Near-infrared spectroscopic analysis for rapid evaluation of major chemical components in sugarcane bagasse

Abstract

In this study, near-infrared (NIR) spectroscopy was employed to determine cellulose in terms of glucan, hemicellulose in terms of xylan, and lignin contents in sugarcane bagasse as biomass for energy purpose utilization. We investigated by using three sample groups which were consisting of (A) twenty simulated samples prepared by mixing cellulose, hemicellulose, and lignin standards in many ratios to expand the range of contents for these analyses, (B) forty-seven sugarcane bagasse samples of wild species of Saccharum spontaneum and Erianthus, and their hybrids obtained from Khon Kaen Field Crops Research Center, and (C) seventy sugarcane bagasse samples collected from various sugar factories in Thailand. All samples were measured in the NIR region of 1,100–2,500 nm using reflectance mode. Partial least square (PLS) regression models for the quantitative determination of glucan, xylan, and lignin contents in sugarcane bagasse samples were calculated from data of NIR spectra and of analyzed contents detected by reference methods. The best PLS calibration models for cellulose (correlation coefficient (R) = 0.94, standard error of prediction (SEP) = 4.31%), xylan (R = 0.88, SEP = 1.50%), and lignin (R = 0.94, SEP = 2.08%) in sugarcane bagasse samples obtained from the model using actual bagasse samples, in which they developed from multiplicative scattering correction, second derivative, and second derivative pretreated NIR spectra, respectively. This study shows that the matrix of actual sugarcane bagasse in the NIR calibration model was necessary for getting the accurate results than those obtained by using a more comprehensive concentration range of analyzes obtained from the NIR calibration model, including the simulated samples. Essentially, NIR spectroscopy can be used to predict the necessary chemical constituents of sugarcane bagasse prior to converting biomass to substitute energy with fast detecting and reducing the use of chemicals.