Item

Abstract

Growing demands for products with constant high quality make continuous crystallization very attractive. Carrying out a process at a steady state can guarantee the continuity of product properties. This might be especially important, when one considers drugs produced in an enantiopure form. Within this contribution the idea of a Mixed Suspension Mixed Product Removal (MSMPR) crystallizer for preferential crystallization is analyzed both experimentally and theoretically [1]. Preferential crystallization is applied to resolve one enantiomer from a racemic mixture of both. The separation process takes place in the metastable zone, where spontaneous nucleation is suppressed kinetically. Preferential crystallization is initiated by introducing seeds of only one enantiomer to the solution. The surface of these seeds drives resolution due to their growth. Executing preferential crystallization in a continuous mode requires a continuous supply of seeds together with racemic feed solution and a continuous withdrawal of the product suspension containing the seeded target enantiomer. To gain insight into the process a mathematical model was developed assuming an ideally mixed vessel, uniform conditions inside the tank, identical crystals size distribution of crystals in the vessel and in the product stream. Population balance and mass balance equations as well as kinetics for crystal growth, primary and secondary nucleation were used to simulate the behavior of the solid and liquid phases. A simulation study using parameters of the enantiomer of threonine [2] was done to investigate the effects of different conditions on the process performance, including the saturation and crystallization temperature as well as mass flow rates of seeds and the feed solution. An optimum set of process conditions was chosen considering also the aspect of reducing the duration of the dynamic start-up phase. The experimental part of the contribution will focus on the following aspects: first, observed resolution results of L- and D-threonine will be compared with model predictions. Effects of selected conditions will be discussed to identify the influence of key process parameters on the length of the dynamic start-up phase and the product quality (purity and mean characteristic size of the product). A comparison of the continuous process with the performance of a corresponding batch process will conclude this contribution. [1] Jones A.G., Crystallization Process Systems, Butterworth-Heinemann, Oxford, 2002 [2] Qamar S., Elsner M.P., Hussain I., Seidel-Morgenstern A. Seeding strategies and residence time characteristics of continuous preferential crystallization Chemical Engineering Science, 71, 5-17, 2012