The development of smart reactors for flow chemistry: the role of additive manufacturing and online analysis for automated optimisation
thesisposted on 02.07.2018 by Matthew J. Harding
In order to distinguish essays and pre-prints from academic theses, we have a separate category. These are often much longer text based documents than a paper.
This thesis investigates the application of online monitoring for the optimisation of flow chemistry, as well as how additive manufacturing can aid the integration of analysis and confer new functionality to flow reactors. The additive manufacturing (AM) processes used were stereolithography (SL) and the metal printing techniques selective laser melting (SLM) and ultrasonic consolidation (UC). Chapter 1 contains a short literature review, intended to give a clear background to the work contained herein. The literature reported gives a brief introduction to flow chemistry and some of the instrumentation used to perform it. Additionally, the evolution of reactor design is investigated leading to an overview of the use of AM for custom reactors. The subsequent use of online analytical technologies and how they relate to the enhancement of flow chemistry is discussed, as well as some of the protocols that have been employed to date to facilitate automated reaction optimisation. Chapter 2 investigates the design and manufacture of flow cells capable of online spectroscopy, as well as the integration of spectroscopic monitoring capability directly into reactors. In addition, the use of AM to produce accessories, not necessarily part of the wetted flow path, was investigated and showed that many useful parts such as fibre optic holders and screws could be produced. The capability of these flow cells was assessed through standard material analysis as well as through the online analysis of flow chemistry. In particular, the use of SL has enabled the production of flow cells with features smaller than 100 microns. This allowed in situ spectroscopy to be performed by embedding fibre optics directly adjacent to the flow channel, offering a new way for reaction monitoring by ultraviolet (UV) spectroscopy to be performed cheaply, and with full user control over the flow cell specification. No additional quartz features were required for these cheap and highly customisable parts. Flow cells of larger path lengths were also produced, and their performance tested, identifying designs and materials suitable for the inline analysis of flow chemistry. These designs were then successfully incorporated directly within the flow channels of larger scale reactors, tailored specifically to commercial flow equipment, for true inline analysis of flow chemistry. Chapter 3 examines the use of metal reactors formed through more expensive printing processes, SLM and UC. As the parts these techniques produce are fully dense, chemically resistant and thermally stable, they were used to perform high temperature chemistry, taking solvents substantially above their boiling points to accelerate reactions and perform them in a fraction of the time of the batch process. UC was also used to produce a reactor with a copper flow path and the possibility of reaction catalysis performed with active metal sections was investigated, revealing that chemical modification of the reactor surface greatly improved the reaction yield. UC was also utilised to produce a flow reactor incorporating a thermocouple in the main body, close to the flow channel to enable accurate reaction temperatures to be measured, a significant improvement over the temperature control offered through the flow instrument. This represents the first use of UC for the production of complicated geometry flow reactors and this work has shown that many more applications of the technique for flow chemistry should be investigated. The ability to perform light mediated coupling reactions in AM produced reactors was also demonstrated successfully for the first time, and further to this that the extended UV curing of SL reactors is crucial for improved robustness of these parts. Chapter 4 centres on the use of online analytical methods to provide rapid, selective, and quantitative online analysis of flow chemistry. This chapter also outlines some of the steps required for automation to be possible, including equipment specifications and the coding approach undertaken to integrate multiple different instruments. A combination of online nuclear magnetic resonance (NMR) spectroscopy analysis and automated experiment selection was then used to optimise a pharmaceutically relevant, photoredox catalysed, C-N coupling reaction between amines and aryl halides, performed under continuous flow conditions for the first time. This optimisation required minimal user input, operating completely unattended, and revealed that lower concentrations of catalyst could be employed than previously identified, reducing the amount of toxic and expensive metal salts required, while achieving high conversion of the starting material. In summary, this thesis has demonstrated that AM, in particular SL, can be used for the production of new high resolution microfluidic flow cells, as well as larger scale flow cell designs which can be integrated into the body of large reactors, not easily performed with other manufacturing methods. SL has also been used to produce reactors capable of performing light catalysed reactions directly, with no further modifications. The use of metal printing AM techniques has allowed in situ catalysis and high temperature, high pressure reactions to be carried out with ease. Finally, the use of online NMR with computer control and experiment automation has allowed the rapid optimisation of a pharmaceutically important C-N coupling reaction.