Arnab Chaudhuri

8.1. SUMMARY OF FINDINGS
In this thesis, multiphase systems have been investigated in the rotor-stator spinning disk reactor (RS-SDR) to discern the reactor’s capability for intensification and for possible future industrial application. In this section, the results obtained in this thesis have been summarized.

In Part I, three different industrially relevant reactions have been investigated to illustrate the potential of the RS-SDR for industrial adoption. Chapter 2 presents the first example, where we have illustrated the use of the RS-SDR for the production of fatty amine N-oxides. These compounds are industrially important surfactants which are used in a wide range of consumer products. We began our study by first investigating the kinetics of both the formation and degradation reactions of the compound. In current industrial practice, the reaction is usually carried out in solvents such as n-Propanol, over the course of hours and in batch/semi-batch fashion. The use of such a solvent system unnecessarily complicates this reaction as it requires the use of downstream processing for the removal of the solvent. Keeping in line with green chemistry principles, we investigated the kinetics of the reaction in water (resulting in a liquid-liquid system). The reaction was found to be autocatalytic in nature due to the formation of the surfactant and we were able to model the kinetics based on models in literature. We also identified the possibility of accessing novel process windows with the use of higher reaction temperatures and reactant concentrations in a solvent-less, continuous system. With the use of the RS-SDR, both the mixing and heat transfer rates were intensified, allowing us to access those novel processing windows. At 80 °C, 57% conversion could be achieved in four minutes at a rotation speed of 2000 RPM. The reactor could be operated safely with high initial reactant concentrations and high reactor temperatures. The resulting specific productivity was calculated as 1960 kg h−1 m−3. In comparison, typical batch/semi-batch reactors in the industry operate in the time scale of hours. The RS-SDR is also able to handle the gel-like product obtained at higher conversions. We however, did find that the reactor would be best suited to be used only during its initial mass transfer limited stage. Upon reaching approximately 50% conversion, the use of a PFR becomes more attractive for the process. Nevertheless, the RS-SDR remains an attractive reactor for this process since it can easily be scaled up for higher productivity by stacking reactor stages. While, here we have demonstrated the use of three stages, further scaling can be carried out with correlations for energy dissipation rates which are presented in literature. Further optimization of this system, for instance by: studying the viscous behavior of the amine oxides, using peroxide stabilizers to prevent degradation, studying the possibility of product recycle, and further intensifiying temperatures, can make the RS-SDR an even more attractive reactor for this continuous and solvent-less process. The results obtained here can also be applied to similar autocatalytic reactions and further illustrates the potential of the RS-SDR as a tool for process intensification.

In Chapter 3, we looked into a gas-liquid reaction, a common multiphase system in the industry. We were also particularly interested in using an alternative means of activating molecules other than temperature. This led us to the idea of using photochemistry in tandem with the RS-SDR, as in recent years, various researchers have illustrated the importance of flow chemistry in this field. The main goal of this work was to develop a multipurpose and small footprint reactor which enables the scaling of photochemical transformations. Scaling photochemical reactions is often complicated with heat and mass transfer issues. Further complications include the attenuation of light (Bouguer-Lambert-Beer Law). However, due to the excellent mass transfer characteristics of the RS-SDR, we showed that a photo rotor–stator spinning disk reactor (pRS-SDR) enables the efficient scaling of challenging gas–liquid photochemical transformations (up to 1.1 kg day−1 productivity). The challenges associated with microreactors, e.g. high pressure drop and variable flow patterns, often prevent the technology from being applied at industrial/pilot scale. In contrast, the pRS-SDR operates without any significant pressure drop and the formation of dispersed bubbles depend mainly on the rotation speed of the disk, rather than the imposed flow rate. This allowed us to obtain exceptional high yields and selectivity for the photochemical [4 + 2] cycloaddition between α-terpinene and singlet oxygen, yielding ascaridole within only 27 s residence time. We were able to also convinc-ingly show that the increased mass transfer rates observed in the pRS-SDR are crucial to obtain these excellent results.

