Dynamic PCA-based MSPC charts for nucleation prediction in batch cooling crystallization processes

Real time monitoring of unit operations in manufacturing of pharmaceuticals is one of the main aspects included in the concept of process analytical technology (PAT). Crystallization is an important purification unit operation in manufacturing of pharmaceuticals. To obtain the desired product properties, e.g., size, shape and polymorphic form, the crystallization process should be monitored in real time. Crystalline product properties affect the physical, chemical and biological behaviour of the product. In-situ ATR-FTIR spectroscopy has been used for solute concentration measurement in crystallization processes. The onset of the crystallization, i.e., the moment when the first crystals are introduced into the system by nucleation, should also be monitored to enable the predictive control of the crystallization process. In addition, if the nascent polymorph could be predicted before any nuclei have been formed, the process could be driven to the direction of desired polymorphic form of the product. In this study, the application of dynamic PCA-based MSPC charts is introduced for monitoring the onset of crystallization process, primary nucleation from in-situ ATR-FTIR data gathered from batch cooling crystallization process. The objective was to set up the alarm for approaching nucleation for crystallization in order to control the process. T2 and Q charts are used to build up the alarm of approaching nucleation. T2 and Q contributions are applied to visualize the wave number ranges in the measured spectrum that are sensitive for the approaching nucleation. Further, when different polymorphs were formed, differences in collected spectra could also be detected. The developed tool enhances the control of batch cooling crystallization process since the new kind of real time information on the onset of the crystallization process can be obtained using the method described.