Development of a spatially offset Raman spectroscopy probe for monitoring pharmaceutical drying

Al-Attili, Mais and Ferreira, Carla and Price, Chris and Faulds, Karen and Chen, Yi-Chieh (2023) Development of a spatially offset Raman spectroscopy probe for monitoring pharmaceutical drying. Chemical Engineering Research and Design, 192. pp. 510-520. ISSN 0263-8762 (https://doi.org/10.1016/j.cherd.2023.02.041)

[thumbnail of Al-Attili-etal-CERD-2023-spatially-offset-Raman-spectroscopy-probe-for-monitoring-pharmaceutical-drying]
Preview
Text. Filename: Al_Attili_etal_CERD_2023_spatially_offset_Raman_spectroscopy_probe_for_monitoring_pharmaceutical_drying.pdf
Final Published Version
License: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 logo

Download (5MB)| Preview

Abstract

Spatially offset Raman spectroscopy (SORS) is a subset of Raman spectroscopy devised for probing subsurface compositions in non-homogenous media. An example of such media is the wet filter cake during pharmaceutical drying. This non-homogeneity poses a challenge for process monitoring as it could render the determined solvent content during drying, and the end point, inaccurate. In this study, a SORS probe was developed for the monitoring of pharmaceutical drying. The probe includes a 45˚ illumination point and 0-5 mm equidistant collection offsets. Using the SORS probe, solvent signal detection through variable thicknesses of dry paracetamol was examined. The solvent content during the drying of paracetamol in anisole was then monitored using the SORS probe. Partial least squares regression (PLSR) analysis was applied to evaluate the performance of offset configurations in monitoring the solvent content during drying. The results showed that the solvent signal is detected through thicknesses beyond 6 mm of paracetamol from the larger offsets of 4-5 mm. A more accurate prediction of the solvent content was obtained from larger offsets. PLSR models using offset spectra showed a decrease in estimation error up to 50% compared to backscattering spectra, with a further decrease upon using standard normal variate pre-processing. This suggests that SORS could offer improved monitoring of pharmaceutical drying processes.