Microstructured Reactors for Rapid Process Development and Scale-Up

Published on: 
Pharmaceutical Technology, Pharmaceutical Technology-09-01-2010, Volume 2010 Supplement, Issue 4

The authors discuss a continuous-flow reactor that avoids parallel channels and enables economic plant setup. This article is part of a special issue on API Development, Formulation, Synthesis and Manufacturing.

Pharmaceutical and fine-chemical production is dominated by batch processes, which enable many kinds of reactions and can handle multiphase systems in multipurpose plants. Operators can arrange vessels in various configurations to establish several reaction routes and even certain work-up steps such as distillation or extraction. Production quantities of pharmaceutical intermediates or active pharmaceutical ingredients (APIs) vary from a few grams in the first medicinal studies to several hundred tons per year for successful pharmaceutical products. This wide production range demands various manufacturing devices at several scales for process transfer.

(LONZA)

The first laboratory studies work with just a few grams of reagents in glass flasks, but continuous-flow devices with integrated microchannels also produce these quantities (1). Process development leads from small-scale production, to sample production at pilot scale, to large-scale production (2). Often, during the scale-up of the chemical process, a wide range of stirred vessels is used to perform the reaction. Heat transfer and mixing rate, however, are often limited in stirred vessels, which need high dilution, long operation or dosing times, and sometimes do not permit highly exothermic reactions. One solution is to combine the versatility of batch vessels with the safety, reproducibility, and high transport capabilities of continuous-flow microstructured equipment.

Microstructured devices

Microstructured devices with small internal volumes and high surface-to-volume ratios offer transport capabilities for rapid mixing, enhanced heat transfer for good temperature control, and intensified mass transfer (3). The proper control of these often harsh conditions is not only essential for the safe operation of chemical equipment, but also necessary to enable an economical chemical process. Harsh conditions with an unusual range of temperature, pressure, concentration, mixing time, and residence time make new process windows and process routes available, and enable the manufacture of new chemical intermediates. These conditions and routes are not feasible in batch vessels or cannot be maintained during process scale-up (4–6). But harsh reaction conditions can be handled safely in closed systems with small internal volumes.

Microstructured devices operate under continuous flow conditions. Generally, continuous processes offer many advantages ranging from controlled process conditions to high flow rates and mass throughput. Continuous operation enables bulk-chemistry processes to have high production capacities. Fluid dynamics determine the characteristics of continuous-flow equipment such as pressure loss, residence time, heat-transfer characteristics, and mixing time. The combination of continuous-flow processes with microstructured devices brings benefits to the laboratory and production environments. This article discusses microreactors with a single rectangular channel, typical diameters from 0.2 to 2.0 mm and higher, moderate flow rates (e.g., 10–300 mL/min), and Reynolds numbers in the transitional regime (i.e., 100–3000) from straight laminar to turbulent flow. The authors describe process design and related issues with generic examples, as well as with real chemical examples from Lonza's (Basel) experience.

Equipment design for single-channel microreactors

Lonza designed a small, compact plate device called the LabPlate reactor to visualize the flow inside the microstructured channel. The device enables personnel to design reactor channels and develop processes with low flow rates when reagent availability is limited. Process conditions are similar to those in capillary chemistry, as well as in larger reactor devices, with the advantage that the reaction zone can be inspected and viewed. Multiphase reactions can be observed with phase distribution, as can the precipitation of metastable intermediates.

The fluid entering the microchannel within the reactor plate passes through the entrance, contacting element, and several mixing- and residence-channel elements, each with its own design (see Figure 1). The entire LabPlate reactor consists of a cooling block with cover plate, microstructured plate, view glass, and flange housing (see Figure 2). The fluids are sealed against the external environment by conventional O-rings; the reagents and heat-exchange medium are separated by metal walls. The modular setup allows the integration of several microstructured plates, as well as the integration of the reactor into other flow equipment. The entire device can be integrated into a modular microreactor system (MMRS, Ehrfeld Mikrotechnik BTS, Wendelsheim, Germany) (7).

