# Experimental and Numerical Study to Enhance Granule Control and Quality Predictions in Pharmaceutical Granulations

## TL;DR

This study uses a simulation model and AI to better understand and control the quality of granules in drug manufacturing.

## Contribution

A novel discrete element method model with AI integration is developed to predict and control granule quality in pharmaceutical processes.

## Key findings

- The DEM model accurately simulated granule growth kinetics matching experimental results.
- The EDEMpy AI tool effectively analyzed agglomerate size distributions for process efficiency.
- The model can predict granule quality based on equipment geometry and operational conditions.

## Abstract

Background/Objectives: The pharmaceutical industry demands stringent regulation of product characteristics and strives to ensure the reproducibility of granules manufactured via the wet granulation process. A systematic model employing the discrete element method (DEM) was developed herein to gain insights into and better control this process. Methods: The model comprehensively simulates particle behavior during granulation by considering the intrinsic properties of the powder material, the specific geometry of the granulation equipment, and various operational conditions, including impeller speed and chopper use. Notably, this approach can simulate dynamic interactions among particles and integrate complex phenomena, such as cohesion, which is crucial for predicting the formation and quality of granules. Results: To further support process optimization, an EDEMpy artificial intelligence (AI) tool was developed as a posttreatment routine to monitor and analyze agglomerate size distributions, proving essential for assessing the efficiency of the granulation process and the quality of resulting granules. The DEM model was evaluated by comparing its output with experimental data collected from a 0.5 L high-shear granulator. The model reproduced the granule growth kinetics observed experimentally, confirming the agreement between the experimental and numerical analyses. Conclusions: This underscores the model’s potential in predicting and controlling granule quality in wet granulation processes, enhancing the precision and efficiency of pharmaceutical manufacturing.