摘要：The prediction power of partial least squares (PLS) and multivariate curve resolution-alternating least squares (MCR-ALS) methods have been studied for simultaneous quantitative analysis of the binary drug combination – doxylamine succinate and pyridoxine hydrochloride.Analysis of first-order UV overlapped spectra was performed using different PLS models – classical PLS1 and PLS2 as well as partial robust M-regression (PRM).These linear models were compared to MCR-ALS with equality and correlation constraints (MCR-ALS-CC).All techniques operated within the full spectral region and extracted maximum information for the drugs analysed.The developed chemometric methods were validated on external sample sets and were applied to the analyses of pharmaceutical formulations.The obtained statistical parameters were satisfactory for calibration and validation sets.All developed methods can be successfully applied for simultaneous spectrophotometric determination of doxylamine and pyridoxine both in laboratory-prepared mixtures and commercial dosage forms.
关键词：doxylamine succinate;pyridoxine hydrochloride;binary mixture;partial least squares (PLS);partial robust M-regression (PRM);multivariate curve resolution-alternative least squares (MCR-ALS)
