(240c) Membrane Cascades for Downstream Processing

It is proposed here that properly designed counterflow cascades of permselective membranes can compete successfully with existing downstream processing techniques in a significant number of applications. Process chromatography is now well entrenched, largely for lack of a suitable alternative, even though it is inefficient for batch adsorption and many examples of gradient elution, both very widely used. Simulated moving beds are promising for both of these applications, but their inherent complexity and limited percolation velocities ressult in high process costs. Moreover all asorptive processes are poorly suited for processing the large solutes of increasing coomercial interest: large proteins, plasmids, viruses and even eukaryotic cells. Membrane processes have already replaced chromatography for manay size exclusion separations even though even though they are used almost exclusiely in single-stage applications. Available selectivities are rapidly increasing, and high performance tangential flow filtration (HPTFF) has now achieved useful protein fractionations. It will be shown here that the effectvieness of HPTFF can be greatly increased through incorporation of individual modules into a modification of the ideal cascades pioneered by the nuclear industry, and that the already high efficiency of such cascades can be further increased by incorporating cross-flow morphologies into into individual modules. The modules in turn consist of a diafilter for solute fractionation coupled with a solute impermeable filter designed for solvent management. The separating power of such a system will be demonstrtated by specific examples using data already available in the published literature. The primary practical problem of implementing such a process scheme is coordinating the flows of the individual membrane units, and this in turn requires an effective process control system. It is believed that the engineering sophistication of the industries involved has now reached a state putting this goal within reach. Moreover the scale of many promising applications is now large enough to justify a major design effort. A major advantage of this new technology is its insensitivity to diffusional limitations that become ever more stirngent as we deal with larger molecules and even more complex moieties such as viruses,organelles and whole cells. Membranes already exist that are suiteble for this entire range.