Process-induced inadvertent phase change of an active pharmaceutical ingredient in a drug product could impact chemical stability, physical stability, shelf life, and bioperformance. In this study, dispersive Raman spectroscopy is presented as an alternative method for the nondestructive, high-throughput, at-line quantification of amorphous conversion. A quantitative Raman method was developed using a multivariate partial least squares (PLS) regression calibration technique with solid-state nuclear magnetic resonance (ssNMR) spectroscopy as the reference method. Compositionally identical calibration tablets containing 20% w/w total MK-A drug in varying weight proportions (0%-50% w/w based on total MK-A) of amorphous and crystalline MK-A were compressed at 10-45 kN force. PLS predictions of amorphous content of tablets using Raman spectroscopy correlated well with ssNMR quantification. The predictive accuracy of this model led to a strong correlation (R2 = 0.987) with a root mean-squared error of prediction of 1.5% w/w amorphous MK-A in tablets up to 50% w/w amorphous conversion in compressive stress range of 60-320 MPa. Overall, these results suggest that dispersive Raman spectroscopy offers fast, sensitive, and high-throughput (<5 min/tablet) method for quantitating amorphous conversion.
Keywords: amorphous quantification; chemometrics; dispersive Raman spectroscopy; multivariate analysis; partial least squares; process-induced amorphization; solid-state NMR.
Copyright © 2017. Published by Elsevier Inc.