This thesis focuses on developing quantitative spectroscopic methods to determine products from hydrolysis and fermentation reaction. A dispersive Raman spectrometer with 785 nm incident light is used to measure saccharides and ethanol in complex matrices. An appropriate pretreatment step is critical prior to hydrolysis reaction for reducing spectra interference. Different pretreatment methods are compared systematically and the glucose detection limits in hydrolysates are calculated.

Multipeak fitting model is also constructed in order to analyze two main sugars produced from cellulose and hemicellulose. Different ratios of sugar mixtures are used to examine the model's accuracy. Ammonia-pretreated corn stover is hydrolyzed by cellulase and hemicellulase, and concentrations of glucose and xylose in hydrolysate are determined by the multipeak fitting model.

The measurement is sensitive despite the complicated matrices from biomass conversion reactions. This protocol requires minimal sample preparation and analysis time with future automation capability. Raman spectroscopy is ideally suited for screening different biomass types, growing conditions or reaction conditions because it only takes seconds to minutes for each measurement and minimal spectral processing is required.

The developed multipeak fitting model enables accurate determination of glucose and xylose in hydrolysis broth. Similar methods could be applied to measure total sugar yields from other plant material with appropriate pretreatment. Raman spectroscopy has the potential to be used for real-time monitoring of reaction progress in hydrolysis broths that have complicated matrices.