Granule Moisture Content Prediction During Fluid Bed Drying.

Area of Science:

• Pharmaceutical Manufacturing
• Process Engineering
• Analytical Chemistry

Background:

• Manufacturing process duration impacts pharmaceutical product quality.
• Accurate endpoint determination is crucial for consistent granule quality.
• Traditional methods include time- and property-defined endpoints.

Purpose of the Study:

• To evaluate humidity-temperature probes and near-infrared (NIR) spectroscopy for monitoring granule moisture content.
• To compare the accuracy of these methods for endpoint determination in a commercial-scale fluidized bed dryer (FBD).
• To identify a practical and precise solution for real-time moisture monitoring during pharmaceutical drying.

Main Methods:

• Experiments conducted on a commercial-scale fluidized bed dryer (FBD) processing 250 kg of wet granules.
• Utilized humidity-temperature sensor probes at inlet/outlet air ducts for continuous data collection.
• Employed near-infrared (NIR) spectroscope with a fiber optic probe for manual spectral data acquisition.
• Developed prediction models for moisture content (MC) using data from both methods.
• Evaluated model accuracy using the root mean square error (RMSE).

Main Results:

• The humidity-temperature probe model demonstrated higher predictive accuracy for MC with an RMSE of 0.11.
• The NIR spectroscopy model achieved a lower predictive accuracy with an RMSE of 0.29.
• Sensor-based monitoring proved effective for real-time MC tracking throughout the drying process.

Conclusions:

• Humidity-temperature probes provide a precise and practical solution for endpoint determination in commercial-scale FBD operations.
• Sensor-based moisture content monitoring offers superior accuracy compared to NIR spectroscopy for this application.
• This method enhances control over the drying process, potentially improving pharmaceutical product quality.