(356e) Spherical Agglomeration Process: Mechanistic Understanding and Mathematical Modelling
AIChE Annual Meeting
2019
2019 AIChE Annual Meeting
Particle Technology Forum
Population Balance Modeling for Particle Formation Processes: Nucleation, Aggregation and Breakage Kernels
Tuesday, November 12, 2019 - 2:10pm to 2:35pm
The focus of the current work is on the agglomeration in suspension rate processes as there is no comprehensive mechanistic understanding of this process. Based on the similarities between wet granulation and spherical agglomeration, the following rate processes may be proposed: wetting/nucleation of primary crystals by the bridging liquid, consolidation/growth of agglomerate nuclei, and attrition/breakage of the agglomerates. The aim of the current study is to identify the influence of key material properties and process parameters on the final properties of agglomerates. This was achieved by a) integration of different targeted, systematic small-scale experiments designed to observe and quantify agglomerate evolution, and b) development of predictive and robust mathematical models to inform future process design tools through a bottom-up approach.
Spherical agglomeration experiments were conducted in a stirred vessel and an oscillatory baffled reactor. A novel microfluidic system and a contracting nozzle device were proposed and devised to respectively investigate wetting/nucleation and attrition/breakage of the agglomerates in isolation. A completely novel mathematical model is developed to investigate the kinetics of wetting and nucleation (via an immersion mechanism) and to identify different regimes of the nucleation process. gPROMS FormulatedProducts® was used to develop a population balance model for the agglomeration in suspension process, including a customised agglomeration kernel. The effect of material properties and process conditions on different rate processes, as well as size distribution of the produced agglomerates, was determined through the experiments and the developed models.