Date of Thesis
Producing biofuels from lignocellulosic biomass (i.e., non-edible plants) can reduce greenhouse gas emissions. The main challenge in converting lignocellulose to fuel, however, lies in the bioconversion process of biomass to sugars. To overcome the recalcitrance of the plant cell walls, pretreatment (e.g., ball milling) can be coupled with enzymatic hydrolysis to produce sugar monomers (e.g., glucose and xylose). The research question in this thesis study was “Can Fourier Transform Infrared Spectroscopy Attenuated Total Reflectance (FTIR-ATR) applied to untreated and physically pretreated lignocellulosic biomasses coupled with chemometric analysis predict sugar yields from enzymatic hydrolysis?” PLS models were constructed by correlating the X matrix—i.e, the FTIR-ATR spectra of raw and physically pretreated samples—to the Y vector—measured values of 72-hour glucose and xylose yields. It was determined that the PLS models constructed from the fingerprint region of FTIR-ATR spectra (800-1800 cm-1) of the five raw biomasses which underwent various physical pretreatment levels (no treatment, 1 hour of ball milling, 2 hours of ball milling, and shatterbox for five minutes) were able to predict the glucose yields (g sugar per fraction of Total Solids). The initial glucose model resulted in a coefficient of determination for cross-validation (Q2) value of 0.8262 with four latent variables. Using regression coefficients and variables important of projection (VIP) scores, regions of the spectra were truncated. These truncated regions were associated with bonds, identified as contributing regions for predicting glucose yields, suggesting that PLS regression models were created based on real chemical information and not chance correlation.
physical pretreatment, FTIR-ATR, lignocellulosic biomass, enzymatic hydrolysis, PLS regression, chemometrics
Master of Science in Environmental Engineering
Alaparthi, Rajasri, "Using FTIR-ATR to Predict Saccharification from Enzymatic Hydrolysis of Physically Pretreated Lignocellulosic Biomass" (2020). Master’s Theses. 242.