NIR spectroscopy for the in-line monitoring of a multicomponent formulation during the entire freeze-drying process

•NIR spectroscopy was able to extract product information in an icy environment during freeze-drying.•In-line NIR monitoring of all components during an entire freeze drying process.•Comparison of chemometric techniques for data-analysis.

Freeze drying is a complex, time consuming and thus expensive process, hence creating a need for understanding the material behaviour in the process environment and for process optimization. Near-infrared (NIR) spectroscopy offers the opportunity to monitor physicochemical changes of the formulation during freeze-drying. The aim of this work was to examine whether NIR spectroscopy allows in-line monitoring of all components during the entire freeze-drying process of a multi-component pharmaceutical formulation (a solution of fenofibrate and mannitol in a mixture of tertiary-butyl alcohol, and water). To extract useful information of all components in the formulation from the large multivariate data-sets obtained during in-line spectroscopic monitoring, several spectral pre-processing techniques and spectral data analysis techniques such as the mean of selected wavenumbers (Mws), the correlation coefficient (CorrCoef) and principal component analysis (PCA) have been evaluated and compared. To find out whether these chemometric techniques are also able to differentiate between changes in the process settings influencing the freeze-drying process of the formulation, freeze-drying processes were performed at four different conditions. Results demonstrated that in-line measurements using NIR spectroscopy were possible in an icy environment and that a further process understanding could be obtained. Data-analysis revealed the crystallization behaviour of each of the four components. In addition, using the three pre-processing techniques allowed observe the sublimation of the solvents. Mws and CorrCoef have proven to be adequate methods for monitoring the main physicochemical changes of product during the processes; this affirmation was confirmed by observing the outputs of PCA for entire processes.

Graphical abstractFigure optionsDownload full-size imageDownload as PowerPoint slide