The described methods are particularly applicable for use in cross flow filtration. Results show that good mixing in a relatively short time can be achieved for all conditions tested if sufficiently high recirculation flow rates are used. In many situations, this can be attained in three minutes or less.
In Application Note 29-0187-32, we have investigated the phenomenon using a tubing loop to recirculate solutions via convective forces in disposable 100 L and 200 L 3D rectangular bags. In Application Note 29-0187-31 studies are presented that demonstrates recirculation of 1 cP and 5 cP viscous solutions in 50 L, 2-dimensional (2D) pillow bags using a tubing loop and peristaltic pump.
The growing interest in using disposables in biopharmaceutical processing can be seen in the increasing use of single-use bags instead of stainless steel vessels. The driving force behind this switch is the desire to minimize cleaning and cleaning validation for product and buffer containers. Disposable systems also afford increased design flexibility. GE Healthcare Life Sciences, ReadyToProcess product platform brings a plug-and-play alternative to multiple bioprocessing unit operations including cross flow filtration.
The change from fixed-geometry stainless steel vessels to disposable bags generates new mixing considerations. Apart from magnetically-driven impellers, which often are the option when employing disposable bags, an alternative method of mixing is to simply recirculate the solution within the reservoir bag. Because cross flow filtration generally requires that the bulk suspension (retentate), processed by the filter, is actively circulated, in-bag recirculation is suitable for this mixing application. Cross flow retentates are sometimes higher in density and/ or viscosity than their respective feed streams. Thus, mixing behavior in disposable cross flow filtration reservoir bags is of particular interest.
In Application note 29-0187-31 the mixing of homogeneous solutions in 50 L, 2D pillow bags is demonstrated. The effects of liquid density and viscosity differences, vessel working volume, and power input per unit volume (i.e. recirculation flow rate) are presented. Of special interest is determining the minimum working volume of the reservoir bags. Application note 29-0187-32 describes studies on the mixing of homogeneous solutions spiked with a high salt concentration solution in 100 L and 200 L 3D rectangular bags. It examines the impact of parameters such as liquid density and viscosity, vessel design and power input (i.e. recirculation flow rate).
Results and conclusions
All results are presented in detail in the two Application Notes. The conclusions from the studies using the 50 L, 2D pillow bags were that during cross flow filtration, good mixing in relatively short time spans (T95 < 6 min) is generally achieved. Good mixing is achieved at low as well as high viscosity, with both upper and lower inlet configurations, and over a broad range of recirculation flow rates. In some cases, T95 is attained in three minutes or less. Therefore, the necessity of the use of an exogenous mixer (impeller) can be avoided with one single reservation: to assure proper mixing at high viscosity, low working volumes should be avoided or external mixing (bag manipulation) applied. For 50 L, 2D bags, decreased convective flow and high viscosity lead to increased mixing times. Very low working volumes affect mixing and are due to changes in bag geometry.
There were similar conclusions from the studies using the 3D bags. Good mixing in a relatively short time span (T95 < 6 min) can be achieved for all conditions tested for typical flow rates used in hollow fiber processes. In many situations, this can be attained in three minutes. Mixing in disposable bags during cross-flow filtration operations can therefore be achieved using recirculation flow without having to resort to bags with an incorporated impeller. For both 3D bag sizes (100L and 200L), the recirculation return port has to be directed into the bulk liquid in order to enable mixing.
In both cases, the contour plots from the Design of Experiment studies generally reveal large operating ranges with a wide window of conditions. Process developers can therefore choose the correct conditions that will give them good control over the mixing. When extreme conditions leading to poor results are avoided, good mixing can be achieved in most situations.