Tablet Characteristics Prediction by Powder Blending Process Analysis Based on near Infrared Spectroscopy

Abstract

Prediction of powder blend homogeneity and pharmaceutical characteristics such as powder flowability, mechanical strength and disintegration time of compressed solid dosages was carried out via on-line near infrared spectroscopy measurements. Blend homogeneity was evaluated in the pre-blending phase using both partial least-squares (PLS) and model-free regression methods. Covariance analysis between consecutive spectra was shown to represent a faithful monitoring of homogeneity as a model-free method, which coincided well with the results obtained from the PLS regression method. Similarly, during the post-blending phase with lubricant, pharmaceutical properties were accurately and precisely predicted using PLS regression models. These prediction models could be implemented in the determination of optimal blending time for a lubricant in view of improving flowability and of ensuring adequate mechanical strength and tablet disintegration time.