Microreactors Eyed for Industrial Use

Chemistry in microreactors has been touted as faster, safer, and cheaper, so it's not surprising that interest in such systems continues to intensify. Microreactors typically consist of plates containing microstructures that define a very small reaction volume, as well as pumping accessories for continuous operation. Now that microreactors are available off the shelf at modest cost, more and more companies engaged in fine chemicals production are testing the technology.

Recently, Sigma-Aldrich took the plunge. Last February, it installed a standard Cytos Lab System microreactor at its R&D facilities in Buchs, Switzerland. Developed by CPC (Cellular Process Chemistry Systems GmbH,, Frankfurt), the system has a list price of about $187,000. Sigma-Aldrich is in good company: Sixteen other pharmaceutical and fine chemicals producers are interested in CPC's technology, according to Sigma-Aldrich's Fabian Wahl, manager of R&D for Europe. Wahl's group is spearheading the firm's technology evaluation.

The key feature of microchemical systems is the high ratio of surface area to volume. That ratio is 200 for a microreactor with a reaction volume of 1.5 mL, such as that of CPC's, and only 0.6 for a 1-m³ reactor. Mass transfer, heat transfer, and mixing are vastly more efficient in a microreactor, allowing far more precise reaction control and far better product quality control than can be achieved with conventional reactors, Wahl says.

In addition, an optimized process can be run at any scale without further R&D by simply running in parallel as many reactors as are required. Such a mode of operation is easily executed because microreactor systems typically are modular.

The technology would reduce the cost of scale-up and resources deployed to develop safe and stable processes, Wahl concludes. It is useful and practical for Sigma-Aldrich's businesses, particularly for developing and making catalog products.

Many of Sigma-Aldrich's catalog products are produced under typical lab conditions in flasks of up to 20 L. Of the more than 2,000 compounds in this portfolio, about 800 could be produced in microreactors with little or no process modification, Wahl says. For such cases, microreactors would reduce reaction time and cost.

For example, the condensation of 2-trimethylsilylethanol and p-nitrophenyl chloroformate to produce 2-(trimethylsilyl)ethyl 4-nitrophenyl carbonate requires 14 hours to complete in a conventional setup but only 18.4 minutes in a microreactor. Because the time is so brief, the possibility that the product will degrade or that by-products will form is vastly reduced, Wahl says.

Even more attractive is the opportunity microreactors offer to run problematic chemistries, such as those that are highly exothermic, produce unstable products, or form difficult-to-separate side products. Wahl's team has explored use of microreactors for two such reactions.

The first is ester hydrolysis to produce an alcohol that readily degrades. Wahl declines to identify the ester or the alcohol. He says only that Sigma-Aldrich could not keep up with the demand for the alcohol because yield deteriorates as the reaction is scaled up: 70% at 5 L, 35% at 20 L, and 10% at 100 L.

Because Sigma-Aldrich's original process requires an insoluble component, it could not be carried out in a microreactor. "We had to change the chemistry, but we did not know in which way," Wahl says. With microreactors, they found out quickly.

IT TOOK LESS than a day to run a model reaction under 12 different conditions. A substrate that yields a stable alcohol was used so as not to complicate process development. After the best conditions were identified for the model system, it took only two hours to optimize the reaction of the actual substrate.

The second problematic chemistry tested by Wahl's team is preparation of methylenecyclopentane from a substrate that Wahl does not wish to disclose. The reaction is highly exothermic, and process control is not good under conventional conditions. Furthermore, up to 30% of the yield consists of the more thermodynamically stable product, 1-methylcyclopentene, which is difficult to separate. For these reasons, production of methylenecyclopentane had been discontinued.

Using the CPC microreactor, Wahl's team devised a reaction that gives 70% conversion, no by-product, and a throughput of 300 g per hour. A 70% conversion with no by-product is better than a higher conversion with some by-product because separating the product from the precursor is easier than separating it from the by-product, he explains.

Wahl expects microreactor technology also to have a positive impact on Sigma-Aldrich's custom synthesis business, in which rapid process development is a competitive advantage. His expectation that microreactors would reduce development time by 40% has been met so far. Thus, Sigma-Aldrich is inclined to adopt microreactor technology more widely, he says.

The major drawback is the inability to work with gases and insoluble reagents. A minor one is the fact that the CPC microreactors are fabricated from steel, which is chemically sensitive. CPC is now testing microreactors made with Hastelloy, a chemically resistant material.

The CPC microreactors are easy to work with, Wahl says. The most common problem is blockage, which often is solved with one call to CPC's hotline. Changeover from one process to another requires only a five-minute rinse with solvent.

Last month, Wahl visited Massachusetts Institute of Technology for a close look at the microchemical system being developed by Klavs Jensen, a professor of chemical engineering. The MIT system is different from CPC's in three major ways. First, the reaction volume is in the microliter range, making lots of experiments possible with minuscule amounts of materials. Second, the reactor is made of silicon, which can be oxidized to form glass surfaces familiar to chemists. Because silicon and silicon dioxide (glass) surfaces are easy to modify, it is possible to fabricate reactors that are preloaded with catalysts, Wahl says. Third, the structured surfaces of the MIT system are covered with glass, so reactions can be observed as they take place.

The MIT system is not yet available commercially. But when it becomes available, Sigma-Aldrich will be considering it for possible adoption into its business operations, Wahl says. Meanwhile, Sigma-Aldrich will be closely following the further development of MIT's technology, hoping for a chance to test a prototype.