Statistical correlation of spectroscopic analysis and enzymatic hydrolysis of poplar samples

Spectroscopic characterization of poplar wood samples with different crystallinity indices, lignin contents, and acetyl contents was performed to determine changes in the biomass spectra and the effects of these changes on the hydrolysis yield. The spectroscopic methods used were X-ray diffraction for determining cellulose crystallinity (CrI), diffuse reflectance infrared (DRIFT) for changes in C-C and C-O bonds, and fluorescence to determine lignin content. Raman spectroscopy was also used to determine its effectiveness in the determination of crystallinity and C-C and C-O bond changes in the biomass as a complement to better-known methods. Changes in spectral characteristics and crystallinity were statistically correlated with enzymatic hydrolysis results to identify and better understand the fundamental features of biomass that influence enzymatic conversion to monomeric sugars. In addition, the different spectroscopic methods were evaluated separately to determine the minimum amount of spectroscopic data needed to obtain accurate predictions. The principal component regression (PCR) model with only the DRIFT data gives the best correlation and prediction for both initial rate of hydrolysis and also the 72-h hydrolysis yield. The factor that most affects both the initial rate and the 72-h conversion is the O-H bond content of the sample, which directly relates to the breakage of structural carbohydrates into smaller molecules.