This study compares the use of wet milling and indirect ultrasound for promoting nucleation and controlling the particle size during the continuous crystallization of a hard-to-nucleate active pharmaceutical ingredient (API). Both an immersion and an external wet mill installed on a recirculation loop were investigated. It was found that all methodologies significantly improved the nucleation kinetics, and the effects of key process parameters (e.g., mill speed, temperature, and ultrasound intensity) on particle size were experimentally investigated. A minimum d 50 of 27 and 36.8 μm was achieved when using the wet mill and ultrasound, respectively. The effectiveness of wet milling was demonstrated in a three-stage mixed suspension mixed product removal continuous crystallization of the API that was operated continuously for 12 h (eight residence times), achieving a steady state with minimal fouling. Strategies for improving the overall robustness of the setup in routine manufacturing are discussed.

Precompetitive collaborations on new enabling technologies for research and development are becoming popular among pharmaceutical companies. The Enabling Technologies Consortium (ETC), a precompetitive collaboration of leading innovative pharmaceutical companies, identifies and executes projects, often with third-party collaborators, to develop new tools and technologies of mutual interest. Here, we report the results of one of the first ETC projects: the development of a user-friendly population balance model (PBM)-based crystallization simulator software. This project required the development of PBM software with integrated experimental data handling, kinetic parameter regression, interactive process simulation, visualization, and optimization capabilities incorporated in a computationally efficient and robust software platform. Inputs from a team of experienced scientists at 10 ETC member companies helped define a set of software features that guided a team of crystallization modelers to develop software incorporating these features. Communication, continuous testing, and feedback between the ETC and the academic team facilitated the software development. The product of this project, a software tool called CrySiV, an acronym for Crystallization Simulation and Visualization, is reported herein. Currently, CrySiV can be used for cooling, antisolvent, and combined cooling and antisolvent crystallization processes, with primary and secondary nucleation, growth, dissolution, agglomeration, and breakage of crystals. This paper describes the features and the numerical methods of the software and presents two case studies demonstrating its use for parameter estimation. In the first case study, a simulated data set is used to demonstrate the capabilities of the software to find kinetic parameters and its goodness of fit to a known solution. In the second case study, the kinetics of an antisolvent crystallization of indomethacin from a ternary solvent system are estimated, providing a practical example of the tool.

Scaling up and technology transfer of crystallization processes have been and continue to be a challenge. This is often due to the stochastic nature of primary nucleation, various scale dependencies of nucleation mechanisms, and the multitude of scale-up approaches. To better understand these dependencies, a series of isothermal induction time studies were performed across a range of vessel volumes, impeller types, and impeller speeds. From these measurements, the nucleation rate and growth time were estimated as parameters of an induction time distribution model. Then using machine learning techniques, correlations between the vessel hydrodynamic features, calculated from computational flow dynamic simulations, and nucleation kinetic parameters were analyzed. Of the 18 machine learning models trained, two models for the nucleation rate were found to have the best performance (in terms of % of predictions within experimental variance): a nonlinear random Forest model and a nonlinear gradient boosting model. For growth time, a nonlinear gradient boosting model was found to outperform the other models tested. These models were then ensembled to directly predict the probability of nucleation, at a given time, solely from hydrodynamic features with an overall root mean square error of 0.16. This work shows how machine learning approaches can be used to analyze limited datasets of induction times to provide insights into what hydrodynamic parameters should be considered in the scale-up of an unseeded crystallization process.

No abstract

Extracting quantitative information about highly scattering surfaces from an imaging system is challenging because the phase of the scattered light undergoes multiple folds upon propagation, resulting in complex speckle patterns. One specific application is the drying of wet powders in the pharmaceutical industry, where quantifying the particle size distribution (PSD) is of particular interest. A non-invasive and real-time monitoring probe in the drying process is required, but there is no suitable candidate for this purpose. In this report, we develop a theoretical relationship from the PSD to the speckle image and describe a physics-enhanced autocorrelation-based estimator (PEACE) machine learning algorithm for speckle analysis to measure the PSD of a powder surface. This method solves both the forward and inverse problems together and enjoys increased interpretability, since the machine learning approximator is regularized by the physical law.