Multivariate optimization of solvent extraction with 1,1,1-trifluoroacetylacetonates for the determination of total and labile Cu and Fe in river surface water by flame atomic absorption spectrometry

A procedure for simultaneous solvent extraction of Cu(H) and Fe(Ill) from river surface waters as diethyldithiocarbamates (DDC) 1,1,1-trifluotoacetylacetonates into isobutyl methyl ketone (IBMK) has been developed prior to their determination by flame atomic absorption spectrometry. Experimental design approaches were used in order to obtain the best compromise conditions for simultaneous solvent extraction. Variables such as pH, sodium DDC or 1,1,1-trifluoro-2,4-pentanodione (H(TFA)) concentrations, reaction time, IBMK volume and extraction time have been optimized. First, Plackett-Burman designs involving 13 experiments and 2 replicates were used as screening designs to determine which variables were significant. DDC or H(TFA) concentration, as well as pH and IBMK volume were found statistically significant and they were finally optimized by applying a central composite design (15 experiments and 2 replicates). Optimum values for these variables were selected for compromise extraction conditions of Cu(II) and Fe(III) species. An optimum pH of 5.3 was chosen for Cu-H(TFA) and Fe-H(TFA) formation with an optimum H(TFA) concentration of 0.20% (m/v). The optimum IBMK volume for extraction was 8 ml, allowing a pre-concentration factor of 10. A microwave-assisted peroxydisulfate oxidation was used to break down the metal-organic matter complexes in river surface waters in order to assess the total Cu and Fe contents. Applying the experimental design approach, optimum conditions was an irradiation for 5.0 min at a microwave power of 500 W using 0.5 g of ammonium peroxydisulfate. The method was applied to determine total Cu and Fe contents and also labile Cu(II) and Fe(III) contents in several river surface water samples. (C) 2002 Elsevier Science B.V. All rights reserved.