Modeling the performance of pilot-scale countercurrent chromatography: scale-up predictions and experimental verification of erythromycin separation.

Abstract

Biosynthesis of polyketide antibiotics, such as erythromycin A (EA), can result in the formation of analogues of the main compound that are chemically and structurally extremely similar. The large-scale purification of these antibiotics by conventional high-performance liquid chromatography (HPLC) can be prohibitively expensive due to the large volume of both solvent and adsorbent required. This study examines the feasibility of using a novel pilot-scale countercurrent chromatography (CCC) machine as an alternative to HPLC. CCC is a low-pressure (typically <4000 kN m(-2)) liquid-liquid chromatographic technique that allows the separation of solutes on the basis of their partitioning between two immiscible liquid phases. The effects of mobile phase flow rate, column rotational speed, and sample injection volume on the attainable yield and purity of EA were investigated. Our results show that, at a mobile phase flow rate of 40 mL min(-1), a rotational speed of 1200 rpm, and an injection volume of 100 mL (10 g total erythromycin), EA could be satisfactorily fractionated with a purity of approximately 92% (w/w) and a recovery yield of approximately 100% (w/w). The total solute throughput was estimated to be 0.41 kg day(-1). More importantly, we demonstrated simple and predictive linear scale-up of the CCC separation based on data obtained from a single laboratory-scale CCC chromatogram, and verified this experimentally. The retention time and peak width of the target compound at the pilot scale could be predicted to within 4% for operation at a range of mobile-phase flow rates and injection volumes. This predictable nature of CCC separations, unlike HPLC methods, can greatly reduce process development times and enable a complete process-scale operating scenario to be planned.