A quantitative comparison of in-line coating thickness distributions obtained from a pharmaceutical tablet mixing process using discrete element method and terahertz pulsed imaging

Jul 5, 2018

Pei, Chunlei, Hungyen Lin, Daniel Markl, Yao-Chun Shen, J. Axel Zeitler, and James A. Elliott.

“A quantitative comparison of in-line coating thickness distributions obtained from a pharmaceutical tablet mixing process using discrete element method and terahertz pulsed imaging.” Chemical Engineering Science (2018).

Abstract

The application of terahertz pulsed imaging (TPI) in the in-line configuration to monitor the coating thickness distribution of pharmaceutical tablets has the potential to improve the performance and quality of the spray coating process. In this study, an in-line TPI method is used to measure coating thickness distributions on pre-coated tablets during mixing in a rotating pan, and compared with results obtained numerically using the discrete element method (DEM) combined with a ray-tracing technique. The hit rates (i.e. the number of successful coating thickness measurements per minute) obtained from both terahertz in-line experiments and the DEM/ray-tracing simulations are in good agreement, and both increase with the number of baffles in the mixing pan. We demonstrate that the coating thickness variability as determined from the ray-traced data and the terahertz in-line measurements represents mainly the intra-tablet variability due to relatively uniform mean coating thickness across tablets. The mean coating thickness of the ray-traced data from the numerical simulations agrees well with the mean coating thickness as determined by the off-line TPI measurements. The mean coating thickness of in-line TPI measurements is slightly higher than that of off-line measurements. This discrepancy can be corrected based on the cap-to-band surface area ratio of the tablet and the cap-to-band sampling ratio obtained from ray-tracing simulations: the corrected mean coating thickness of the in-line TPI measurements shows a better agreement with that of off-line measurements.