Identification of Dimethoate-containing Water Using Partitioned Dispersive Liquid-liquid Microextraction Coupled with Near-infrared Spectroscopy

抄録

  A simple, rapid and efficient extraction procedure, partitioned dispersive liquid-liquid microextraction, has been developed in combination with near-infrared spectroscopy for the extraction and discrimination of dimethoate from aqueous samples. For this technique, the appropriate mixture of extraction solvent (CCl₄) and disperser solvent (THF) was utilized. Partial least squares discriminant analysis was applied to build the model with several pre-process methods over the wavenumber regions between 7100 cm⁻¹ to 7300 cm⁻¹. The best model gave satisfactory classification accuracy, 98.6% for calibration set (n=74) and 97.6% for prediction set (n=42), using preprocessing of standard normal variate followed by Savitzky-Golay first derivative. The method was successfully applied to bottled water, tap water, lake water and farm water samples. The results demonstrated the possibility of near-infrared spectroscopy after partitioned dispersive liquid-liquid microextraction for the identification of water contaminated by dimethoate.