The Cutting Edge of Continuous

Oct. 12, 2010
Manufacturers look outside their walls for flow chemistry expertise.

Continuous flow processes are gaining a foothold in drug manufacturing, yet manufacturers tend to have little expertise in house. Outsourcing is a natural solution. Earlier this year, i2Chem was founded to help manufacturers synthesize compounds via continuous processes.

The company has roots in the Novartis-MIT collaboration, and is based upon technology developed by Klavs Jensen, Head of the Chemical Engineering Dept. at MIT, and Peter Seeberger, formerly of MIT and now of the Max Planck Institite in Berlin. i2Chem is using continuous flow reaction technology to optimize reactions, develop new synthetic routes, and even provide compounds and production systems for manufacturers, says president James N. Little, who stepped down as chairman of Microfluidics Corp late last year.

We spoke with Little about what i2Chem does, and what manufacturers are looking for. (For a PDF presentation detailing some of the specific flow technology work that i2Chem does, click here.)

PhM: You’re a new company, but your technologies have been around for some time, correct?

J.L.: Our technical founders started working in this area in the late ‘90s. They have about 100 man-years of experience in flow chemistry. This allowed i2Chem to start work immediately on projects for customers. Usually a startup has to spend a year or more doing research before they are ready to go. We started working on customer projects day one.

PhM: Is it safe to say we’ve reached a tipping point in the acceptance of continuous flow chemistry in pharma? (Put another way, why did you start i2Chem this year?)

J.L.: I don’t know if we have reached the tipping point yet. People are reading about it, thinking about it, but only a few have tried flow chemistry. It is an ideal time to enter the market as many understand the benefits and are willing to give it a try.

PhM: Is there a real lack of expertise within most pharma companies in terms of flow chemistry and how to convert batch processes? Do companies need to hire outside help such as yours if they’re going to convert?

J.L.: Most large pharma have invested in flow chemistry, some quite substantially. Others only have a small but growing commitment. Some reach out for help, while others are self sufficient. The biotechs are the ones that outsource much of their chemistry today, but any manufacturer with non-ideal results in batch are willing to try flow.

PhM: You’re always asked, can batch be done in flow? How often is the answer yes? Is there really any limit to the types of processes that can be converted to flow, and thus optimized?

J.L.: The no-no in flow is solids. These are very difficult to do in flow unless particles are small and don’t agglomerate. We use temperature and pressure to speed up reactions, usually reducing hours to minutes in a flow system. But if temperature and pressure don’t speed up a reaction significantly, you should stick to batch. 

PhM: The type of equipment needed for continuous processes is extremely chemistry- and application-dependent. Has this been an issue in the industry’s adoption of flow chemistry, that processes must be highly specialized? Do you envision more standardization at some point?

J.L.: The technology is still very new and is very chemistry- and application-dependent. The field is not ready for standardized products yet. It is difficult to make a lot of chemistry reactions fit a standard product. We feel this is one of our strengths—we figure out the best chemistry in flow then pick the flow components to keep the chemistry optimized.

PhM: Give us your best ROI arguments for continuous flow processes. What do you tell manufacturers with new compounds who are weighing batch vs. continuous?

J.L.: This is still very reaction-dependent. The one thing that does stand out is that flow systems take up less space in a plant.

PhM: If I came to you today with, say, Compound X, and asked you to set up a continuous operation, what would be involved? What essential steps would you take, and what’s a ballpark timeline by which I could have a continuous line set up on my plant floor?

J.L.: We first look at all the steps of the reaction. We know that some steps are better left to batch; so a final production system could have a mix of batch and flow. We then propose what steps can be converted to flow. We run experiments on each step and figure out the best flow conditions for each step. Once this is done, we then optimize how to connect each of the flow steps together in an optimized continuous process. From this, we can now spec out the components for the production flow system. The timeline is dependent on the number of steps being converted. If only one step, it could be up and running in production in 3-4 months. If more steps are involved the timeline is extended.

About the Author

Paul Thomas | Senior Editor