One aspect which will determine if the RS-SDR is eventually adopted for use in the industry is its ability to be scaled-up. In the previous two studies, we have mentioned how the reactor can easily be scaled by stacking stages. In Chapter 4, we looked to properly characterize and illustrate the scalability of the RS-SDR by investigating the transesterification of triglyceride. The product, fatty acid methyl esters, commonly referred to as biofuel, has a wide range of applications. In this study, we have illustrated the performance of the RS-SDR for the homogeneous base catalyzed transesterifcation reaction in both a single stage RS-SDR and in a multistage RS-SDR. The high shear forces present in the RS-SDR intensified the mixing and for the homogeneous base catalyzed reaction, this resulted in high specific productivity (5.44 mol m−3 min−1 of FAME in a 1-stage RS-SDR and 6.97 mol m−3 min−1 of FAME in a 3-stage RS-SDR). We also observed a synergistic effect of scale up on the specific productivity due to the increasing PFR characteristics of the reactor (more tanks in series). We also extended the study to heterogeneous base catalyzed transesterification reaction for the production of FAMEs. We performed the reaction using sodium silicate as the slurry catalyst. We were able to reach 72 % conversion in the RS-SDR with a τ of 8 mins. using a high slurry concentration (20 wt.%). We did not observe any clogging of the reactor during operation. In comparison, in batch, only 47% conversion was reached in 8 mins. of residence time. The results of this study illustrate the potential of the RS-SDR for not only FAME production but also for other multiphase reactions of similar nature.

In the second part of the thesis, we looked to characterize the RS-SDR to be able to better design future reactors and processes. For instance in an effort to further optimize the productivity of the reactor, in Chapter 5 we have demonstrated how the structural modifications on the rotor and stator of the RS-SDR can have effects on the hydrodynamics of the reactor. The macromixing results illustrated that certain rotor modifications can be used to obtain more backmixing characteristics in the reactor. This was also linked to a higher average energy dissipation rate. Using micromixing studies we have illustrated that in some cases the increase in the energy dissipation rates translated to higher εloc. This indicates that perhaps further enhancements to mass transfer rates can be brought about using these types of modifications. On the other hand, certain reactor modifications, though illustrating worse micromixing characteristics, illustrate much better PFR type behavior. Essentially, the use of structural modifications presents as another degree of freedom with which processes involving the RS-SDR can be designed. This will have particular applicability for multistage RS-SDRs.

In Chapter 2, we observed very viscous flow behavior under intensified conditions. To better understand the flow behavior in the RS-SDR with viscous fluids, in Chapter 6 we investigate the macromixing performance. We used residence time distribution measurements to gauge the effectiveness of the reactor. We observed that while the viscous Newtonian fluid illustrated behaviour which was consistent with prior investigations, the non-Newtonian fluid demonstrated quite aberrant behavior. Even at high rotation speeds, the reactor model deviated drastically from previous investigations. We concluded that this was due to the presence of significant dead volume fractions in the throughflow and transition regimes. Further investigations should look into the operation limits (viscosity and flowrates) to discern when significant deviations to the reactor model start to appear.

In Chapter 7 we have illustrated the effect of various operating parameters on the gas hold up in the reactor, with particular emphasis on the role of viscous fluids. Using a gravimetric technique, we have illustrated that in the case of top to bottom flow, the hold up values can decrease up to 15% with rotation speed, depending on the operating conditions and the nature of the fluid. For bottom to top flow, a much more stable hold up is observed with rotation speed for water. However, viscous fluids illustrate the same decreasing behavior. The flow structures forming for bottom to top flow are irregular in comparison to the top to bottom flow structure. This should be further investigated to better understand the reason for the observed behavior. Furthermore, the pressure drop across the reactor and the energy dissipation rates were investigated with viscous gas-liquid flows. We found results which deviate from previously proposed correlations. This is most likely to due to operation in the laminar region rather than the turbulent region and this should be further investigated as well. Furthermore, to develop a better understanding of the bubble formation and deformation dynamics the pressure drop experiments should be carried out with a higher frequency of sampling.

8.2. OUTLOOK
The results from this thesis demonstrates the potential of the RS-SDR with specific examples. Similar investigations into multiphase reactions will draw more attention to the field. There should be particular applicability for fast reactions, exothermic reactions. This thesis, as well as countless studies in literature, have also demonstrated the importance of proper kinetics studies and scope for using green chemistry concepts. Future implementation of these concepts and coupling them with intensified reactor can vastly help improve the efficiency of chemical processes. In this section, we recommend some future research directions for the RS-SDR, propose how such reactors can be used in the future, and speculate how chemical plants of the future, using the RS-SDR (or other intensified reactors), may look like.

