Reconfigurable refrigerator sized flow platform for end-to-end manufacture of multiple drugs

Last month Adamo et al published a hot Science paper on a compact continuous flow synthesiser for four drugs: luoxetine hydrochloride, lidocaine hydrochloride, Diazepam and diphenhydramine hydrochloride. The group had previously reported an end-to-end manufacture of aliskiren hemifumarate (see Angew. Chem. Int. Ed., 2013, 52, 12359–12363), which included the synthesis, purification, and tableting of the drug.  The Science paper represents a significant advance in terms of reducing plant footprint and showing that a continuous processing setup can also achieve the similar flexibility as a batch manufacturing plant, it includes the upstream (reaction) and downstream (purification and formulation) processing. Researchers at MIT have also published an economic analysis of a batch vs flow integrated flow process (see Ind. Eng. Chem. Res., 50, 10083-10092 (2011) doi:10.1021/ie2006752), presenting a persuasive argument for the drive to continuous manufacturing.

Opening the supporting information really does highlight the that this truly is an integrated and uninterrupted processing strategy, with real-time process monitoring. The reactor unit is modular in the size of a refrigerator, thus enabling a plug-and-play platform and the flexibility needed for flow to be competitive with established batch manufacturing. Many of the components used in the configuration were custom built, including PFA tubing within aluminium casing and membrane separators with high pressure ratings. The system includes in-line, real time analysis using FlowIR from MettlerToledo for qualitative reaction analysis, automated gravity separators were incorporated for monitoring the organic and aqueous phase separation, and in-line ultrasonic monitoring for measuring ultrasound velocity for formulation.

There is a table within the ESI which shows the processing time for the different unit operations with the upstream processes all taking less than 2 hours for each API and the downstream processing taking significantly longer (about 10 hours for each drug), and it is the downstream processing achievements is where this contribution is the most powerful.

I don’t envisage a future of having a TV size reactor inside everyone’s house dispensing personalised medicines as this would be a regulatory nightmare, and in reality too many stock chemicals are necessary to access the drug catalogue. I think the portability of flow is needed in hospitals for personalised medicines and for on-demand delivery in hospitals, such as for the generation of radiopharmaceuticals. To dream further, another application would be in the human spaceflight to Mars anticipated to take place by 2035.

The researchers are now attempting to reduce the footprint further and widen the drug catalogue. It will be interesting to see how much re-optimisation of the reactor unit and chemistry will be needed to access more drugs. I look forward to seeing the future results.

Read the Science article online: Adamo et al., Science, 2016, 352, 61-67, DOI: 10.1126/science.aaf1337

Unlocking the potential of biocataysis using a microfluidic toolbox approach

Christopher Hone

Enzyme-catalysed reactions are normally conducted under mild conditions (dilute concentration and low temperature) and limited solvent compatability. But industrially we strive to reduce waste and costs. Nature does not provide a clear answer. The question synthetic chemists need to answer is “how do we achieve this compatibility?”

Researchers at the Technical University of Denmark outline a holistic solution using in situ product removal (ISPR) for optimal process design.

A common cited example of batch processing is its flexibility. The answer in flow is a plug and play approach.

Substrates and products often have inhibitory and toxic effects in enzymatic processes. Removal of product as it is formed could enable process intensification in flow. Separation proccesses which exploit the differences in physiochemical properties (size, volatility, charge and solubilty). Each of these physical properties have unit operations associated with them, for instance membranes and filtration for size separation. The unit operations are made available as modular units for experimentation.

The application of the approach formalised by Heiritz et al could reduce the barrier in designing biocatalytic reactions in flow and result in the design of more efficient processes.

Read the article online at
A Microfluidic Toolbox for the Development of In-Situ Product Removal Strategies in Biocatalysis
Søren Heintz, Aleksandar Mitic, Rolf H. Ringborg, Ulrich Krühne, John M. Woodley and Krist V. Gernaey, J. Flow Chem, ASAP.

Chinese Lucky Knot: A Novel Microdevice for Non-Linear Concentration Gradients


The Chinese lucky knot dates back to ancient times but a recent article has reported it as inspiration for the development of a micromixer for non-linear concentration gradient generation. The project is a joint collaboration between Chinese and Brazilian research groups.

