Author ORCID Identifier

https://orcid.org/0009-0008-9935-9094

Defense Date

2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Chemical and Life Science Engineering

First Advisor

Dr. Thomas Roper

Second Advisor

Dr. Frank Gupton

Third Advisor

Dr. Lane Carasik

Fourth Advisor

Dr. Nima Yazdanpanah

Fifth Advisor

Dr. Charles McGill

Abstract

Continuous-flow synthesis offers major advantages in pharmaceutical manufacturing, including improved heat and mass transfer, enhanced safety, and the potential for real-time process control. However, impurity formation remains a central challenge, particularly in strongly acidic solvolysis systems where side reactions are highly sensitive to temperature, acid loading, and residence time. This thesis develops an end-to-end workflow for impurity understanding, prediction, and control in continuous-flow processing, using the di-alkylation of 4-hydroxybenzoic acid as a representative case. First, a systematic process optimization campaign mapped the influence of key operating variables on yield and impurity formation. Mechanistically informed kinetic modeling was then used to characterize the primary pathways leading to desired and undesired products, establishing how impurity formation evolves across process conditions.

Building on these mechanistic insights, machine-learning models were trained to predict critical impurity concentrations directly from process variables, enabling accurate, condition-dependent impurity forecasting without requiring full ODE model evaluation. These predictive models were subsequently integrated into a fully automated flow platform that executed real-time diversion of off-specification material based on live temperature, flow, and residence-time data. The resulting system-maintained product quality under both steady-state and perturbed conditions, demonstrating closed-loop, impurity-aware process control. Together, these developments illustrate a pathway toward predictive, self-correcting continuous manufacturing systems in which quality control is embedded directly into process operation, supporting the broader goals of Quality by Design and advancing the field toward autonomous pharmaceutical production.

Rights

© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission

12-12-2025

Available for download on Wednesday, December 11, 2030

Share

COinS