(668b) Successes and Challenges in Constructing a Crystallization Model for Prediction of Supersaturation and Crystal Size From Initial Seed Distributions

Control of powder properties is often a critical step in API processes. The ability to control crystal characteristics such as form, shape and size distribution is often required in order to formulate the drug product to within set specifications. Manipulation of parameters during the crystallization process such as seeding, residence time and antisolvent addition protocol can often be used to achieve desired results. In order to gain more insight on the intrinsic mechanisms governing our crystallization and to help define the design space for our crystallization parameters, a modeling effort based on supersaturation and crystal growth kinetics was developed. We have used as a basis the DynoChem size distribution model to formulate a new approach to modeling both the transient concentration and crystal growth of our system. Fitting the population balance to FBRM data was unsuccessful for this system most likely due to agglomeration inherent to the subject API. Alternatively the method of moments was used to solve the population balance based on an initial particle size distribution of the seeds. With the initial boundary condition set, kinetic parameters for nucleation and growth were fit to concentration data in order to obtain supersaturation profiles as well as particle size predictions. The utilities gained from this model enable the user to design a growth-dominated crystallization which will help deliver desired results on both lab and plant scales.