Use of near-infrared spectroscopy to quantify drug content on a continuous blending process: influence of mass flow and rotation speed variations

Abstract

The aim of this study was to develop a quantitative Near-Infrared (NIR) method which monitors the homogeneity of a pharmaceutical formulation coming out of a continuous blender. For this purpose, a NIR diode array spectrometer with fast data acquisition was selected. Additionally, the dynamic aspects of a continuous blending process were studied; the results showed a well-defined cluster for the steady state, and the paths for the start-up and emptying stages were clearly identified. The end point of the start-up phase was detected by moving block of standard deviation, relative standard deviation, and principal component analysis. A partial least square (PLS) model was generated for the quantification of the drug, with a standard error of prediction of 0.2% m/m. The PLS model was successfully applied for monitoring the drug level at the outlet of the continuous blender. Furthermore, the PLS model was tested under different flow and stirring rates. Flow and stirring rate variations caused different powder flow dynamics, which were reflected on the NIR measurements. Therefore, the PLS model was sensitive to changes in mass flow and rotation speeds.