NIR spectroscopic methods for monitoring blend potency in a feed frame - calibration transfer between offline and inline using a continuum regression filter

Abstract

A material sparing method for near-infrared (NIR) calibration was developed using an offline apparatus coupled with a calibration transfer method to enable a partial least squares (PLS) model to monitor the concentration of active pharmaceutical ingredients (API) in the feed frame of a rotary tablet press. The offline apparatus was designed to simulate the powder flow dynamic and NIRS measurement environment of a tablet-press feed frame. A comprehensive experimental design, including calibration and testing, was employed to determine blend inhomogeneity. NIR spectra were collected at both the feed frame (inline) conditions and the simulator (offline) conditions. The simulator conditions were designed to mimic the density and powder flow in the feed frame during the actual tableting process. The offline data were pretreated by an orthogonalization-based calibration transfer algorithm, a continuum regression filter (CR filter), before being subjected to PLS modeling. This study demonstrated: (1) calibration for inline application can be generated using an offline apparatus, and (2) the CR filter, as an innovative calibration transfer method, can generalize the offline method for multiple feed frame conditions.