Kuijpers, K.P.L.

This thesis outlines how microreactor technology can be beneficial for photocatalytic systems. Owing to the mild reaction conditions, e.g. operation at room temperature and visible light irradiation, photocatalysis proves to be an attractive strategy to achieve selective transformations of organic compounds. In order to boost reaction kinetics, uniform irradiation of the reactor is desired. Continuous flow microreactors, with small characteristic dimensions, were employed to achieve ideal irradiation of the reaction mixture.

A part of this thesis focusses on bridging the gap from lab scale to large scale production for photoredox catalysis. Unfortunately, the small dimensions, necessary for efficient application of photocatalysis, directly imply challenges in the throughput of these type of reactors. This limits the use to lab scale production. Due to limitations in the absorption of light, implied by the Bouguer-Lambert-Beer law, standard scale-out strategies, i.e. increasing the characteristic dimension, are inefficient for photon-induced reactions. Alternative scale-up strategies, which keep the small dimensions intact, are therefore presented in this thesis.

The work described in this dissertation can be divided into two topics: scale-up strategies for photocatalytic reactions and automation of continuous flow microreactor systems. Chapters 2, 3 and 4 present the design and application of scaled-up micro and mesoreactor systems for photocatalytic gas-liquid reactions, while chapter 5 presents how automation of a microreactor based platform can assist analytical measurement techniques. In chapter 6 a solar based, scaled-up microreactor is combined with automation elements, which results in a stand-alone mini-plant.

Especially photocatalytic gas-liquid reactions are challenging to scale. Simply increasing the reactor length, while keeping the residence time constant, will result in significant pressure drop and changing flow patterns over the reactor length. Therefore, in chapter 2, a numbering-up strategy for the scale-up of gas-liquid photocatalytic reactions, that circumvents these challenges, is presented. A reactor system is designed, constructed and tested by applying a photocatalytic aerobic oxidation reaction that converts thiophenol to a disulfide. The modular design of a single photomicroreactor allowed us to systematically scale the photochemical reaction within 2^n parallel reactors (up to n=3 in this work). The flow distribution was first tested without applying a chemical reaction, using just an ethanol-oxygen mixture, resulting in a standard deviation of less than 5% among the 8 reactors. Secondly the oxygen consuming aerobic oxidation reaction was applied, which resulted in a very good flow distribution with a flow distribution less than 10%. The yield of the target disulfide in the numbered-up assembly was comparable to the performance of a single photomicroreactor.

In chapter 3 investigations on the robustness of the numbered-up reactor system are presented. In such a system potential disturbances can occur, such as channel blockage and light source failure. Channel blockage leads to relatively large changes in both flow distribution and yield. The standard deviation on flow distribution increased from 10% for no blockage to 15% for single blockage and even to 25% for the blockage of two channels. However, we found that the overall performance can be accurately predicted thus making it possible to adjust the reaction parameters to obtain certain output targets. Light source failure did not lead to large variations in the mass flow distribution, highlighting the importance of the flow distributor section. Since the reaction is photocatalyzed, the impact on the reaction yield was significant in the reactor where the light failure occurred.

Chapter 4 presents a novel high-throughput reactor designed for photochemical transformations. In this chapter, for the first time, a rotor-stator spinning disk reactor is employed a photochemical reactor. The reactor consists of an enclosed rotating disk with a narrow gap (2 mm) around the disk. The combination of the narrow gap size and the rotational energy of the disk enables effective mixing, which is necessary to enhance multiphase reaction systems. Also, by keeping the gap size small and embedding a quartz window as top cover we were able to efficiently irradiate this novel reactor. The performance of the reactor is studied with respect to many parameters; like rotation speed, liquid flowrate, gas flowrate, catalyst concentration, substrate concentration, gas holdup, gas bubble size, and energy dissipation rate. The conversion and selectivity for the synthesis of ascaridole, an anthelmintic drug, increase from 37% to 97% and 75% to 90% respectively with an increase of rotation speed from 100 to 2000 RPM. Compared to conventional photochemical reactors such as the batch reactor or the microreactor, the photo-rotor-stator spinning disk reactor has much higher productivity (270 mmol·h-1 or 19.2 mol·h-1·m-2) and higher selectivity (> 90%), with the latter illustrating the impact of mixing on selectivity. By improving mass transfer and with enhanced mixing in our reactor system, we were able to reach 1.1 kg·day-1 levels of productivity with full conversion of the substrate, without even optimizing the reactor design and light intensity.

Automation of processes can reduce time and expensive labor while increasing accuracy and reproducibility. In chapter 5 a fully automated continuous-flow platform for fluorescence quenching studies and Stern–Volmer analysis is reported. All the components of the platform were automated and controlled by a self-written Python script. With this system automated screening of novel quenchers or Stern–Volmer analysis can be performed, thus accelerating and facilitating both reaction discovery and mechanistic studies. This chapter presents the development, calibration and application of this system. Two case studies are presented: the photocatalytic decarboxylation of a,b-unsaturated carboxylic acids and the photocatalytic decarboxylative alkylation of N-containing heteroarenes with N-(Acyloxy)phthalimides with recently developed inexpensive, carbazolyl based, organic photocatalysts.

In chapter 6 the first steps in developing an off-grid, solar-driven mini-plant are taken. A scaled-up, red luminescent solar concentrator photomicroreactor (LSC-PM) is constructed and coupled to a control system which regulates the conversion of the oxidation of L-methionine to its sulfoxide. Methylene Blue is used as photocatalyst, which matches perfectly with the emission spectrum of the dye used to dope the LSC-PM. The control system makes sure that the conversion is kept above a minimum target conversion, with very little deviation due to the feedforward nature of the control system. A 99% average conversion was obtained under fluctuating irradiations in a lab environment. Furthermore, the reactor was coupled to a battery that acts as an energy buffer and power convertor, in order to enable all necessary electronic equipment. Also, the battery can be charged with a solar panel, making it possible to operate the reactor off-grid.

Outlook
In this thesis microreactor technology is used to scale-up photocatalytic reactions. However, such microreactors seem to be mostly suitable for lab scale quantities. When larger scale production levels are desired, the more compact photo-rotor-stator spinning disk reactor design seems to be the better option. In this work production scales larger than 1 kg·day-1 could be reached. As the system was still operated under light limited conditions, further scale-up is definitely possible. As in this work especially the part on the light source could be improved. The light source can be optimized on alignment, irradiation surface of the reactor and intensity as well. The combination of these variables will improve the throughput significantly, assuming mass transfer and reaction rate will not be limiting. Also the pRS-SDR itself can be increased in size, although this will be a minor improvement compared to the light source optimization. In this way, production levels for industries such as the pharmaceutical and fine chemical industry are within reach. I believe 100 kg/day throughputs can be achieved in the near future with a single reactor unit. Hopefully, this will open the path for implementation of photochemical processes in the industry.

Also, automation is applied to a microfluidic platform. In this way photocatalytic activity could be screened and reaction mechanisms could be studied efficiently. Even better systems could be developed when combined with inline analytics that can aid the search for novel synthesis, especially when combined with Design of Experiments (DoE) and Machine Learning (ML) algorithms. DoE and ML can also be used to optimize reaction conditions.

LSC-PMs are limited by the solar irradiation intensity. This energy density is not high enough for bulk production of chemicals. However, it can find its use in special applications, such as medicine production in remote places. In such places a reactor should be self-sustaining as an electricity grid is not available. By combining LSC-PMs, solar panels and a regulating system, energy neutral production is possible.