In the pharmaceutical industry, antisolvent crystallization has been applied in the synthetic process of active pharmaceutical ingredients owing to its moderate thermal requirement and capability to produce high yields. Various attempts have been taken to understand the cooling crystallization and antisolvent crystallization processes to advance their application in the pharmaceutical industries. Among the physical properties of the active pharmaceutical ingredients, simulation of the crystal size before and after large-scale experiments has become an area of interest due to its importance in the industry.
Presently, crystallization modeling based on the nucleation and crystal growth is among the available methods for crystal size simulation. However, in a pharmaceutical industry, crystal sizes are measured by the laser diffraction/ scattering method due to its wide dynamic range and high reproducibility. Considering the characteristic of the laser diffraction method, a huge difference between the measured values and the calculated values obtained with crystallization models have been reported. Thus, an urgent solution is highly desirable to ensure effective transformation between the measured and calculated data.
Recently, Takanori Kodera from the Pharmaceutical Science and Technology Core Function Unit in collaboration with Dr. Masanori Kobari and Dr. Izumi Hirasawa from Waseda University developed a new antisolvent crystallization model to simulate the crystal size in pharmaceutical industries. The model was simple, concise and entailed only six experimentally determined kinetic parameters including the rate constants and orders of crystal growth, and primary and secondary nucleation. The work is currently published in the journal, Chemical Engineering & Technology.
A methodology for determining the growth rate constant and order was developed. It included an appropriate treatment of the size distribution data obtained through the laser scattering/ diffraction method. Eventually, the determined growth rates were used to simulate the crystal size where the feasibility of the simulation by crystallization modeling in the practical application of the pharmaceutical industry was confirmed. On the other hand, the growth rate parameters were first determined experimentally as they can be obtained independently based on the assumption that the nucleation is negligible. The assumption was proved experimentally using a seed chart.
From the seed chart, the authors confirmed that the nucleation was made negligible by sufficient amounts of seed crystals. The population density of the seed crystals from the raw data was equivalent diameter distribution of spheres in volume. As such, a number coefficient was introduced to adjust the optical and shape properties of the seed crystal to determine its population density. Additionally, since the proposed model was based on population and mass balances, the number mean diameter was useful in determining the crystal size.
Utilizing the obtained growth rate parameters, a good agreement was exhibited between the measured values and the calculated number mean diameters of the seed and product crystals. Furthermore, the parameters could be successfully used to simulate the crystal size in other solvent compositions. Reflecting on the study success, Kodera the first author in a statement to the Advances in Engineering community, is hopeful that the simple and concise approach presented will advance antisolvent crystallization in pharmaceutical industries.
Kodera, T., Kobari, M., & Hirasawa, I. (2019). Modeling and Growth Kinetics of Antisolvent Crystallization Applied to the Pharmaceutical Industry. Chemical Engineering & Technology, 42(7), 1458-1465.