In recent years, the FDA has called on pharma and contract manufacturing organizations to switch from batch processes to continuous flow, which offers enhanced control, smaller reaction volumes, and time and cost savings. Continuous flow is particularly suitable for fast, potentially highly exothermic reactions, where safety concerns may prohibit the use of large-scale batch protocols. However, with no commercially available flow calorimeters, measurement of the thermal properties of these reactions remains a challenge. This article discusses how existing high sensitivity calorimetry systems can be adapted to obtain real-time calorimetry data under continuous flow conditions.

The principle of flow chemistry is most easily understood when compared with batch chemistry, the traditional approach for chemical reactions. Batch chemistry involves loading reagents into a single container—often a round bottom flask or jacketed reactor vessel—while heating and stirring to ensure the reaction proceeds to completion. Depending on the process, this is followed by multiple individual steps to separate and purify the final product. In contrast, in flow chemistry, reagents are continuously pumped through a temperature-controlled tube, pipe or microfluidic chip, and the generated product can be collected as the reaction proceeds. Compared to batch processes, this approach offers enhanced reaction control and reproducibility, due to the combination of optimized reactor geometry, superior temperature regulation and rapid, efficient mixing.

Exploring new frontiers
Transferring batch chemistry reactions to a flow regime has opened up new possibilities that simply could not be achieved with conventional techniques, allowing scientists to explore new chemical spaces and develop previously inaccessible reaction schemes, such as telescoping multi-step reactions. This streamlined workflow—almost impossible, or at least laborious, using batch chemistry—potentially opens the door for the discovery of exciting new molecules.

Flow chemistry can also easily accommodate other techniques, such as photo- or electro-chemistry, with one notable application using flow electrochemistry to directly synthesize metabolites. Once a medicinal chemist has achieved a ‘hit’ compound, the candidate can be dissolved and pumped through a flow electrochemical cell to yield a range of selectively oxidized metabolites. This simple operation eliminates the need to develop new synthetic pathways to produce the metabolites, potentially saving months of time.

Not every batch process can be translated to a flow regime, but the constant stream of high quality academic papers published over the last five to 10 years is testament to the rapid adoption of flow chemistry thanks to the unique benefits it offers. For instance, the small volume flow reactors are much easier to pressurize than larger batch set-ups, enabling elevated solvent boiling points and higher reaction temperatures, therefore increasing the rate of reaction. In addition, the fact that only a small volume of reagents is reacting at any one time makes the whole process a lot safer, particularly when working with hazardous or highly exothermic reactions. This process miniaturization also makes flow chemistry ideal for medicinal chemists, who often only require small volumes of product for characterization, helping to speed up discovery pipelines.

Bridging the gap between academia and industry
A deciding factor for many scientists investigating flow chemistry systems is the ability to rapidly optimize reaction conditions, reducing costs and saving valuable time. For example, the reaction time, temperature, concentration and reagent ratio can all be easily adjusted, allowing multiple reaction conditions to be investigated sequentially, with a simple solvent flush required between experiments. The addition of a liquid handling robot enables automated screening of reagents, further increasing the scope of reaction optimization.

Reaction optimization becomes even more critical as processes are scaled up for further testing, characterization and manufacturing of the compound of interest, but this is where technology has, to date, lagged behind. The main reason for this is that a majority of continuous flow processes have been developed in academic environments, with little need for the development of pilot- or production-scale processes. However, for these technologies to gain widespread acceptance in an industry setting, it is crucial that processes developed on the microliter scale can be safely and effectively translated to a manufacturing environment. Few laboratories have much experience in this area, and groups at the forefront of flow chemistry, such as the Kappe Laboratory (Institute of Chemistry, University of Graz, Austria), are looking at how to overcome the limitations of traditional batch reaction characterization methods.

Understanding the infinitely variable
A good example of this is reaction calorimetry. Calorimetry offers a unique view into the course of a chemical reaction. Instead of following the appearance of a new molecule or the disappearance of a starting material to measure the kinetics of the reaction, calorimetry studies the heat generated within the reaction environment. If the only ‘event’ within a reaction vessel is the actual chemical transformation, then the profile of heat generation shouldn’t differ from that of reagent consumption or product generation. However, the heat within a batch or flow reactor isn’t necessarily constant, and the chemical reaction is unlikely to be the only source of heat. For example, if you’re performing a reaction in a matrix of concentrated acid, and the process generates water, then there is a simultaneous physical interaction between the water and the acid. This will contribute additional heat to the reaction, but its contribution will decrease as more water is produced and the environment becomes more dilute, creating a moving target in terms of the heat.

There can also be other simultaneous physical effects that affect the overall reaction temperature. For example, dissolving a solid starting material is an endothermic process, whereas a solid product precipitating is likely to be an exothermic process. Heat within a chemical reactor can therefore come from many different places, not just the chemical reaction itself. Understanding the contribution of these various factors to the overall heat of a reaction is crucial for safe and effective scale-up, as this heat must be effectively managed and removed. Unfortunately, as reaction volume increases, the efficiency of strategies to remove the heat decreases, as the overall volume increases faster than the surface area.

