(85e) A Combined Experimental and Computational Approach to the Scale-up of High-Shear Wet Granulation

High-shear wet granulation is widely used in the pharmaceutical industry to achieve particle size enlargement, which leads to improved properties such as flowability and bulk density. Scale-up of wet granulation is always a challenge due to a wide range of geometrical designs of granulators and the varying amounts of shear rates in them, which in turn dictates the dynamics of granule formation and granule characteristics.

The objective of the current study is to find the smallest appropriate scale for conducting formulation and process optimization studies that can then be considered scalable. The ability to conduct key experiments at a small scale can lead to significant reduction in time and resources. As a first step, a full factorial design of experiments was conducted on three key process parameters to gain understanding of their effects on granule characteristics at 1-L scale. There were four center-point repeats and each parameter (water amount, impeller speed, wet massing time) was studied at two levels excluding the centers. The formulation used consisted primarily of microcrystalline cellulose, lactose, hydroxypropyl cellulose (added dry), and croscarmellose sodium. The main responses studied included granule porosity (mercury intrusion), granule particle size (sieve analysis), flow (Erweka® flow tester), bulk density, compactability (Stylcam® tablet press simulator), and dissolution. It was found that granule particle size was relatively insensitive to changes in process conditions (other than water amount), but those changes were well-captured by granule porosity and compactability. Statistical analysis was conducted on the responses and an optimized 1-L process was selected based on granulation flow and compactability.

Scale-up experiments were conducted using different impeller speed power laws, such as constant tip speed, constant Froude number, and constant shear stress, as scale-up rules. Three sizes of granulators were used for this study: 1-L, 10-L and 65-L. Water amount and water addition time was kept constant. Scale-up criteria included achieving similar granule characteristics (porosity and compactability) and a similar tablet dissolution profile across scales. It was found that in addition to impeller speed, wet massing time played a key role when comparing granule characteristics (especially pore size distribution) between scales.

A multi-dimensional population balance model that incorporates kernels based on fundamental physics and chemistry has been developed. The model will be validated and thereafter used to predict evolutions and distributions of granule particle size and porosity, thus alleviating the need for potentially labor and capital intensive experiments.