h1

%0 Thesis
%A Kaven, Luise Friederike
%T In-silico and in-situ optimization for enhanced synthesis of functional microgels
%V 31
%I Rheinisch-Westfälische Technische Hochschule Aachen
%V Dissertation
%C Aachen
%M RWTH-2024-06268
%B Aachener Verfahrenstechnik series - AVT.SVT - Process Systems Engineering
%P 1 Online-Ressource : Illustrationen
%D 2024
%Z Veröffentlicht auf dem Publikationsserver der RWTH Aachen University
%Z Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2024
%X Microgels are functional polymers with the potential for versatile applications. Each application requires tailored properties of the functional microgels. To enable a tailor-made production, insights during the reaction via process monitoring are essential, as well as mathematical modeling for predicting and optimizing microgel properties. Thus, this thesis provides advancements in microgel synthesis regarding monitoring, modeling, and optimization to unlock the full potential of these versatile polymers. First, concentration monitoring via Raman spectroscopy is established for the continuous microgel production mode. Transferring this monitoring technique from batch to continuous flow setups poses challenges that are systematically addressed. A quality criterion for Raman spectra is derived to allow functional outlier detection during continuous synthesis. Hence, overall, a guideline for in-line monitoring transfer is established. Second, the capabilities of Raman spectroscopy are enhanced by presenting a method to determine the microgel size from Raman measurements. Recent developments in machine learning are leveraged to improve size determination quality. The resulting accuracy is comparable with state-of-the-art off-line analysis tools. Thus, combining Raman spectroscopy and advanced machine learning methods is a promising approach for in-line polymer size determination. Third, Raman spectroscopy is also enabled for monitoring the charged microgel synthesis. Applying indirect spectral hard modeling resolves the complexities caused by the multi-component solutions with dissociated and undissociated states. Therefore, a detailed insight into the reaction phenomena during the charged microgel synthesis is enabled. Fourth, a mechanistic, dynamic model for the synthesis of microgels is extended to account for integration of functional epoxy groups. Unknown parameter values are calculated within a parameter estimation. By strategically including quantum mechanically calculated parameter values, the fit between model prediction and experimental measurements can be improved while reducing the calculational effort. By applying the identified synthesis model, the distribution of functional epoxy groups within the microgel is predicted. Fifth, a mechanistic, dynamic model is enhanced to capture the synthesis of charged microgels with regard to pH changes during the process. Again, missing parameter values are calculated within a parameter estimation, including quantum mechanically computed values strategically. The developed model presents a robust framework for predicting and optimizing the performance of charged microgels in diverse scenarios, paving the way for designing more efficient and tailored microgel-based systems. Sixth, a data-driven hardware-in-the-loop approach is presented to synthesize microgels of a desired size successfully. Data-driven approaches, particularly Bayesian optimization, are employed for microgel synthesis optimization for multiple objectives regarding product and process properties simultaneously. The proposed framework includes global deterministic optimization and has the potential for efficient microgel development by minimizing the number of experiments and modeling efforts needed. The thesis brings together advancements in the fields of process analytical technology, mathematical modeling, and data-driven optimization while combining experimental real-world development (in-situ) with theoretical considerations (in-silico). Therefore, this thesis's findings provide one step toward synthesizing tailored microgels with specific compositions or functionalities at increased production scale. Finally, many findings of this thesis are relevant not only for other polymer systems but also for method development in the field of spectroscopy-based size determination and hardware-in-the-loop optimization for all kinds of (chemical) systems.
%F PUB:(DE-HGF)11 ; PUB:(DE-HGF)3
%9 Dissertation / PhD ThesisBook
%R 10.18154/RWTH-2024-06268
%U https://publications.rwth-aachen.de/record/988558