Online control development and process intensification in continuous chromatography (MCSGP)

Abstract

The multi-column countercurrent solvent gradient purification (MCSGP) process is studied in this work. The process was developed some years ago by Aumann and Morbidelli [1], some further process development is still going on. In this work, the process itself was applied for two challenging separation problems; a monoclonal antibody (mAb) charge variant separation and the purification of rare earth elements (REE). The charge variants of three mAbs were separated, namely Avastin®, Herceptin® and Erbitux®. Especially Herceptin® should be mentioned, as the bio activity of the charge variants was measured for this mAb in the past [2]. Therefore, the benefit of removing inactive variants could be quantified. All three mAb separations were performed successfully in cation exchange batch chromatography and MCSGP, with MCSGP having an up to 4 fold higher productivity. The activity of the mAb Herceptin® could be increased by 30% in comparison to the original mixture which is available as a therapeutic. Additionally, the stability of the MCSGP process was verified applying feed which was enriched in impurities. It was found that the feed composition was not influencing the product quality significantly. The separation of REE is becoming an increasingly important challenge, as the demand of REE is increasing worldwide and several production sides have been closed due to environmental concerns and too high costs [3]. In the past, chromatography was an often applied tool for the purification of REE, but was replaced by extraction in the last 40 years. Implementing a continuous chromatographic process as the MCSGP providing higher productivity than batch chromatography the purification of REE based on chromatography could become economically interesting again. In addition, the chromatographic steps could significantly lower the environmental concerns due to milder conditions and thereby reinitiate the production of REE in western countries. In this work, three REE species were separated (Praseodymium, Cerium and Lanthanum) applying batch and MCSGP chromatography. The MCSGP process could outperform the batch significantly, improving the productivity by factor 5 to 15. However, the cation exchange chromatography only reached moderate absolute productivities, as the solubility of the REE in the mobile phase was very low. To overcome this problem, a different mobile phase would have to be applied as proposed recently by Hansen et al. [4]. The implementation of this mobile phase could increase the productivity of the MCSGP process to the order of kilograms per day and liter of resin. MCSGP in its original setup is able to separate three fractions, which is sufficient for most encountered problems. In rare cases where multiple products in a mixture are valuable, the separation of more than three fractions would be beneficial. In this work, the theoretical design of a MCSGP capable of separating n fractions and its design starting from batch chromatography is discussed. In theory, applying a MCSGP process with n columns can separate n fractions as well, in experimental application; the maximum number of fractions was limited to 4 due to pressure drop issues. The four fraction separation was validated applying a model mixture of four proteins and a mAb charge variant separation. In both cases, the yield could be increased applying MCSGP keeping the purity at the same level.