(185b) Model-Based Analysis of Breakage in Fluid Bed Drying of Continuously Produced Pharmaceutical Wet Granules

In continuous pharmaceutical wet granulation, subsequent drying is a necessary step for evaporating the granulation liquid from the granules and rendering these processable downstream and in tableting. In the ConsiGmaâ„¢-25 wet granulation this is achieved through semicontinuous fluid bed drying. The desired effect of drying to the ideal moisture content is accompanied by breakage of granules during this process (De Leersnyder et al., 2018). The generated fines have a negative impact on the processability downstream, such as possibly reduced content uniformity and flowability, as well as to the drying stage itself where the fine material gets stuck on the filter that entrains particulate material in the outgoing air. The latter could affect the duration limit of continuous operation as the pressure across the filters increases due to the fouling.

Not only does this breakage cause lower processability, it is also complex to capture and predict. It namely occurs in different locations in the process, within each of a different nature. Wet granules exiting the granulator are sent to the dryer through gravitational loading in a vertical ConsiGmaâ„¢ setup, whereas in horizontal configurations these are pneumatically conveyed through a tube. Through contact with the tube wall granules undergo size reduction, dependent on their current properties governed by composition, size and moisture content. The liquid moreover provides the granules with different elastic properties than in a dry state. The next stage is during the fluid bed drying itself, as granules are dropped into the fluidizing air from the top of the dryer and collide with the dryer cell walls or already present granules. This breakage extent is affected by dryer process settings and granule properties. Finally, often the largest size reduction of the product occurs in the discharge of the dried granules to the subsequent evaluation module. A more forceful pneumatic conveying transports the granules through another tube, termed dry transfer line, causing breakage in function of granule properties, most prominently remaining moisture content.

As this problem is too complex to forecast, mathematical models can be used to predict the extent and nature of breakage occurring due to location, material and granule properties. A change in particle size distribution (PSD) can be modeled by means of a population balance model (PBM). Breakage models can be implemented in this framework, with kernels describing (combinations of) different fragmentation mechanisms that are plausible for the system at hand. The extent of breakage and other parameters of these functions can be mathematically linked to the modeled materials’ properties by using PSD data.
The quality of such developed models is however not only dependent on the supplied data, also applying the right objective function plays a key role in finding the model and mechanism that describes the system in the most robust way. PSD data contains an abundance of information, it is therefore difficult to find an objective function that can take all of it into account. Also the experimental variability of PSD results could be used in model calibration in order to emphasize the most certain experimental information in the distributions. This work therefore implements techniques from the compositional data field to calculate variances on experimental PSD outcomes in function of size class. The (inverse of) the measurement uncertainty is then taken into account in the objective function. It has been shown that this allows to include information in both number- and volume-based PSD for mechanistic purposes (Ghijs, 2020).

Recent data collections have been carried out on the ConsiGma™-25 that varied the active pharmaceutical ingredient (API) and its powder properties with a fixed combination of excipients of 30% lactose monohydrate (Pharmatose 200M, DFE Pharma, Veghel, The Netherlands), 15% MCC (Avicel® PH-101, FMC International Health and Nutrition Cork, Ireland) and 5% HPMC (Methocel™ E15, The Dow Chemical Company, Midland, U.S.A.). The applied APIs showed different degrees of hydrophilicity and solubility. For each API, granules were produced at different levels of liquid addition. Part of the study also involved producing granules of similar properties based on different APIs, in order to isolate the effect of the applied API. Drying experiments were then carried varying the dryer inlet air temperature, mass loading and drying time. The experimental outcomes were characterized by high resolution PSD measurements (dynamic image analysis, QICPIC, Sympatec Gmbh, Clausthal-Zellerfeld, Germany) of the incoming and outgoing granules of the dryer.

In combination with the breakage PBM including measurement uncertainty in model calibration and validation, this study elucidates the effect of the tested API properties, granule properties (granule size and moisture content) and dryer process settings on their breakage behavior through the ConsiGmaâ„¢-25 system. The more robust calibration helps in picking out the most appropriate breakage kinetics out of the abundance of PSD data. Moreover, a predictive model is generated that can take process settings, material and granule properties into account in forecasting breakage over the dryer for the selected formulations. This constitutes a pivotal modelling step towards a formulation-generic breakage model.

References

De Leersnyder, F., Vanhoorne, V., Bekaert, H., Vercruysse, J., Ghijs, M., Bostijn, N., Verstraeten, M., Cappuyns, P., Van Assche, I., Vander Heyden, Y., Ziemons, E., Remon, J. P., Nopens, I., Vervaet, C., & De Beer, T. (2018). Breakage and drying behaviour of granules in a continuous fluid bed dryer: Influence of process parameters and wet granule transfer. European Journal of Pharmaceutical Sciences, 115, 223–232. https://doi.org/10.1016/j.ejps.2018.01.037

Ghijs, M. (2020). Modelling and experimentation on fluid bed drying of granules in the context of pharmaceutical continuous wet granulation. Universiteit Gent. Faculteit Bio-ingenieurswetenschappen.