Conventional batch blending processes in pharmaceutical manufacturing processes coupled with long quality analysis times increases the cycle time of secondary pharmaceutical manufacturing processes. Continuous dry powder blending offers the benefit of continuous monitoring that can be used in optimizing the process for the right quality of the product at the end of the process. In order to understand continuous dry powder blending processes, the microscopic particle level properties must be related to the macroscopic process variables.

A continuous double helical ribbon blender was used to perform two component blending. Two different active ingredients (Caffeine and Acetaminophen) and two different excipients (Lactose and Microcrystalline Cellulose) were studied. Within Lactose, two different grades of lactose were studied. The change in macroscopic process variables like flow rate, rotation rate of the ribbon, angle of incline were all found to influence the blending process by affecting the fill weight and the mean residence time of the powder in the blender. Microscopic properties like particle size, shape, adhesion/cohesion were found to affect the dispersion coefficient and mean residence time of the powder in the blender. Near Infrared Spectroscopy was used to measure the concentration of the powder near the exit in real time.

Residence time distribution data were modeled using an axial dispersion model were found to describe the experimental data. A model for predicting the variance reduction ratio (ratio of variance at the input to that at the output) from residence time distribution was developed and tested. The measurement errors arising from the residence time experiments were found to propagate in the predictions which were found to be in conformance with experimentally determined variance reduction ratios within the range of measurement errors.