CHO cell deposition during initial stages of clarification using the D0HC commercial depth filter in a Micro 20 capsule

Depth filters are used extensively for initial clarification of cell culture broth in bioprocessing, both as a stand-alone unit operation and/or after centrifugation. Although a number of efforts have been made to apply classical fouling models to describe the performance of different depth filters, there is still considerable uncertainty over the nature and location of cell deposition on/within the depth filter. In this work, scanning electron microscopy and fluorescence microscopy were used to evaluate both the location and extent of Chinese Hamster Ovary (CHO) cell deposition within the Millistak+® D0HC depth filter in Micro 20 capsules during the initial phase of filtration. No cell deposit / cake was visible on the external filter surface, with the cells captured within the depth of the D0HC media. The fluorescent images show a significant spatial variation in cell deposition, with a much higher number of cells deposited at the bottom part of the filter when the Micro 20 capsule is oriented vertically (with the feed flow perpendicular to gravity), due to the greater density of the cell suspension compared to that of the buffer initially used to flush the filter. This behavior was well-described by the analysis of the velocity profiles within the Micro 20 capsule by computational fluid dynamics accounting for buoyancy effects on the flow. This spatial variation was absent when the capsule was oriented horizontally (with the feed flow parallel to gravity). Cell capture within the depth of the D0HC media was highly non-uniform, with some regions of the filter showing nearly close-packed cell deposits. The number of cells decreased with depth through the filter in both orientations, consistent with classical models for depth filtration. These results have significant implications in further understanding the fouling mechanisms during depth filtration for initial clarification of cell culture fluid.