ISBN 9783843954099

978-3-8439-5409-9, Reihe Technische Chemie

Mareike Schreiber-Choo
Perspectives for Autonomous Downstream Process Synthesis of Biochemical Products on a Robotic Platform

142 Seiten, Dissertation Technische Universität Dortmund (2023), Softcover, A5

Zusammenfassung / Abstract

Products newly developed, complex fermentation broths, incomplete data basis and time pressure are key drivers for the design of new process synthesis methods. For that purpose, robotic devices enhance process synthesis by minimizing time, effort and costs. The advancement of robotic platforms from experimental support to systems with abilities for autonomous decision-making is a promising perspective. In the field of biochemical process design, a high number of repetitive screening experiments and decisions affecting overall process costs are required especially during early stages of downstream process development.

To simplify extensive screening experiments, this thesis focuses on the development of screening strategies with autonomous decision-making by a custom-built robotic platform. Three different approaches are developed for solvent screening for liquid-liquid extraction, adsorbent selection for liquid-phase adsorption and desorption agent screening. The methods are based on heuristics, decisive physical properties, key performance indicators with economical focus and supported by automated experimentation.

For extraction, the screening method is verified by extraction of progesterone out of a fermentation broth. Thereby, the need for extension of the method is highlighted and established by considering ecological factors as well as decisive parameters. The adsorption/desorption screening procedure is demonstrated for the adsorption of caffeine out of an aqueous solution with several impurities. An adaptive screening methodology for adsorbent screening is developed and implemented on the robotic platform guided by key performance indicators. Furthermore, perspectives for autonomous desorption agent screening are addressed. Overall, a path towards holistic process synthesis conducted by robotic devices is outlined within this work.