Advancing dynamic frequency domain fluorescence measurements for reaction progress monitoring of photoreactions in complex media
Date
2023
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
This dissertation describes the development of a comprehensive analytical method- ology for investigating the kinetics and mechanistic aspects of photochemical processes in complex media using the photodegradation of the photosensitizer, rose bengal (RB), as a model. By using multi-wavelength frequency-domain dynamic fluorescence measurements and advanced data mining techniques for reaction monitoring, the proposed approach provides in-depth insights into complex photochemical reactions, significantly enhancing the interpretation of the acquired data. ☐ The core methodology involves the use of a multichannel frequency-domain spectrofluorometer for data acquisition and the implementation of various data mining strategies, such as singular value decomposition, generalized rank annihilation, and non-negative matrix factorization. Applying this methodology to investigate the photodegradation of RB in air-saturated alcohols under simulated solar radiation, the study successfully identifies the influence of solvent properties on the photochemical activities and photodegradation rates of RB. Data mining proves effective in mitigating distortions caused by spectral overlap in sample mixtures, and the approach’s performance is further validated by comparing results to previously reported data. ☐ Additionally, the research explores the photodegradation of RB in neat octanol, buffer-saturated octanol, and phosphate buffer under varying oxygen concentrations. The study showcases the versatility of the methodology in extracting spectra, decay times, and photodegradation or photoproduction rates of each component in the sample solution. The findings reveal that RB photodegradation pathways are influenced by sol- vent polarity and viscosity, with photooxidation competing against diffusion-controlled photolysis in the presence of reactive oxygen species. ☐ Lastly, the feasibility of incorporating compressive sensing techniques in the de- sign of multichannel frequency-domain spectrofluorometers is thoroughly evaluated. The study assesses the potential for achieving rapid response times, compactness, ruggedness, and cost-effectiveness in various applications, including real-time monitoring and analysis in diverse scientific and industrial fields. A systematic exploration is conducted to identify the effective algorithms, optimal coded signal lengths, and signal-to-noise (S/N) ratios, which are essential to establishing the viability of the proposed compressive sensing approach for spectrofluorometers. ☐ In conclusion, this dissertation establishes a robust and versatile analytical methodology for studying photochemical processes by integrating advanced spectroscopic and data analysis tools. The findings contribute to a deeper understanding of the underlying mechanisms governing these processes and highlight the potential for further optimization and wide-ranging applications of reaction monitoring techniques. This work underscores the value of combining sophisticated instrumentation and data mining methods for studying photo-induced processes and reactions in a variety of contexts.
Description
Keywords
Compressive sensing, Matrix-formatted frequency-domain Fluorescence, Multivariate analysis, Photochemistry, Photo-induced processes