THIRD ORDER MULTIVARIATE CALIBRATION APPLIED TO THE QUANTITATION OF AZINPHOS-METHYL IN FRUITS BY EXCITATION-EMISSION-KINETIC MEASUREMENTS

M. MONTEMURRO; G. G. SIANO; M.J. CULZONI; H.C. GOICOECHEA

Congreso; XVI Chemometrics in Analytical chemistry; 2016

Azinphos-methyl (AZM) is a widely used organophosphate insecticide and acaricide, with demonstrated negative impacts on the environment [1]. The weak natural fluorescence of AZM in aqueous solution can be enhanced in a variety of ways, including UV photolysis, inclusion into cyclodextrins, and base hydrolysis. Upon absorption of UV-A radiation, this molecule undergoes photolysis to the highly fluorescent compound N-methylanthranilic acid, which undergoes subsequent photolysis to photochemically stable products [2].Four-way data generated by excitation-emission fluorescence matrices (EEFMs) measured as a function of reaction time can be modelled with appropriate higher order algorithms to achieve the second order advantage and improve both the selectivity and sensibility of the method.In this work, we developed a fluorescent kinetic method for the determination of AZM in fruit samples. It is based on third-order data, obtained by measuring the time evolution of the EEFMs of the photolysis of AZM in alkaline medium. The experiments were carried out with an instrument consisting in an optic fiber connected to an UV radiation source, coupled to a fluorescence spectrophotometer. The matrices were recorded in a quartz cell of 1 cm of path length. The readings were made in the excitation range of 220 to 320 nm every 5 nm, and emission from 320 to 500 nm every 5 nm at a scan rate of 24,000 nm min-1. EEFMs were measured every 30 seconds for six minutes, thus having 21 × 33 × 13 data points.Calibration and validation sets, consisting in five concentration levels of AZM each, were analyzed. The validation samples were prepared by adding fuberidazole, bitertanol and thiabendazole as uncalibrated interferences. Data modelling was performed with the algorithms MCR-ALS, PARAFAC and U-PLS/RTL. Relative error prediction and analytical figures of merit were calculated. The results obtained were similar for the three algorithms.The method was further used for the quantitation of AZM in apple, pear, peach and plum. For this purpose, the standard addition calibration method was used in order to overcome matrix effect, which involved spiking three levels of AZM standard to the real samples. The sample preparation procedure consisted in an extraction with acetonitrile followed by dispersive liquid-liquid microextraction. Recoveries and figures of merit were also calculated for real samples.