(438a) Predicting the Productivity of Chromatography Processes By Repeated Cyclic Operations or By Continuous Column Switching Operations

Chromatography processes are commonly optimized in terms of the productivity P defined as P = (mass recovered)/[(bed volume)(cycle time)]. P can be increased by either pseudo-continuous multi column operations such as simulated moving bed (SMB) chromatography or a single column repeated cyclic operation (RCO). It is first needed to optimize a RCO process in order to evaluate the benefit of continuous chromatography. We have developed a method for predicting the P in RCO isocratic elution processes by considering the distribution coefficient K and HETP as a function of mobile phase velocity. A model system chosen was reversed phase chromatography of two polyphenols with an ethanol-water mixture mobile phase. In this system the retention and the selectivity (the difference between the two distribution coefficients, ΔK) can be changed by the ethanol concentration markedly. The calculation showed the highest P value (considering mobile phase consumption) at 24-26% ethanol concentration. Based on these optimized conditions, we performed numerical simulation of a standard 4-zone SMB separation process. The column height of SMB was one-fourth of the RCO. The sample and the solvent feed rates were calculated from the optimized RCO conditions. At 24% ethanol concentration where ΔK was large(ca.2.5), the performance of RCO was better than that of SMB. At 26% ethanol concentration (ΔK=1.5), the performance of SMB was similar to that of RCO. At 28% ethanol concentration (ΔK=0.9), SMB productivity was higher by 10-20%. These results show that RCO is suited as a pseudo continuous chromatography operation when the selectivity is high (large ΔK).