Jani Siitonen

24 Aug 2012


2009 Master of Science, Lappeenranta University of Technology, Finland.

2009 Project engineer, Lappeenranta University of Technology, Finland.

2010– Doctoral student, Lappeenranta University of Technology, Finland. Topic:

Theoretical analysis and design methods for advanced separation processes based on chromatography and membrane filtration.

2011 Research visit at Nanyang Technological University, research group of Asst Prof. Arvind Rajendran, Singapore.

Abstract


Hybrid separation process – steady-state recycling chromatography with an integrated membrane filtration unit

A process concept where a membrane filtration unit is integrated to a steady-state recycling chromatography (SSR) is analyzed theoretically. In an SSR process, the performance of single column chromatographic separation is enhanced by recycling the unresolved part of the chromatogram. The pure leading and trailing sections of the elution profile are collected, while the unresolved middle part is mixed with fresh feed and re-injected into the column. The recycling fraction is typically more dilute than the fresh feed. This limits the amount of fresh feed that can be processed per cycle. The process performance can be enhanced by removing some of the solvent from the recycle fraction, e.g. by using membrane filtration (SSR–SR). The membrane filtration unit can be integrated to different positions of the process. In this study, the performance of the following three SSR–SR configurations is investigated: solvent removal from: I) the fresh feed, II) the recycle fraction, and III) the actual feed solution into the column (obtained by mixing the fresh feed and the recycle fraction).

A method is developed to choose the cut times for fractionating the outlet stream of the chromatography column and the capacity of the membrane filtration unit such that user-given purity and/or yield requirements are satisfied. The amount of fresh feed that can be processed per cycle and the injection volume into the column are identified as the only free operating parameters. It is shown that the three SSR–SR configurations have identical performance with the same operating parameters. In contrast, the configurations differ with respect to the maximum amount of fresh feed as well as the range of feasible injection volumes.

The effect of concentration limits on applicability and performance of different SSR–SR configurations is studied. The following two types of limits for the extent of solvent removal are discussed: maximum concentration achievable in the membrane filtration unit (solubility or osmotic pressure limit) and maximum concentration of the solution fed into the column (viscosity limit). It is observed that the process configuration where solvent is removed from the column feed (configuration III) is typically the most flexible one with respect to the operating parameters and provides highest productivity. In addition, it is shown than an SSR chromatography with a membrane filtration unit yields higher productivity and lower eluent consumption than an optimized batch chromatography process that employs a similar solvent removal unit or a conventional SSR process without solvent removal.