Contribution of crystal-impeller and crystal-crystal collisions to secondary nucleation

A secondary nucleation model mainly based on the findings of Ottens (1973) and Evans (1974) considering both crystal-impeller collisions and crystal-crystal collisions was investigated using experimental data obtained from two crystallizers, a 22-liter Draft Tube (DT) crystallizer and an 1100-liter Draft Tube Baffled (DTB) crystallizer, which are both operated continuously in an evaporative mode for the crystallization of ammonium sulfate. Since the two crystallizer types differ in scale and configuration, not only the effect of the impeller frequency but also the effect of scale on the crystal size distribution (CSD) could be investigated. The CSD-prediction obtained using dynamic process simulations is consistent with the measured data for all investigated experiments. Not only the changes in the CSD due to different impeller frequencies, but also the changes due to scale and configuration are well described and model is able to capture the sustained cyclic behavior in DTB crystallizer. It was found that three model parameters i.e. the number of nuclei per unit energy, the lower bound of integration for the crystal-impeller and for the crystalcrystal collisions are a function of the impeller frequency and should therefore be altered accordingly. Therefore the predictive capability of this model is limited. In the 22-liter DT-crystallizer crystal-impeller collisions are dominating while crystalcrystal collisions are of less importance. Nevertheless, crystal-crystal collisions cannot be neglected and are especially pronounced at low impeller frequencies. In the 1100-liter DTB-crystallizer the term of the secondary nucleation rate equation describing crystal-crystal collisions is more important. But this effect cannot be attributed only to crystal-crystal collisions. Since the supersaturation in the DTBcrystallizer is higher, and the circulation time is larger compared to the DTcrystallizer, it is expected that also the surface breeding in combination with a fluid shear mechanism is responsible for the production of secondary nuclei.