Membrane-based filtration is widely used to harvest therapeutic proteins from their host cells in biomanufacturing. However, it is often plagued with fouling issues that require frequent filter replacement and contributes to low product recovery because of the nonspecific binding to the membrane surface. An industrial-scale spiral inertial microfluidic device developed by researchers at MIT appears to be able to overcome those issues.

Writing in Lab on a Chip, the scientists demonstrated a fast, clog-free CHO cell clarification process with a macroscopic volume process rate of 1 L min−1, which suggests it can potentially scale to a 1,000-L bioreactor. The device also offers high cell-clarification efficiency of approximately 99% (depending on the CHO cell density). Cells can be removed selectively.

“Because we can achieve clogging-free cell retention using the developed plastic inertial spiral device, we can save enormous costs for membrane maintenance and also prevent the loss of product,” first author Hyungkook Jeon, PhD, tells GEN. “We believe the developed microfluidic cell retention device could replace the conventional membrane-based filtration method, improving biomanufacturing efficiency.

“While microfluidics technology has been around for more than 20 years, we started to apply its unique process capability to large-volume applications only recently,” Jeon says. “This is partially inspired by the development of inertial microfluidics, which allows one to enjoy much higher throughput even at the single-chip level. In this work, we are multiplexing to achieve truly industrial volume processing capability.”

Industrial-scale deployments of inertial microfluidics fabricated in a soft elastomer (polydimethylsiloxane, or PDMS) have been difficult because of the time- and energy-consuming fabrication process and channel deformation. Yet the cost of fabricating a hard-plastic device has been too high. To solve those problems, the researchers developed a framework to translate the library of inertial microfluidics designs into deformation-free, mass-producible plastic equivalents.

“To our best knowledge, this is the first attempt to apply a microfluidic system to industry-level bioprocessing, so there are some technical issues we have to consider or verify further,” Jeon says. The need to deal with sterilization of the plastic device, automate the fabrication of the multiplexed plastic spiral unit, and verify the long-term operation of that unit during actual biomanufacturing are examples.

The technology also may have applications beyond biopharmaceutical manufacturing such as removing microplastics and microalgae, filtering yeast, and high-throughput blood fractionation for blood transfusion.