作者: Dipak Mainali , Jane Li , Peter Yehl , Nicholas Chetwyn
DOI: 10.1016/J.JPBA.2014.03.001
关键词: Water content 、 Analytical chemistry 、 Calibration (statistics) 、 Process engineering 、 Single product 、 Spectroscopy 、 Chemistry 、 Active ingredient 、 New product development 、 Near-infrared spectroscopy 、 Partial least squares analysis
摘要: Abstract Near infrared (NIR) spectroscopy has been widely used for the determination of water content in a wide variety samples. With few exceptions, all methods employ calibration model developed and applicable single product. The current study describes NIR method using single, comprehensive to predict tablets containing different active pharmaceutical ingredients (API). was range 2–13% (w/w) three APIs formulation compositions. To develop robust model, individual models were sequentially starting from simple one product including projects final partial least square analysis method. Data pretreatments spectral region selections performed during development optimize number factors correlation coefficients cross-validation prediction by model. reliably predicted tablet samples these products, can be updated measurements new drug products adding two purpose.