(380c) Optimization of Dissolution Time in a Stirred Tank: A Case Study Using the Population Balance Method

Dissolution of solid particles in liquids using stirred tanks is a common process in many pharmaceutical, consumer products, and chemical industry applications. Homogeneous suspension of particles is the first step towards maximizing the mass transfer from the particles to the liquid. Many process parameters affect the dissolution rate such as overall flow field, agitation speed, and the particle size distribution among others. Optimizing the operating conditions of the stirred tank for dissolution is needed for process design, e.g. scaleup from lab scale to production scale.

In this study, a CFD simulation methodology is developed for modeling particle dissolution using the Eulerian multiphase model combined with Population Balance Method. Two approaches of the Population Balance Method are compared in terms of speed and accuracy, namely the Discrete Method and Quadrature Method of Moments (QMOM).

Predicting the correct cloud height and solid particle suspension is the first step needed to accurately predict the mass transfer between the solid particles and the liquid. One main contributor towards the accuracy of the cloud height predictions is the turbulent flow field. Several studies iterated the importance of accounting for the turbulent flow impact on drag force acting on the suspended particles. Turbulent dispersion force is also shown to have significant impact on the cloud height prediction in stirred tanks. Various models for including additional liquid-solid interaction forces on particle distribution in the stirred tank are explored. In addition, the influence of various parameters such as agitation speed, particle size, solubility in liquid etc., on the particle dissolution time and dissolved species blend time is studied.

A baseline design will be first presented under given process conditions and its results will be presented. Then, once it is validated, a response surface-based optimization algorithm is implemented to study the effects of the inputs, e.g. particle size and solubility, in order to optimize the dissolution time. The response surface-based approach starts by creating a Design of Experiments (DoE) to optimally fill the design space with design (simulation) points. The design points will then be used to construct the response surface, which in turn can be efficiently used to find the optimum operating conditions with respect to the dissolution time.

The conclusion of this work will contain the lessons learned when it comes to modelling particle suspension, turbulent dispersion in solid suspension tanks, population balance, and finally the mass transfer (dissolution) between the solid particles to the liquid.