Near Infrared Spectroscopy as a Screening Tool for Sugar Release and Chemical Composition of Wheat Straw

Abstract

Bioethanol production would benefit from a rapid screening method to determine the ability of feedstock to be processed into fermentable sugars. The aim of this study was to relate near infrared (NIR) spectra of straw to the release of sugars for ethanol production from cultivars of winter wheat, and to establish a calibration model to quickly determine the content of structural carbohydrates, lignin, and ash. We applied a high-throughput pretreatment and enzymatic hydrolysis (HTPPH) assay, involving hydrothermal pretreatment (180 � C for 17.6 min) and enzymatic hydrolysis, to establish the release of glucose and xylose from 20 cultivars grown in two replicates at two sites; in total 79 samples were measured. The NIR spectra could explain 56% of the variance in sugar release with a root mean square error of cross-validation (RMSECV) of 0.014 g g −1 dm. NIR calibrations predicting content of structural carbohydrates and lignin could explain only about 25% of total variance, whereas calibrations predicting ash content could explain 94% of total variance. The relatively low percentage of explained variance of sugar release was due mainly to uniformity of samples, which rendered the uncertainty of HTPPH method to be large compared with variance between samples. NIR spectroscopy, therefore, has potential to assess sugar release of wheat straw. Improved prediction of carbohydrates and lignin require better compositional analysis for homogeneous material. Despite successful prediction of ash content, site-specific cross-validation indicated that there might be problems with model transferability from site to site.