic content, monomer concentration, and molecular weight). Near infrared and mid infrared spectra of the ground barks were used to develop calibration models using partial least squares regression. Calibration models for gross calorific value and total phenolics demonstrated good predictive performance for both techniques based on their coefficient of determination (R2 > 0.82) and root mean square error of cross-validation (RMSECV). Hot water extract yield and molecular weight were not well predicted by either technique (R2 < 0.69). Although both techniques performed similarly, near infrared spectroscopy may be more practical in an industrial setting. Near infrared spectroscopy combined with partial least squares regression may be a useful approach for incoming bark quality management in a bark biorefinery operation.