(189l) Filtration Studies Combined with Mechanistic Modelling to Reliable API Process Understanding and Scale-up

Process development and manufacturing in Pharma Industry comprises several unit operations that need to be studied and optimized along a project lifecycle. One of the main stages present in all manufacturing processes is the isolation of the product or side-products in solid form. Among the different solid-liquid separation techniques that can be applicable, one of the most common is the filtration of the product suspension. This unit operation is normally carried out easily at laboratorial or small scale but often brings concerns when increasing scale, or changing equipment, mostly due to lack of process understanding. The combination studies of scale dependent and scale independent parameters allied with mechanistic modelling can bring real gains to the process, including optimization of process conditions, time saving and risk mitigation associated with scale-up.

Unlike the empirical or statistical models, the mechanistic model of a filtration relies on data from specific experiments related with the mathematical description of the physical processes during filtration. Thus, the model provides accurate answers to important questions raised when developing and scaling-up a filtration, namely the cake resistance and the compressibility index, and allows a deeper knowledge of the process.

Our methodology produces experimental data that will support the filtration development and the mechanistic model using DynoChem. Filtration data gathered from previous large-scale runs was used to validate the mechanistic model by direct comparison of key process variables.

We demonstrate the application of this methodology by showing case studies where specific experiments were performed to estimate the filtration scale-independent parameters and scale-up of the process. The examples will show the evolution of filtration processes with high cake resistance during development and the consequent results for the large scale runs, and the using of the models to support the scale-up of filtration processes with accurate predictions and consulting on the best strategies to optimize cycle time and reduce process risk.