Two multivariate strategies applied to three-way kinetic spectrophotometric data for the determination of mixtures of the pesticides carbaryl and chlorpyrifos

Two pesticides, carbaryl and chlorpyrifos, have been simultaneously determined using second-order kinetic spectrophotometric measurements upon alkaline oxidative degradation. In spite of the complexity of the system and of the serious spectral overlap among the reagents and products, calibration and prediction is possible thanks to the power of second-order multivariate techniques. Strategies such as parallel factor analysis (PARAFAC) and multivariate curve resolution coupled to alternating least-squares (MCR-ALS) have been employed, which adequately exploit the second-order advantage. They allow for a correct determination of the analytes both in synthetic binary samples and in a commercial formulation, in this latter case even in the presence of unmodeled interferents. Multi-way partial least-squares (n-PLS) produced good results only on synthetic binary mixtures but could not be applied to a commercial sample because it contained an uncalibrated component.