Understanding of Wetting Mechanism Toward the Sticky Powder and Machine Learning in Predicting Granule Size...

Area of Science:

• Pharmaceutical Technology
• Materials Science
• Chemical Engineering

Background:

• High shear wet granulation (HSWG) is crucial for traditional Chinese medicine (TCM) powder processing.
• Understanding binder wetting mechanisms and their impact on granule size distribution (GSD) in sticky TCM powders is limited.
• Developing predictive models for GSD is essential for intelligent manufacturing of TCM.

Purpose of the Study:

• To investigate the wetting mechanism of binders in HSWG of TCM powders.
• To explore the influence of various parameters on GSD during HSWG.
• To establish a predictive model for GSD in TCM granulation.

Main Methods:

• Utilized permeability and contact angle experiments with molecular dynamics (MD) simulations to study wetting mechanisms.
• Employed machine learning (ML), specifically Extreme Gradient Boosting (XGBoost), for GSD prediction.
• Performed correlation analysis to assess parameter influences on GSD.

Main Results:

• Water increased powder viscosity, while ethanol acted as a wetting agent, reducing contact angle.
• XGBoost demonstrated superior accuracy in GSD prediction.
• Binder concentration and amount significantly impacted GSD, whereas impeller speed had minimal influence.

Conclusions:

• Elucidated the distinct roles of water and ethanol in TCM powder wetting during granulation.
• Developed an accurate ML-based GSD prediction model for HSWG of TCM.
• Provided insights into controlling GSD by optimizing formulation and process parameters for TCM intelligent manufacturing.