Figure 1: (Left) A LabPlate reactor (Lonza, Basel) with microstructured channels and thermal fluid and reagent connections. (Top right) A typical mixing channel with nozzle-type contacting element and tangential mixing elements. (Bottom right) A single mixing channel during a gas–liquid test reaction in the LabPlate microreactor. (IMAGE IS COURTESY OF LONZA)

Lonza developed a plate stack reactor for reactions that typically take minutes to complete (see Figure 2) (8). The reactor is based on the multiscale approach, which adapts plates of various sizes to the reaction needs. For example, operators may use a tiny channel at the start of the reaction, when heat generation is strong, and gradually increase the size of the plates to accommodate slower reaction rates (i.e., less heat evolution). This design optimizes heat transfer, minimizes pressure drop, and greatly increases volume (i.e., by as much as several mL). In addition, the reactor may be combined with conventional heat exchangers and tube equipment to gain several liters of volume and several minutes of residence time. Lonza manufactures and distributes the reactor in cooperation with Ehrfeld Mikrotechnik BTS (7).

Figure 2: A microreactor (LabPlate, Lonza, Basel) for development purposes, including a microstructured plate with heat-exchange block and connections for the heat-exchange medium (left) and a typical plate-stack reactor setup with A6 or A5 size (right). (IMAGE IS COURTESY OF LONZA)

The microstructured reactor plates are made from corrosion-resistant material and can perform various tasks in a modular setup. Plates are designed for mixing and heat exchange. Mixing plates include a mixing channel as well as wider channel elements to provide sufficient reactor volume for heat exchange and appropriate residence time. In an effort to standardize Lonza's proprietary MicroReactor design, the company chose production-plate sizes based on the European A4, A5, and A6 standard paper sizes. Each size features a plate area double that of the previous size, thus doubling the heat-exchange area and reactor volume. The sizes enable scale-up, which is related to the reaction's kinetics (9). Thus, for rapid reactions with typical reaction times of less than 1 s (i.e., mixing controlled, Type A), the aims are the following:

  • To ensure sufficient cooling between the reactor plates and mixing channel
Advertisement

  • To provide short mixing times in tiny, complex channels with a comparable high-pressure drop.

For rapid reactions with typical reaction time between several seconds and 10 min. (i.e., kinetically controlled, Type B), the aims are the following:

  • To maintain volume for sufficient residence time for the reaction and the same area-to-volume ratio for enhanced heat transfer

  • To optimize mixing quality by choosing a pressure drop as low as possible.

Slow reactions with a characteristic time of more than approximately 10 min. (i.e., Type C) can be performed in conventional equipment such as static mixers or tube reactors. However, for rapid reactions, all the reactor plates are microstructured because mixing and heat exchange are the dominant factors. For intermediate reactions, the same plate depth is used for the various plate formats, thereby keeping the surface-to-volume ratio constant. The gradual size increase of reactor plates (i.e., the multiscale approach) and appropriate channel geometry allow the MicroReactor to operate at flow rates as high as 600 mL/min.

The modularity and versatility of the single channel and plate-in-series approach allow the development of plates appropriate to a multitude of applications. Lonza designed, manufactured, and applied plates for rapid mixing, gas–liquid dispersions, and multi-injection applications in various projects, thus demonstrating the technology's suitability for various purposes. The right side of Figure 2 shows the individual reactor plates made from Hastelloy C-22 sandwiched between aluminum plates with high thermal conductivity, a design that yields a compact reactor. In this way, the thermal fluid layer is not directly fixed onto the reactor plates, thus allowing for cost-effective manufacturing, as well as quick and easy adaptation to various reaction conditions. The overall reactor is robust, allows high flow rates of the heat exchange fluid, and can sustain pressures higher than 100 bar on the reaction side.

In many reactions, especially with viscous systems and low-temperature applications, the pressure drop may become important at high flow rates. In addition, the mixing zone often is the plate section that consumes the largest pressure drop. Consequently, an enlargement of mixer elements at higher flow rates drastically reduces the overall pressure drop. In general, no loss of performance is observed as long as the same energy-dissipation rate in the mixing zone is maintained. Thus, the mixing zone is the only scaled factor that is considered in this reactor technology, and it must be properly designed and adapted. It is essential to operate a microreactor with one single channel and completely avoid device parallelization to scale up processes successfully from laboratory to pilot scale.

