Abstract

Two sensitive analytical atomic spectrometry methods, electrothermal atomic absorption spectrometry (ET-AAS) and hydride generation coupled to atomic fluorescence spectroscopy (HG-AFS), were optimized for determining total antimony in soils and plant (alfalfa) matrices. The dry soils were digested with HNO₃–HCl–HF mixture, while, for the freeze dry alfalfa samples, HNO₃–H₂SO₄–H₂O₂ or HNO₃–H₂O₂ mixtures were used. The microwave oven digestion procedures chosen allowed the total dissolution of the matrices. The experimental parameters of both spectrometric techniques were optimized using standard solutions of Sb(III) and/or Sb(V), and digested solutions of soil and alfalfa samples. Since in the antimony determination by HG-AFS the kinetic of the hydride generation is dependent on the antimony oxidation state, a chemical reduction of Sb(V) to Sb(III) was carried out prior to the stibine generation. For this purpose, KI and L-cysteine were used as reducing agents, assaying different experimental conditions. The reduction of Sb(V) in plant solutions by a KI–ascorbic acid mixture can be performed at room temperature, while the Sb(V) reduction from soils solutions was quantitative when the procedure was accomplished in a microwave oven or at 90 °C in a water bath. For antimony determination by HG-AFS, the simple calibration mode was used, because this technique is less sensitive to interferences. For antimony determination by ET-AAS the use of a chemical modifier is unavoidable. Similar amounts of nickel or palladium were effective in stabilizing the antimony species present in soils and plant solutions; however, the best analytical signals were obtained using mixtures of this metals with NH₄H₂PO₄ and citric acid. Due to the matrix interference for determining antimony by ET-AAS, the standard additions method was used. The accuracy of the proposed methods were assessed by analyzing two certified reference soils (CRM) from NIST, San Joaquín soil (SRM 2709) and Montana soil (SRM 2710) and a reference vegetal material, Virginia tobacco leaves (CTA-VTL-2). In allcases the results obtained by both techniques agreed with the certified values. Under the optimized conditions, a detection limit of 0.08 µg l⁻¹ of Sb(III) was achieved by HG-AFS, with a precision of 4.3% for0.5 µg l⁻¹Sb(III); the calibration graph was linear from 0.25 to 250 µg l⁻¹. The detection limit obtained by ET-AAS, [injecting 20 µl Sb(III) solution and 10 µl chemical modifier mixture (2 µg Ni + 100 µg NH₄H₂PO₄ + 50 µg citric acid)] was 9 pg Sb, with a precision of 4.7% for 100 pg Sb. The proposed methods were successfully applied to the Sb determination in soils and alfalfa samples, from the Valparaíso region in Chile. In all samples the antimony concentrations found were higher than the average reported for Sb concentration in soils and vegetable.