(461e) A Combined Numerical Modelling and Laboratory Investigation into the Relationship between the Power to Mass Ratio, Turbulent Eddy Dissipation and Particle Size Distribution | AIChE

(461e) A Combined Numerical Modelling and Laboratory Investigation into the Relationship between the Power to Mass Ratio, Turbulent Eddy Dissipation and Particle Size Distribution

Authors 

O'Dwyer, K. - Presenter, APC - Dublin
O'Sullivan, J., APC Ltd
Crystallisation is of great important in the pharmaceutical, food, and fine chemical industries. In the pharmaceutical industry specifically, it is used to separate and purify intermediates and Active Pharmaceutical Ingredients (APIs). Changes to crystal morphology is well understood in terms of the chemical factors such as solubility, metastable zone control, and antisolvent addition to achieve optimisable crystals. However, crystals also change because of the hydrodynamics in a vessel. The relationship between fluid mechanics and crystal characteristics (mean crystal size, particle size distribution (PSD), morphology) are yet to be understood. In industry, the main challenge in a crystallisation process is to obtain the desired PSD of crystals. The PSD is an important factor to consider as it will impact the product quality and purity. It also affects the downstream processes such as filtration, drying, and milling. However, the hydrodynamics in a crystallisation process are not well understood. The gap in this understanding can reduce the flexibility in the production of pharmaceuticals.

Mechanical stress parameters (turbulence eddy dissipation (TED), strain rate, power draw) were estimated using computational fluid dynamics (CFD) to determine the effect on the particle size distribution (PSD) in a cooling crystallisation of benzoic acid. In CFD the distributions of the TED and strain rate were outputted, and maximum values found for different impellers (anchor impeller, pitched blade impeller, retreat curve impeller, half-moon impeller), baffled and unbaffled reactors and using two different power draws (0.15 W/kg and 0.01 W/kg). Experiments, to replicate the simulations in the laboratory, were then run.

The experimental PSD was found to be related to an increase of TED and strain rate (found in CFD simulations). The PSD increased 26 -73 % at the high-power draw power draw with the introduction of baffles. A combination of high-power draw and change in impeller type could decrease the PSD by 3-51 %. It was determined to achieve the smallest PSD the baffled half-moon impeller configuration must be used. The mechanical stress outputs (TED and strain rate) were the largest from the CFD simulations for the half-moon impeller. This suggests the increase in the mechanical stress led to increased attrition and a smaller PSD as a result. The retreat curve impeller produced the largest PSD, which was attributed to its poor mixing performance, leading to less attrition and overall poor energy transmission. A workflow from the variables (impeller, baffles and power draw) was developed to aid in the optimisation of the PSD using mechanical stress factors.

This integrated approach between mathematical models and experiments allowed for the rapid assessment of potential practical solutions on a laboratory scale unit operation before large scale pharmaceutical manufacturing runs. The TED rate, strain rate, and power draw have been deemed important characteristics that possibly cause attrition of crystals, reducing the PSD as they are increased. This offer researchers and manufacturing additional means to control the PSD of pharmaceutical products.