Treffer: NIR-Based Real-Time Monitoring of Freeze-Drying Processes: Application to Fault and Endpoint Detection.

In the pharmaceutical industry, freeze-drying is crucial for the stability of active pharmaceutical ingredients (APIs). Monitoring this complex process presents challenges as traditional methods often lack real-time insights, potentially leading to quality issues and batch rejections. Effective monitoring is then essential for optimizing process parameters and minimizing waste, thus saving costs and resources. This study evaluated the application of Near-Infrared (NIR) spectroscopy for the real-time monitoring of the freeze-drying process: NIR spectra were acquired in-line via a specially designed flange in the freeze-dryer. Two approaches were investigated. The first involved freeze-drying monitoring using control charts, thus creating a reference model based on cycles under normal conditions. A PCA model was developed using these reference cycles, and an intentional fault cycle was performed to test the system's ability to detect deviations. Multivariate control charts, utilizing Hotelling's T² and DModX metrics, were shown to effectively monitor process deviations, enhancing the understanding of freeze-drying dynamics. The second approach was based on the use of NIR spectroscopy for assessing residual moisture during lyophilization. By integrating Partial Least Squares (PLS) regression with in-line NIR spectra, we estimated endpoints and detected faults in both reference and faulty cycles. The results showed strong correlations between PLS estimates and the Pirani–Baratron method, highlighting NIR's applicability for monitoring drying phases. This research advocates for broader NIR implementation in pharmaceutical development, emphasizing its potential to monitor the process, ensure quality, and reduce out-of-specification product. [ABSTRACT FROM AUTHOR]

Copyright of Processes is the property of MDPI and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)