Novel simulated moving bed processes for antibiotics purification

Chad Austin Farrenburg, Purdue University

Abstract

Purification is an important process in pharmaceutical production. Maximizing yield while maintaining required purity is paramount to reducing purification costs, but this goal is difficult to achieve when the product and its impurities are very similar. Crystallization is often used is this case, but results in a substantial loss of yield. This research develops a simulated moving bed chromatography (SMB) process for purifying an important antibiotic, Clarithromycin (6-o-methyl erythromycin), from a similar impurity, 6,11-o-methyl erythromycin, using a systematic design approach. First, appropriate adsorbent and mobile phase combinations (adsorption systems) are found from preliminary experiments. A series of single-column chromatography experiments are then used to estimate the adsorption and mass transfer parameters. Second, a design method for four-zone SMB processes that have non-linear adsorption isotherms and significant mass transfer effects was developed in this research. This method provides the highest throughput and the lowest solvent consumption for the isocratic four-zone SMB process. Rate model simulations confirm that the design method can guarantee high purity and yield. It is demonstrated that this design method can be applied to both Langmuir and Bi-Langmuir adsorption isotherm behavior. Because most real adsorption systems have these properties, this is an important development. This method is especially important when selectivity is low and cost minimization is of great importance. A five-zone SMB process is a novel development of this research that improves upon the typical four-zone SMB process by substantially reducing solvent consumption. The non-linear, non-ideal design equations developed for four-zone SMB processes are modified to determine optimal operating conditions for the five-zone SMB process. This process is tested using a series of lab-scale SMB experiments. These lab-scale SMB experiments are also used to validate the design method and intrinsic parameters. These intrinsic parameters are used to develop plant-scale designs of a four-zone SMB process and a five-zone SMB process using the most promising adsorptions systems. From these plant-scale designs, cost estimations are used to determine the most cost-efficient process for removal of 6,11. In addition, further processes are proposed for removing the remaining impurities.

Degree

Ph.D.

Advisors

Wang, Purdue University.

Subject Area

Chemical engineering|Pharmacology

COinS