Author ORCID Identifier

https://orcid.org/0009-0005-6154-8349

Semester

Summer

Date of Graduation

2026

Document Type

Dissertation

Degree Type

PhD

College

Statler College of Engineering and Mineral Resources

Department

Chemical and Biomedical Engineering

Committee Chair

Debangsu Bhattacharyya

Committee Member

Srinivas Palanki

Committee Member

John Hu

Committee Member

Nagasree Garapati

Committee Member

Jingxin Wang

Abstract

For decades, many chemicals have been produced from fossil resources, contributing to resource depletion, pollution from extraction and refining, and hazardous byproducts. Growing environmental concerns and declining fossil reserves have intensified the search for sustainable sources of fuels, energy, and chemicals. Lignocellulosic biomass, the only abundant renewable carbon source, is therefore being explored as a feedstock for second-generation biofuels and high-value chemicals.

Currently, global wood-panel industry heavily depends on petroleum-derived commercial adhesives like phenol-formaldehyde, urea-formaldehyde, etc. These formaldehyde-based chemicals cause hazards to both human health and environment through emission of toxic volatile organic compounds. Typically, lignin is considered as one of the most widely used raw materials for bioadhesive production due to the presence of both alcohol- and phenol-type hydroxyl groups in its complex molecular structure. It has been observed that crosslinking soy protein with lignin can be an effective way to overcome the disadvantages associated with soy protein adhesive alone, such as low water-resistance, poor mechanical properties, etc. and improve adhesive performance. Therefore, this work focuses on the production of a formaldehyde-free, sustainable bioadhesive by crosslinking soy protein isolate with a partially degraded lignin/high-molecular-weight oligomer solution derived from base-catalyzed depolymerization (BCD) of kraft lignin under high pressure and moderate temperature conditions with a mild alkaline catalyst. A lumped kinetic model for the BCD of kraft lignin is developed, with kinetic parameters optimally estimated through dynamic optimization by using experimental data reported in the literature. Its low-temperature behavior is then checked against an in-house experimental observation. A representative stoichiometric model for the proposed reaction scheme is also developed. Several plant-wide process model configurations for bioadhesive production are proposed, including a laboratory-scale batch reactor and a continuous-scale reactor, with rigorous heat integration incorporated into the models. Additionally, an economic model is developed for techno-economic optimization. An equation-oriented approach is implemented for minimizing the levelized cost of adhesive (LCOA). It is found that the plant consisting of a plug flow reactor for the BCD reaction of kraft lignin along with a hybrid heat integration approach via vapor compression and supplementary low pressure steam results in the best economic performance. The minimal LCOA is found to be $634.21/mt, which is very competitive considering the current market price of $1000–4000/mt for bioadhesives.

The commercial deployment of biomass conversion technologies remains significantly challenging due to several interacting sources of uncertainty that can have strong impact on process performance, product quality, production rate, and economic viability. These uncertainties may arise from feedstock composition, limited experimental data, uncertain reaction pathways and kinetic parameters, scale-up assumptions, and market-dependent techno-economic factors such as raw material prices, utility costs, and equipment cost correlations, to name a few. As a result, simultaneous parameter estimation and uncertainty quantification (UQ) become essential to support uncertainty-aware process design and decision-making for different biomass valorization pathways. This work also presents the development of a two-stage Bayesian inference framework for simultaneous parameter estimation and posterior UQ, thus enabling scalable and uncertainty-aware decision-making in high-dimensional complex chemical process systems. The proposed approach combines Sobol global sensitivity analysis for screening and dimensionality reduction, followed by variational inference for UQ of kinetic, design/operational, and economic parameters. An integrated multi-software interface is developed for automated simulation and parameter estimation, enabling Bayesian calibration through stochastic gradient-based optimization and automatic-differentiation. The proposed framework, though demonstrated on the bioadhesive production process, is generic and readily generalizable to other biomass conversion pathways.

One of the principal steps in the potential lignocellulosic biomass conversion processes is the degradation of the major polymeric components to relatively smaller oligomeric or monomeric units. Platform C5 and C6 sugars derived from lignocellulosic biomass are key intermediates for a wide range of bio-based chemicals and fuels. These monomeric sugars are produced through the hydrolysis process where the glycosidic linkages between the monosaccharides are cleaved in presence of an acid catalyst. Acid-catalyzed hydrolysis can be performed using either concentrated or diluted acids as the catalyst, while concentrated acid hydrolysis (CAH) preceded by the biomass decrystallization step can result in a near-theoretical sugar yield. Despite several challenges of the CAH process, such as high utilization of concentrated acid and requirement of corrosion-resistant material of construction for reactors, the interest in CAH has been renewed recently due to improved acid recovery methods, lower operating temperature, lesser degradation of sugars and its adaptability to various biomass feedstocks. This work presents an integrated framework combining reaction kinetic modeling, parameter estimation, plant-wide process simulation, and technoeconomic analysis for producing C5 and C6 sugars from biomass via CAH. A species-resolved kinetic model for mixed-acid catalyzed hydrolysis is developed, with parameters estimated using dynamic optimization based on transient experimental data. A scalable commercial scale reactor configuration is also developed and integrated into a plant-wide process model that includes CAH of holocellulose, sugar separation, acid catalyst recovery through high-boiling solvent extraction, and acid reconcentration for recycle. A comprehensive techno-economic analysis is conducted to evaluate economic performance of alternative process configurations using the levelized cost of sugar production (LCSP) as the primary metric and identify economically favorable conditions for sugar production. Results show that acid recovery is the dominant factor influencing process economics, significantly reducing chemical consumption and production cost, while additional heat integration further decreases the overall production expenditures. The proposed integrated configuration provides a minimum LCSP of $197.3/mt, thereby establishing CAH with acid recovery and heat integration as a promising route for sustainable production of platform chemicals and biomass valorization.

Download

Share

COinS