Understanding the fluidization of pharmaceutical powders in simple canonical flow channels

The thesis aims to get a better understanding of the fundamentals of pharmaceutical powder fluidization that was not widely addressed in the literature by investigating the fluidization process from the early stage of powder particles evacuation from the powder bed till dispersion in simple geometries. Such fundamental studies are essential to understand the fluidization process in dry powder inhalers (DPIs) and improve DPI device design and development.

Firstly, an experimental platform was developed using a two-beam laser attenuation method to study the effects of powder properties, turbulence, and forces induced by airflow. Horizontal flow channels of a square cross-sectional area with a side length of 5 mm and 10 mm were used, and the powder bed was subjected to a shear force (powder bed surface was in line with the airflow direction). Airflow with Reynolds numbers of 9000 to 20,000 was utilized to fluidize lactose pharmaceutical powder used as a carrier for the drug particles in dry powder inhalers formulation. Four lactose powders with different median diameters and cohesiveness properties were examined. The results demonstrated that the time taken by the powder to evacuate from the powder bed is associated with the Reynolds number except for high Reynolds number cases in which the increase in the airflow rate only had a slight effect on the evacuation time. The properties of the powder dominated the differences in evacuation time for the low Reynold number cases. The laser voltage signal after the beam passes through the powder pocket was fitted with a polynomial fit. The frequency associated with the evacuation process was calculated by taking the Fourier transform of the subtracted voltage signal (calculated by subtracting the laser voltage signal from the fitted curve). The change in the frequencies represents the temporal variation in the laser light intensity associated with the powder evacuation process, corresponded well with the change in the theoretical frequencies range. Theoretical frequencies were calculated based on turbulent velocity and the characteristic length scale of the powder pocket.

Secondly, the effect of simple design alteration on enhancing the fluidization process effectiveness by placing a grid upstream of the powder bed in the horizontal channel was investigated. The impact of flow turbulence induced by the grids on powder fluidization was examined using direct high-speed imaging and particle image velocimetry (PIV). Flow turbulence is one of the dispersion mechanisms adopted in many dry powder inhaler devices. Direct high-speed imaging provided qualitative information about powder bed morphology and quantitative information such as powder evacuating rates and evacuation percentage time history. PIV was incorporated to obtain the powder particle flow field around the pocket. The effects of grid blockage ratio (ranging from ~25% to ~40%) and grid location relative to the powder pocket on the powder fluidization were investigated. The use of grids reduced the time required for the powder evacuation and residual remaining in the pocket and enabled powder evacuation at a lower flow rate. Grid with a low blockage ratio achieved lower evacuation time and higher local turbulence intensity of the powder particles compared to the other tested grid for the examined flow rates.

Thirdly, the lactose particle's local dynamics after being dispersed from the powder pocket into the channel were studied using particle image velocimetry (PIV) and high-speed, long-distance microscopy (HS-LDM). Airflow with a mean velocity ranging between 13.3 m/s till 66.67 m/s were examined using the four lactose powders. The effects of grids with different blockage ratios upstream of the powder pocket on particle dynamics were investigated. Using grids improved the powder particle dispersion and increased the local powder particle velocities. The mean velocity for powder particles using HS-LDM corresponds well with the PIV measurement. Particles' mean velocity was affected by the percentage of fines and particles' median diameter. Higher mean velocity was observed for powders with higher fines percentages and smaller median diameters.

Finally, the lactose carrier's fluidization in a vertical flow channel where the powder bed was subjected to normal force (inlet airflow was normal to the powder interface) was investigated. Direct high-speed imaging was used to study the powder pocket morphology, while PIV and HS-LDM were used to study powder particle dynamics. Airflow rates with average velocities of ~1.7 m/s to ~7 m/s were examined. The vertical channel has an inlet and a discharge channel with a square cross-sectional area with a side length of 15 mm. Fracture of the powder bed was observed for the cohesive powder (LH200), in which the powder bed broke in the form of coarse plugs. Powder particles experienced higher turbulence intensities at the recirculation region at the discharge channel's left corner. A lower slip ratio for particles with different size bands was noticed in the region close to the powder bed. Particle distribution percentage for different size bands changed more with distance from powder pocket in the region located closer to the powder pocket compared to further downstream locations. In addition, changing the airflow rate for the same powder had a slight effect on the particle distribution percentage for different size bands.