To date, the small number of continuous flow processes that have been scaled up have relied on batch or semi-batch calorimetry data. While this provides a lot of useful information—the overall heat of a given reaction will not change significantly whether it’s a batch or a continuous process—the physical conditions will ultimately differ from flow reaction schemes. The other major difference is that, with a continuous process, it is ultimately the steady state reaction rate that is of interest. With a batch process, the finite availability of resources means the trajectory of reagent conversion is a ‘moving target’, and the reaction environment may change with time; it may start as a single-phase reaction, then switch to a two-phase reaction as it progresses, or vice versa. In contrast, a continuous process should be operating in a steady state at any point, but the exact reaction rate—and therefore the amount of heat generated—will be determined by a broad range of factors, including reagent concentrations, flow rates and reactor geometries. And, with plug flow reactors, the overall heat generation will change significantly over the length of the reactor vessel. This is the major challenge of reaction calorimetry in flow—capturing sufficient, accurate data to enable process optimization.

Understanding the technology
Traditionally, reaction calorimeters have been developed and optimized for the measurement of batch processes. The current industry ‘gold standard’ methods for batch reaction calorimetry rely on heat flow or power compensation technologies, both of which allow the user to measure heat within a reaction while the process temperature remains either under control or constant. The major drawback of this approach is that, although temperature changes are monitored in real time, the overall heat transfer for the reaction is determined through careful calibration of the system before and after the reaction takes place. This means that both methods are extremely sensitive to any change in product composition, liquid level, process temperature, agitation rate or viscosity, which will upset the calibration and distort the results.

An alternative approach is to measure the true heat flow of the reaction using pre-calibrated heat flux sensors that are located on the wall of the reactor vessels. This real-time approach directly measures heat across the reactor wall, without the need for pre- or post-experiment calibration, making it completely independent of the properties or behavior of the reaction mass. As a result of this, calorimetry systems employing this technology (Chemisens, Syrris) are extremely sensitive, allowing accurate measurement of changes in power of as little as 0.0001 watts—less than the energy generated by a dividing cell. This makes real-time calorimetry ideally suited to use with small volume reactors, and, by eliminating the need for calibration probes, this technology can also be easily adapted for monitoring continuous flow processes.

Putting theory to the test
Researchers from the Kappe Laboratory at the University of Graz recently demonstrated how a continuous flow set-up based on a commercially available real-time calorimeter (Chemisens CPA202) can be used to analyze heat generation for a number of common single- and multi-phase reactions.¹ To create a system capable of continuous flow calorimetry, the group placed a tubular reactor inside the thermostated reactor zone. Syringe dosing pumps were used to provide the liquids streams, and a mass flow controller with a fixed value back pressure for gas flow (Figure 1). Depending on the nature of the reaction—gas development during reaction or reaction temperature higher than the atmospheric boiling point of the solvent—back pressure regulators of up to 7 bar were applied. For reactions involving gases and one liquid feed, a simple T-mixer was used, whereas a glass static mixer was used for the experiments with two liquid feeds. The microfluidic chip included a micromixer with two inputs and one output, with a total residence volume of 250 μl. An additional PFA residence time coil was included downstream of the chip to increase residence time volume up to ~5.5 ml, ensuring that all chemical transformations reached complete conversion.

This set-up was used to analyze the calorimetry profiles of a number of common exothermic reactions of commercial interest, including simple mixing of ethylene glycol and water, hydrolysis of acetic anhydrides, nitration of phenols, reduction of aromatic nitro groups by hydrazine, oxidation of alcohols and olefin reductions using diimide. This demonstrated that the real-time calorimeter could be used to measure the total heat flow—and subsequently derive the heat of reaction—under continuous flow conditions in a safe manner for both single- and multi-phase reactions. Although it is difficult to compare data produced by these continuous flow reactions directly with data from related batch experiments—as optimization of continuous flow reactions requires elevated temperatures and/or use of a larger excess of reagent(s) to achieve the necessary short residence times—this study clearly demonstrates the potential for performing direct calorimetry measurements in continuous flow  processes. This offers a very attractive alternative to relying on batch or semi-batch calorimetry data for scale-up of continuous flow processes.

What’s next for flow chemistry?
Enhanced control combined with faster, safer reactions and rapid library synthesis make flow chemistry a versatile technique across the chemistry and drug discovery and development sectors. As more organizations recognize the benefits of this approach, there is now a growing demand for novel continuous flow technologies to meet the varied needs of academia and industry. Continuous innovation in this area will act as a catalyst for the future development of new applications. This not only offers a desirable alternative to batch processes, it also opens up new possibilities for multi-step reactions, allowing exploration of previously unattainable chemistry. 

References

1. Gabriel Glotz, Donald J. Knoechel, Philip Podmore, Heidrun Gruber-Woelfler and C. Oliver Kappe. Reaction Calorimetry in Microreactor Environments Measuring – Heat of Reaction by Isothermal Heat Flux Calorimetry. Organic Process Research & Development 2017, 21 (5), 763-770. DOI: 10.1021/acs.oprd.7b00092.