(100e) Effect of a High-Shear Homogenizer On Blend Time in a Stirred Tank

Blend time defined as the time required to reach a predetermined degree of homogeneity in a mixing system, is a parameter commonly used to determine the effectiveness of a batch mixing process. Non-dimensional blend times for stirred tanks are available in the literature. However, little information is currently available for fully baffled, stirred tanks equipped with two mixing devices i.e., a centrally mounted slowly rotating impeller to generate bulk flow and a second high-shear homogenizer placed off-center to generate a high-shear zone. These systems are commonly used in the pharmaceutical industry to improve processes such as emulsification and dispersion, which are typically conducted in batch stirred tanks.

In this work, the blend time in a fully baffled, stirred tank equipped with a centrally mounted Rushton turbine as the primary impeller and an additional high-shear rotor-stator homogenizer, was experimentally determined for low-viscosity liquids using a fast acid-base reaction in the presence of an indicator together with an imaging analysis technique to monitor the change in color during the process. Blend time was also determined using conductivity measurements. In addition, the velocity distribution generated by the high-shear homogenizer was experimentally obtained using Particle Image Velocimetry (PIV). These data were used, through a mass balance, to quantify the pumping capacity of the homogenizer, which was compared to that of common impellers. The pumping number for the homogenizer was then calculated at different homogenizer rotational speeds.

Results show a significant reduction in the overall blend time when the high-shear homogenizer is present in the tank, as compared to the blend time obtained with the primary impeller alone. Data were collected at different homogenizer speed and primary impeller speeds, and fitted with a semi-empirical model with only one parameter. The model assumes that the tank is divided into two separated areas that exchange material based on the pumping capacity of the impellers. Data fitting to the model (requiring one fitted parameter to be determined separately and independently) shows a good agreement between the model prediction and the experimental results for blend time under different operating conditions.