(410a) Comparison of a New Continuous Chromatographic Solvent Gradient Process for Bioseparations with Conventional Processes

The increasing production of highly valuable biomolecules led to a rising attention to the chromatographic purification of these molecules in the recent past. This is partly due to the fact, that there is on the side a growing cost pressure from the market, on the other side is the chromatographic separation step often the largest cost driver in the downstream-processing. The generic purification problem faced mostly is a 3-fraction separation, i.e. weak and strong adsorbing impurities and the product. Due to the widely different adsorptivities of the components to be separated, a solvent gradient is very often required. This type of separation problem is conventionally solved by using a solvent gradient batch chromatography and a subsequent fractionation of the column effluent. A sophisticated variation of this method is the steady-state recycling, where impure fractions with high product content are recycled to the column inlet. The disadvantage of both these techniques is their low productivity and their high solvent requirement due to the inefficient use of the stationary phase and the dilution during elution. Using the principles of counter-current chromatography could overcome these drawbacks, but established processes like SMB (and its derivatives) can only perform binary separations. In addition, the use of solvent gradients in SMBs is limited to a one-step gradient. A new continuous chromatographic solvent gradient process (called MCSGP) has been proposed by our research group recently. This process combines the advantages of batch chromatography and SMB, yielding a continuous, counter-current process which performs a 3-fraction separation and which can incorporate linear solvent gradients. This new MCSGP-process will be compared with the two conventional processes as mentioned above with respect to productivity and solvent requirement for different bio-separation and -purification problems. The productivity and solvent requirement for each combination of process and separation problem is optimized numerically based on a lumped kinetic column and a multi-component, competitive isotherm model. The results show, that the productivity of the new process is about factor 5 higher and that the solvent requirement is about 80% lower than for the conventional processes.