The stepwise scale-up of the plates is accompanied by a stepwise scale-up of the channel cross section, leading to a scaling factor of approximately 1.4 between each two steps. For example, the typical hydraulic diameters of the mixing channel in the A6 reactor plate are 0.35 and 0.5 mm. The A5 reactor plate therefore contains mixing-channel structures with typical diameters of 0.7 and 1.0 mm. Larger plates are scaled up accordingly, thus creating comparable mixing conditions. To gain residence time, the heat-transfer area and internal volume are increased in corresponding steps of connected plates in series. These measures result in a consistent, versatile, and flexible scale-up strategy that completely avoids parallelization over a wide range of flow rates (10).

In addition to reactor size, the operational conditions are changed accordingly. Small reactors such as the LabPlate and A6 Lonza MicroReactor are well suited for projects in process development. They fulfill the needs of early-phase studies in a consistent and straightforward manner. Larger A5 MicroReactors are suitable for small-scale and pilot-production campaigns.

Project approach for continuous-flow processes

The main goal in laboratory development is to achieve a robust process because chemical systems are often metastable and tend to form deposits that are more or less stable over time. Precipitation or fouling within the channels creates unpredictable pressure drop behavior over the reactor. The reactor technology must facilitate timely and flexible process development and minimize the need to consider chemical engineering during process development. Thus, the use of one single channel, together with robust, pulsation-free high-pressure pumps, ensures correct feed balance and stoichiometry, as well as the ability to clean the reactor during and after operation.

The MicroReactor concept also indicates a certain project-management strategy for converting a batch protocol into a continuous-flow process. The base is a detailed mass balance with stoichiometry, including all starting materials, intermediates, and products, as well as major side products. A preliminary short feasibility study with a LabPlate or A6 MicroReactor indicates the potential advantages of performing a given chemistry in a continuous-flow process. Typical parameters such as concentration and mixing ratios, mixing and residence times, reactor temperature, and pressure can be tested. A decision at the end of each step allows good project control and shortens the time needed to make a decision.

If results are positive, operators can perform a parameter-optimization study to develop a robust process with sufficiently high yield and selectivity. A flexible and modular reactor setup with pumps, sensors, control system, and residence-time loops allows the investigation of a wide range of process parameters and will give a huge amount of data from kinetics and chemistry. These data are important for the design of the final production process. A long run of several hours in the laboratory at the end of the optimization step ensures the achievement of a robust and reliable process that can be scaled up to pilot scale.

The production campaign on pilot or large scale can be performed with the same reactor equipment used in the laboratory, but with longer operation times and appropriate peripheral equipment. The high flow rates typical of production can be tested in the laboratory for a short period to guarantee appropriate reactor performance. This approach enables operators to avoid a parallelization of reactor equipment, including in production campaigns. The approach also provides flexibility and manageable cost scenarios and time limits. In additionally, it avoids the risk of a technology jump from laboratory to production scales.

Scale-up example

All reactor types displayed in Figure 2 are used frequently in Lonza's laboratories and production environment. A real-case example of a two-step organometallic reaction (i.e., lithium–hydrogen exchange and coupling) demonstrates the scale-up of microreactors. The reaction had three feeds: Feed-1 with substrate, Feed-2 with the first reagent, and Feed-3 with the second reagent. The reaction was stoichiometric and operated at two temperature levels. Feed-1 and Feed-2 were precooled to the cryogenic reactor temperature (i.e., –35 °C), and the second reaction was performed without cooling at room temperature. The flow diagram and reaction scheme are depicted in Figure 3.

Figure 3: Reaction scheme of a two-step reaction, corresponding process scheme with two reactors, and image of the plant setup in a pilot-plant environment. Three feed lines and the MicroReactor (Lonza, Basel) with iced surface (–35 °C reactor temperature) are visible. (IMAGE IS COURTESY OF LONZA)

The first reaction was of Type A with an adiabatic temperature rise of more than 75 °C. This reaction was performed in four units: a static mixer, a glass microreactor, and the Lonza A6 and A5 MicroReactors. The second reaction was of Type B and less demanding in terms of heat exchange (ΔTad < 25 °C) and mixing. A microreactor and static mixer performed equally well.

