Abstract
Alkaline pretreatment of wheat straw could significantly augment enzymatic hydrolysis for producing fermentable sugars, which is a pivotal process for the conversion of lignocellulosic biomass into advanced biofuels, biomaterials, or biochemicals. Yet, the enzymatic conversion process system is complex and multivariate, and study on the interaction mechanism of the key parameters in enzymatic hydrolysis is still lacking. Therefore, in this work, multivariate data analysis (MDA) (i.e., principal component analysis (PCA) and partial least square (PLS)) was conducted to reveal the inherent relationship and the significance of these factors in a modified alkali pretreatment system. A robust model, developed from 140 enzymatic hydrolysis datasets, was validated with an additional 20 datasets, demonstrating the predictive prowess of the PLS model. MDA identified that cellulase dosage, mechanical refining, dye adsorption value, and solid content were paramount variables. The integration of cellulase and xylanase notably elevated sugar yields and the conversion rates of carbohydrates, surpassing those of single enzyme treatments. The model's predictive accuracy, reflected in the close alignment between observed and predicted data, underscores its suitability for optimizing and controlling the enzymatic hydrolysis process. This study paves a way for data-driven strategies to enhance industrial bioprocessing of lignocellulosic feedstocks.