Normally large volume devices and relatively high flow rates are necessary to achieve optimal mixing, but these are impractical for handling high value chemicals and analytes. Hence there is a drive for scale-down of microdevices that have predictable mixing characteristics. After comparing many Arabic and Chinese designs, the Chinese lucky knot proved to provide the best mixing. Optimal mixing was achieved for flow rates between 10 and 300 µL/min, which was proven using experimental validation and a computational fluid dynamics study. The device comprises of a central grid matrix. More outlets can easily be added by increasing the size of the micromixer.

The group are currently applying the technology for measuring reaction kinetics and for photoreduction chemistry. Read more about their work using the link below.


Read the article online using the link below

DOI: 10.1556/JFC-D-13-00021

J. Flow Chem., 2014, 4, 61

Immobilised metal-free catalysis for enantioselective synthesis using microreactors

A recent article in the Journal of Flow Chemistry reported the first immobilisation of enantioselective organocatalysts into a microreactor. Rather than immobilise the catalyst within the reactor as a packed bed, the group from the Netherlands used polymer brushes as the catalyst support instead. The main problem with packed bed reactors is the reagent may not go through the beads properly and so does not interact with the catalyst. It can also be difficult to achieve efficient heat removal from the centre to the wall in a packed bed reactor.

The polymer was covalently bonded to the inner walls of the flow reactor using glycidyl methacrylate. The swelling of the polymer beads is crucial to performance. The extent of swelling relates to the accessibility of the substrate to the catalytic sites. Polar solvents (chloroform, dichloromethane, methanol and water) displayed the highest activity.

The first enantioselective catalyst coated to the walls was cinchona alkaloid (chinchonidine) and its catalytic activity was tested using a Diels-Alder reaction between anthrone and N-substituted maleimidines. After 52 min reaction time at room temperature 31% conversion and 17% ee was obtained. A second chiral organocatalyst, a quinidine derivative (not drawn), was anchored to the microreactor wall, and in this case the conversion was doubled to 54% and 39% ee achieved. Reducing the temperature to 0 °C increased ee to 49% but lowered the conversion.

The immobilisation of a proline catalyst was also achieved, and applied to an Aldol reaction between 4-nitrobenzaldehyde and cyclohexanone. The reaction conditions used were water as solvent, room temperature and a 52 min residence time. 23% conversion, anti:syn 83:17, and 93% enantioselectivity were achieved. The group successfully showed that the supported catalyst achieved exactly the same result as homogeneous conditions. After operation for 20 days no drop was observed. Immobilisation of the catalyst had no impact on performance in terms of conversion and enantioselectivity. Polymer brushes Polymer brushes can cause a productivity (kg product/per volume reactor) limitation since there is a limit to the amount of active catalyst (lower overall catalyst loading) which can be attached to the wall. The paper represents a significant advance for the flow chemistry community because this new organocatalyst technology displayed moderate conversion and ee values were obtained which is a promising step forward for scalable asymmetric organocatalyst continuous processing, and the catalyst activity was successfully maintained for a month.

Read the article by the group online using the link below

Rajesh Munirathinam, Andrea Leoncini, Jurriaan Huskens, Herbert Wormeester and Willem Verboom,  J. Flow Chem., 2015, 5, 37–42.

DOI: 10.1556/JFC-D-14-00034

Under Pressure for Antihypertensive Drugs? Decide to Go with the Flow

High blood pressure has been implicated in 16.5% of all deaths for cardiovascular disease. Telmisartan 1, an antihypertensive drug, has become an opportunity for generic manufacturers since its patent has recently expired. Researchers from Virginia Commonwealth University have reported an elegant three-step convergent continuous-flow process for the synthesis of Telmisartan 1 in the Journal of Flow Chemistry.


The Virginia-based group had previously reported a batch synthesis which included the isolation of intermediates en route to Telmisartan 1. Since then, the team had significantly streamlined the process.

A key challenge for multi-step flow processes is identifying conditions for each reaction which are compatible with all subsequent reactions. In this example all the reactions were compatible with basic conditions which circumvented the need for an in-line liquid-liquid separation. A mixture of NMP and water gave improved solubility in comparison to earlier attempts thus prevent the need to handle solids in flow. In addition this solvent mixture could be used for all the reactions so circumvented the need for solvent exchanges.

