PT  - JOURNAL ARTICLE
AU  - Azheruddin Mohammed
AU  - Antoine Cournoyer
AU  - Ryan Gosselin
TI  - An Investigation into the Spatial Distribution of Moisture in Freeze-Dried Products Using NIR Spectroscopy and Chemical Imaging
AID  - 10.5731/pdajpst.2020.012443
DP  - 2023 Mar 01
TA  - PDA Journal of Pharmaceutical Science and Technology
PG  - 55--66
VI  - 77
IP  - 2
4099  - http://journal.pda.org/content/77/2/55.short
4100  - http://journal.pda.org/content/77/2/55.full
SO  - PDA J Pharm Sci Technol2023 Mar 01; 77
AB  - Near-infrared (NIR) spectroscopy (NIRS) is a widely accepted method of measuring moisture in pharmaceutical freeze-dried products, both during the process and in the finished products. Multiple NIR measurement approaches have been introduced to monitor product moisture in freeze-dried vials. However, the spatial moisture gradients within a vial have not been investigated in depth. Like any other point-focused process analytical technology (PAT) tool, a spectrum produced by NIRS represents an average over a given area of the product vial. Implementing a point-focused NIR on any random position without proper understanding of spatial moisture variations within the vial may severely impact the reliability of the results. The present work focuses on understanding the moisture distribution within freeze-dried vials. We performed an investigation using NIR tools, NIR chemical imaging (NIR-CI), and NIRS to understand the spatial variations in moisture on the outer surface (i.e., periphery) of the freeze-dried vials. To achieve this, the moisture distribution within individual vials was mapped using NIR images. Then, NIRS was used to determine the necessity of using multiple measurement points to produce robust models quantifying the moisture inside freeze-dried products. Overall, the results show a simplified version of the phenomenon in which non-homogenous distribution of moisture, as well as the non-uniform drying front, occur within the vials. The findings from the NIRS-based partial least squares (PLS) models indicate that to achieve reliable product/process information, measurements must be drawn from multiple measurement points on the surface of the freeze-dried products.