One particularly interesting field where future work should be focused on is photochemistry. As discussed in this thesis, in photochemical reactions the kinetics can be scaled with the intensity of light. As a consequence, with the emergence of more powerful lamps and LED equipment, the mass transfer rates for these reactions will come under more scrutiny. Furthermore, due to the Bouguer-Lambert-Beer limitations, consistent irradiation of the reactor becomes highly important. The use of continuous flow chemistry has already demonstrated tremendous potential in this field; however, one challenging aspect involved with the use of for instance microflow reactors is scaling. The RS-SDR seems particularly well positioned to be used as a reactor in this field due to the ease of scaling the device. The use of the rotation speed allows the reactor to be scaled independently of the flow. Investigations of the RS-SDR for other multiphase systems involving photochemistry such as solid-liquid-gas systems, will further demonstrate the potential of the reactor. Given the current applicability of photochemistry in the pharmaceutical industry, the demon-stration of the intensification of certain active pharmaceutical ingredients (APIs) can lead to the adoption of this technology in a rapidly growing industrial sector. Further research into the development of other types of multipurpose systems should also be investigated. For instance, the use of a membrane coupled with the RS-SDR can be particularly interesting for equilibrium type reactions [1–5].

In the field of speciality chemicals, reactors such as the RS-SDR can prove vital for transitioning from batch to flow. The high throughputs which can be obtained by this reactor and the scalability should allow for more sustainable and effective production. Initially, the RS-SDRs can serve as secondary reactors which can be used for meeting fluctuating demands. This will allow for the demonstration of the capabilities of the reactor in an industrial setting without high financial risks. RS-SDRs are also particularly applicable for the production of speciality chemicals due to the presence of stiff competition, but high margins, in this sector. Therefore, small improvements in costs and product quality can have large implications on the profit margins. Looking slightly further in the future, the modular aspect of the reactor will allow it to be used in small, localized plants. These small production units can help to reduce the high costs often associated with chemical production facilities. Logistical costs associated with bringing products to market and the transportation costs of bringing raw material to sites can also be reduced. This of course also reduces the carbon emissions associated with the logistical transport of goods. Instead these plants would move production closer to customers, while also being able to respond rapidly to changes in demand due to the modular nature of the reactors. Such plants, also improve on the safety due to the lower inventory and can perhaps even be operated using renewable energy supported with a battery powered grid. The transition to modular manufacturing can therefore transform the chemical plant from huge facilities to much smaller, localized production centers (Fig. 8.1) [6–10].

Figure 8.1: The future of chemical manufacturing?

Figure 8.2: The estimated breakeven time for a RS-SDR based biofuel process

There is a famous saying; "if it ain’t broken, don’t fix it", which perhaps, somewhat understandably, can be used to describe the working mentality of the chemical industry. Most processes are still carried out in batch since it is a proven method and poses relatively little risk. This makes it unlikely that, without outside incentives (e.g. government regulations), manufacturing facilities will look to completely overhaul current production methods to make way for intensified technologies. However, in one field where perhaps inroads can be made is when new industrial facilities are designed/constructed. For instance, it is expected that biofuel demand will rise to 4600 billion liters per year in 2045. To meet this demand, new production facilities have to be installed. It is therefore vital that in these new facilities, the use of intensified continuous reactor technologies is championed. Using the results obtained in this thesis for the biofuel productivity, an ASPEN simulation was used to predict the cost of an RS-SDR based biofuel facility. A facility such as this, with the production capability of 75,000 tonnes/year is expected to have an estimated breakeven point of 2 years. With further design improvements in the RS-SDR, such as the use of magnetically coupling of the shaft, thereby reducing mechanical wear, the lifetime of the reactor should increase even further. This makes the investment into such processes relatively low risk [11].

In conclusion, the RS-SDR is an example of an intensified reactor which has demonstrated tremendous potential for improving the efficiency and sustainability of mass and heat transfer limited reactions. The reactor seems particularly apt for industrial applications due to the ease of scaling. Further integration of various modes/methods of production (i.e. photochemical) can help to push the applicability of the reactor even further. In the future, chemical plants could utilize the high throughput capabilities of the RS-SDR to develop small, localized production centres, which could mitigate the vast amounts of energy expended on logistics and transport. Inherently, these smaller production centers would be safer to operate due to its smaller inventory and would also require less personnel, being a continuous process. Such a shift in the method for the production of chemicals maybe necessary if we are to meet the environmental and development challenges of the current century. However, whether or the not the reactor is eventually adopted into the industry may not depend on the technical capabilities of the reactor but rather on if proper incentives (e.g. regulations or industrial competition) are present to push the industry to change its approach to chemical production.