(246e) A Mindset Change from Batch to Continuous Pharmaceutical Crystallization Process Control: The Residence Time Based Feedback Control

Qinglin Su and Zoltan K. Nagy

Davidson School of Chemical Engineering, Purdue University,
West Lafayette, IN 47907, USA

Abstract

The paradigm shift from batch to continuous crystallization in pharmaceutical manufacturing industry in the last decade has also demanded a mindset change in product quality control strategies. For example, the end-product quality control in batch manufacturing is shifted to a steady-state control in continuous operation.

In batch crystallization processes, the product quality is robustly controlled by following an optimal concentration trajectory, thereby containing the supersaturation within a predefined modest range, using a control strategy that is widely known as the concentration control or supersaturation control. It is pointed out that the solubility is manipulated by cooling or adding antisolvent and the batch time or residence time is extendable to adjust to the uncertainties in the crystallization kinetics [1].

In recent development of continuous crystallization control, research efforts often took it for granted by manipulating the solubility directly to maintain a steady-state supersaturation in continuous crystallization process, e.g., using the mixed suspension mixed product removal (MSMPR) crystallizers. However, it is highlighted in this study that the feed solution or slurry adds one more degree of freedom in continuous crystallization control, resulting in the possibility of decoupling supersaturation from the solubility control. This is significant for product yield and subsequent filtrate recycle since the solubility of the exiting slurry can be maintained constant. Moreover, the residence time can similarly play an important role in addressing the uncertain crystallization kinetics, for example, unexpected slow crystal growth would require a lower supersaturation, and smaller feeding flowrate, thus longer residence time for crystal growth, and vice versa.

In this work, three case studies are discussed to demonstrate the importance of this mindset change in continuous crystallization control experimentally and in simulation. The first case study demonstrated the decoupling of solubility and supersaturation control in a two-stage MSMPR antisolvent crystallization process [2]. Therein, the antisolvent feeding flowrate was adjusted to maintain a constant solubility, while the feeding slurry flowrate was regulated to achieve a constant supersaturation under crystallization kinetic uncertainties. It was found that the decoupling control was robust in handling uncertainties in crystallization kinetics. The second case study introduced the periodic flow operation in a three-stage MSMPR cooling crystallization process, wherein the feeding and withdrawing flows were periodically suspended to allow a modest supersaturation and to extend the mean residence time to allow more crystal growth. Final products with significantly larger crystal size were obtained in model prediction and experimentally [3]. The third case study demonstrated the mindset change in the control of nucleation (Direct Nucleation Control, DNC), wherein the decoupling of solubility and supersaturation control offered a larger design space for total counts control, as well as much more stable and smoother control actions.

References