Investigation of Nucleation under High-Shear Conditions for a Pharmaceutical Compound in an Unseeded System

Abstract

The effect of in situ high energy mixing on primary nucleation of an active pharmaceutical compound (ABT-089) has been investigated by measuring the induction time for 1, 5, and 140 mL scales via different shear-inducing mechanisms, namely, magnetic stirrer, Couette device, and high-shear rotor–stator mixer known as wet-mill in pharmaceutical industry. Experiments performed on previously untested scale and equipment show that the onset time of nucleation decreases with an increase in average energy input per unit volume of the solution (i.e., specific energy input) as previously reported (Liu, J.; Svard, M.; Rasmuson, Å.C. Influence of Agitation on Primary Nucleation in Stirred Tank Crystallizers, Cryst. Growth Des. 2015, 15, 4177–4184). The variability in the induction time is found to decrease with an increase in volume and specific energy input. Studies at different scales are compared using the Rasmuson (Liu, J.; Svard, M.; Rasmuson, ÅC. Influence of Agitation on Primary Nucleation in Stirred Tank Crystallizers, Cryst. Growth Des. 2015, 15, 4177–4184) approach based on the classical nucleation theory. It is shown that the effect of specific energy input can be adequately described by the kinetic term “A” or pre-exponential factor within the classical nucleation theory framework because the critical free-energy term appears to be constant for the same scale. A semiempirical correlation specific to ABT-089 is proposed between the energy input to a system and the average onset of primary nucleation for each of the means of shear induction. The generic approach presented in this work can be used as part of the overall development strategy to scale-up and scale-down crystallization experiments by estimating the specific energy input or the corresponding tip speed requirement to induce primary nucleation within a specified time range. However, given the practical limitations on the amount of data that can be generated and used to develop such correlations, appropriate caution is necessary in defining the scope of their applicability.