Impurity Profiling for a Scalable Continuous Synthesis and Crystallization of Carbamazepine Drug Substance

Abstract

A scalable continuous manufacturing process for the synthesis and crystallization of form III carbamazepine (CBZ) from iminostilbene (ISB) has been established. A high-yielding synthesis was first obtained using a plug flow reactor (PFR) and then scaled up using a continuous oscillatory baffled reactor (COBR). A real-time in-line Raman spectroscopy method was implemented to ensure that the conversion of the starting material ISB to the product CBZ was maintained above 99.0%. The monitored product stream was telescoped into a mixed-suspension mixed-product crystallizer (MSMPR-1) and a filtration unit to isolate the preliminary CBZ form I polymorph. A cooling recrystallization process was designed by using a crystal growth model derived from microscopy measurements. The impurity purging capacities and polymorph attainments were compared for the batch and flow processes. This study outlines the role of process modeling and process analytical technology (PAT) for impurity purging in a telescoped continuous manufacturing process.

Iterative Regression of Corrective Baselines (IRCB): A New Model for Quantitative Spectroscopy

“…Raman and infrared spectroscopies have been applied for an increasing number of use cases. Among others, these applications of spectroscopy include food, − pharmaceuticals, − cosmetics, tobacco, and nuclear waste. , In 2004, the US Food and Drug Administration (FDA) and the International Council for Harmonization (ICH) established an initiative to apply process analytical technologies (PAT), including spectroscopic PAC, for manufacturing quality assurance. , Many studies have reported on the use of spectroscopic analyzers, such as infrared (IR) and Raman, to monitor various stages of pharmaceutical manufacturing. ,,− Spectroscopic PAT is useful both for real-time release and model predictive control. ,,, The linking of PAT to continuous manufacturing for real-time optimization and control using artificial intelligence was referred to by Price et. al as the “holy grail” …”