Spectrophotometric determination of iron species using a combination of artificial neural networks and dispersive liquid-liquid microextraction based on solidification of floating organic drop

Abstract

A dispersive liquid-liquid microextraction based on solidification of floating organic drop (DLLME-SFO) and artificial neural networks method was developed for the simultaneous separation/preconcentration and speciation of iron in water samples. In this method, an appropriate mixture of ethanol (as the disperser solvent) and 1-undecanol (as the extracting solvent) containing appropriate amount of 2-thenoyltrifluoroacetone (TTA) (as the complexing agent) was injected rapidly into the water sample containing iron (II) and iron (III) species. At this step, the iron species interacted with the TTA and extracted into the 1-undecanol. After the phase separation, the absorbance of the extracted irons was measured in the wavelength region of 450-600 nm. The artificial neural networks were then applied for simultaneous determination of individual iron species. Under optimum conditions, the calibration graphs were linear in the range of 95-1070 μg L(-1) and 31-350 μg L(-1) with detection limits of 25 and 8 μg L(-1) for iron (II) and iron (III), respectively. The relative standard deviations (R.S.D., n=6) were lower than 4.2%. The enhancement factor of 162 and 125 were obtained for Fe(3+) and Fe(2+) ions, respectively. The procedure was applied to power plant drum water and several potable water samples; and accuracy was assessed through the recovery experiments and independent analysis by graphite furnace atomic absorption spectrometry.