Personnel observed no adequate temperature control within the static mixer for the first reaction. Roughly the entire adiabatic temperature rise was detected at a medium flow rate (i.e., 148 g/min). Many unwanted side products were formed, and yield dropped visibly compared with that at lower flow rates (i.e., 84% versus 89% on average).

The glass microreactor showed a loss of temperature control at high flow rates, but it was not reflected in the product yield. The short residence time also was important for this process.

Lonza's smaller (i.e., A6) and larger (i.e., A5) MicroReactor technology showed equivalent performance. It was possible to operate the A5 reactor at a speed of 237 g/min with a total flow rate higher than 700 g/min for the second reaction by keeping the pressure drop well controlled. A yield of 88% was achieved. A pilot campaign produced 700 kg of isolated material, thus yielding more than 10 m3 of processed solution through the reactor setup. A second campaign was performed to produce more than 2 tons of isolated material. Both campaigns demonstrated the long-term robustness of the process and the reliability of the installed reactor equipment. Further chemical examples from Lonza's experience demonstrate the versatility of the reactor toolbox described with various kinds of chemistry such as phosgene, ketene or diketene, ozone, or the production of nitrogen-containing compounds (5, 10).

Conclusion

After more than a decade of microstructured devices in chemical production, the implementation of microreactors is essentially understood. The interaction of chemistry, reaction kinetics, involved phases, and transport phenomena plays the crucial role in appropriate reactor and plant design. It is important to determine the limiting steps in time scales and heat management. It is possible to run continuous processes in microstructured devices at the laboratory scale, and these processes can be scaled up to pilot and production scale with low risk. This multiscale production potential from grams to tons allows a high flexibility and offers short time-to-decision in a project.

The Lonza MicroReactor technology represents a consistent and comprehensive scale-up approach and has benefits for the process research and development and manufacturing departments of fine-chemical and pharmaceutical companies. The technology illuminates a clear path from laboratory chemistries to large-scale manufacturing processes, and it completely avoids the parallelization strategies that may result in technical problems. This reactor platform supports rapid process development and production under continuous-flow conditions using microstructured elements. The developed reactor technology is modular, robust, multipurpose, and scalable and has already been tested for several products and processed tons of material during a campaign of a few weeks.

Norbert Kockmann is a senior scientist, Michael Gottsponer is a laboratory chemist, Markus Eyholzer is a laboratory technician, and Dominique M. Roberge* is responsible for business development, all at Lonza, CH-3930 Visp, Switzerland, tel. +41 0 27 948 50 27, fax +41 0 27 947 50 27, dominique.roberge@lonza.com.

*To whom all correspondence should be addressed.

References

1. K. Geyer, J.D.C. Codée, and P. Seeberger, Chemistry 12 (33), 8434–8442 (2006).

2. T.Y. Zhang, Chem. Rev. 106 (7), 2583–2595 (2006).

3. N. Kockmann, Transport Phenomena in Micro Process Engineering, (Springer, Berlin, 1st ed., 2008).

4. J.I. Yoshida, Flash Chemistry, (Wiley-VCH, Weinheim, Germany, 1st ed., 2008).

5. L. Ducry and D.M. Roberge, Angew. Chem. Int. Ed. Engl. 44 (48), 7972–7975 (2005).

6. H. Pennemann, V. Hessel, and H. Löwe, Chem. Eng. Sci. 59 (22–23), 4789–4794 (2004).

7. Ehrfeld Mikrotechnik BTS, www.ehrfeld.com, accessed July 30, 2010.

8. N. Kockmann et al., Chemistry 14 (25), 7470–7477 (2008).

9. D.M. Roberge et al., Chem. Eng. Technol. 28 (3), 318–323 (2005)

10. N. Kockmann and D.M. Roberge, Chem. Eng. Technol. 32 (11), 1682–1694 (2009).