(653f) Modeling Continuous Film Coating Process in Consigma Coater: What Changes?

Tablets are the most commonly used oral dosage form in the pharmaceutical industry. They are often film coated, either to add a protective layer, to mask taste, for cosmetic reasons or to change the release and disintegration of the tablet. The desired effect is heavily dependent on the coating quality (coating thickness and intra and inter tablet coating variability). Predicting these attributes is critical to allow faster and material sparing development. Modeling film coating allows an easier scale up from laboratory scale trials to commercial scale.

We present a model of film coating for the ConsiGma semi-continuous coater. The ConsiGma film coater was developed to be integrated in continuous tableting lines, as a fast mini batch coater. This requires an increased throughput compared to other similar batch size coaters. For this, ConsiGma uses a different approach to distributing the tablets along the drum, making the scale-up from a lab coater (or any other coater) to ConsiGma more complex.

The main challenge in modeling coating is predicting the size of droplet that impacts the tablet, which will provide the weight gain not only of coating material but of water as well. This requires modeling how much solvent evaporated since the droplet left the atomization nozzle. Once the droplet reaches the tablet, it’s also important to understand what happens to the solvent once it reaches the tablet, since it may be absorbed or it may evaporate.

ConsiGma coater brings new challenges, the high throughput makes the assumption of a stationary process impossible. Significant temperature variations occur during the actual coating, from the initial heated state without feed flow to the final cooler state after coating. The model presented achieved an RMSE in predicting the inlet gas temperature of less than 3ºC during the entire process. Predicting the movement of the tablets in ConsiGma is also more challenging, due to the new tablet cascade dynamics. The tablet movement is critical to know the actual distance between the nozzle and the tablets, which will change the droplet size impacting the tablets.

Finally, all these estimated parameters need to be transformed into distributions, not all tablets follow the same path, not all droplets are the same size. This allows the estimation of more complex process quality attributes, such as coating efficiency.

With this new model, we can predict not only the coating quality but also the time required to achieve a desired tablet weight increase.