Novel method for constructing a large-scale design space in lubrication process by using Bayesian estimation based on...

Area of Science:

• Pharmaceutical Science
• Chemical Engineering
• Process Chemistry

Background:

• Establishing a reliable large-scale design space is crucial for pharmaceutical development.
• Traditional large-scale design of experiments (DoE) can be resource-intensive and impractical.
• Scale-up rules are essential for bridging the gap between small-scale and large-scale manufacturing.

Purpose of the Study:

• To develop a robust method for constructing a reliable large-scale design space.
• To integrate scale-up rule reliability into Bayesian estimation for process optimization.
• To validate the approach using the lubricant blending process for theophylline tablets.

Main Methods:

• Conducted small-scale DoE varying Froude numbers and blending times.
• Utilized multivariate spline interpolation, bootstrap resampling, and self-organizing map clustering.
• Applied Bayesian estimation to correct small-scale response surfaces using large-scale data and scale-up rule reliability.

Main Results:

• Developed a reliable large-scale design space without extensive large-scale DoE.
• Demonstrated that the corrected design space is more reliable than the small-scale version.
• Showcased the effectiveness of integrating scale-up rule reliability into Bayesian estimation.

Conclusions:

• The proposed Bayesian estimation approach enhances the reliability of large-scale design spaces.
• This method is valuable for pharmaceutical development when large-scale DoE is not feasible.
• The integration of scale-up reliability improves pharmaceutical quality predictability across manufacturing scales.