Online Raman spectroscopy for bioprocess monitoring

Abstract

Online monitoring of bioprocesses is essential to expanding the potential of biotechnology. In this thesis, a system to estimate concentrations of chemical components of an Escherichia Coli fermentation growth medium via a remote fiber-optic Raman spectroscopy probe was studied in depth. The system was characterized to determine sources of instability and systematic error. A complete first-order error analysis was conducted to determine the theoretical sensitivity of the instrument. A suite of improvements and new features, including an online estimation of optical density and biomass, a method to correct for wavelength shifts, and a setup to increase repeatability and throughput for offline and calibration methods was developed accordingly. The theoretical and experimental ground work for developing a correction for spectrum distortions caused by elastic scattering, a fundamental problem for many spectroscopic applications, was laid out. In addition, offline Raman spectroscopy was used to estimate concentrations of fructose, glucose, sucrose, and nitrate in an oil palm (Elais guineensis) bioreaction. Finally, an expansion of optical techniques into new scale-up applications in plant cell bioprocesses, such as plant call differentiation was explored.