The Vapourtec E-Series flow system, fitted with two 10 mL reactor coils in series, was utilised for the alkylation and deesterification reaction steps. The outlet solution was then collected into a reservoir ready for the next reaction. KOH was then introduced for the deesterification reaction. The alkylated benzimidazole (from the reservoir) and the bromobenzimidazole were introduced via two separate inlet streams to the ThalesNano X-Cube for a Suzuki cross-coupling between the two components. This system can hold packed bed cartridge units, in this case it was a solid-supported Pd catalyst (SiliaCat DPP-Pd). The reactor was operated at 180 °C and a residence time of 5 min to give Telmisartan 1 in 97% conversion.

The flow chemistry route developed by Martin et al. represents a significant advance since it includes no intermediate purification points or solvent exchanges which cuts down on the number of unit operations needed. The absence of these stages drives up process efficiency by reducing costs and waste generated. The approach used by the group is an elegant alternative for accessing this structural class of compounds.

  1. D. Martin, A. R. Siamaki, K. Belecki and B F. Gupton, J. Flow Chem., ASAP

DOI: 10.1556/JFC-D-15-00002

The Flow Reactor for a Green Future

Self-optimisation using continuous flow reactors is an enabling technology for the rapid optimisation of chemical reactions. A self-optimising system combines online analysis, automated reactor control and a feedback algorithm to rapidly optimise reactions. It uses online analytical results to calculate the new input conditions (flow rates, pressure and temperature) for the subsequent experiment. In this example, the Poliakoff group at the University of Nottingham investigate the methylation of 1-pentanol by dimethyl carbonate using supercritical CO2. The study included the optimisation of several reaction metrics – yield, space-time yield (STY), E factor (kg waste/kg product) and a weighted yield function (the product of space-time yield and yield).


Professor Martyn Poliakoff is also a YouTube sensation for videos about the different elements of the periodic table (they have been viewed by millions). He was awarded a Knighthood in the UK’s New Year’s Honours List. An interview with Prof Martyn Polikoff is available by clicking the video below.

You can read the article in full online using the link below.

Denis N. Jumbam, Ryan A. Skilton, Andrew J. Parrott, Richard A. Bourne and Martyn Poliakoff, J. Flow Chem., 2012, 1, 24–27

DOI: 10.1556/jfchem.2012.00019

Flash Chemistry: Achieving Polymer Synthesis Not Possible Under Batch Conditions

By Christopher Hone

The Yoshida group in Tokyo, Japan, have reported the benefits of flash chemistry for a wide variety of chemical transformations. Flash chemistry is based on high resolution reaction time control using small scale flow systems (ability to run reactions on the timescale of seconds!). Flash chemistry is simply not possible using batch apparatus. Small scale flow systems offer precise control of the reaction parameters such as rapid heat transfer for isothermal temperature control. Fast mixing for conducting very fast reactions provided by the T-shaped micromixers and the relatively high flow rates (between 4 mL/min and 10 mL/min in this example).

Nagaki recently reported the use of flash chemistry for the synthesis of fluorine containing anionic block co-polymers incorporating perfluoroalkoxycarbonyl groups. This synthesis is unique because it is an example of a living polymerisation method which does not require a capping agent – which is normally necessary for anionic polymerisation. Fluorine groups display strong electron withdrawing effects hence have unique properties such as low surface energy, oil and water repellence, low refractive index, and high stability.

Lithiation reactions normally need to be at −78 °C in temperature when operated under batch reactors to avoid side reactions. Flow chemistry enables more practical reaction temperatures to be used. Additionally lithium chloride is often added to reduce the number of side reactions when using organolithium reagents. Nagaki successfully synthesised block co-polymers without the need for the addition of lithium chloride, which would otherwise have contaminated the final polymer product. They showed that the concept of flash chemistry could be applied to sequential polymerisation. An example is shown in the Figure for ABA polymerisation perfluoro methacrylate – tert-butyl methacrylate – perfluoro methacrylate, an ABA polymer. Highly defined block polymers were prepared with precise control over the polydispersity index and molecular weight.


You can read the full article online by the Yoshida group using the link below.

Aiichiro Nagaki, Kana Akahori, Yusuke Takahashi and Jun-ichi Yoshida, J. Flow Chem. 2014, 4, 168–172

DOI: 10.1556/JFC-D-14-00017