RESUMEN
A flow injection system with a pre-concentration minicolumn based on a chelating resin was coupled to a flame atomic absorption spectrometer. The focus of this work was the investigation of interference effects and the analytical applicability of the azo-immobilized 8-hydroxyquinoline on controlled-pore glass for the determination of Cr and Mn in mussel and non-fat milk powder. All studied concomitants affected the retention of Cr(III). These effects are probably related to the formation of hydroxo-complexes at the optimum pH range 9.0-10. The positive effect caused by Ca(II) was exploited to increase the retention of Cr(III) species and to improve the slope by 70%. The interferences on Mn(II) retention were less severe. The quantification of Cr and Mn was performed by standard additions. The proposed methodology was validated by analysis of three certified reference materials of mussels (Cr and Mn) and non-fat milk powder (Mn) with a mean relative percent error of <6.5% and mean relative standard deviation of <13%. Chromium and Mn were determined in typical Chilean mussels samples, and Mn was determined in non-fat milk powder samples. Results agreed at the 95% confidence level with those obtained by electrothermal atomic absorption spectrometry (ETAAS) using graphite furnace atomization. The method detection limits for a 30 s pre-concentration time were 0.9 and 1.1 micro g L(-1) for Mn, and 2.2 and 2.5 micro g L(-1) for Cr in acid digested solutions of mussel and non-fat milk, respectively. The methodology is simple, fast (sampling frequency 60-72 h(-1)), reliable, of low cost, and can be applied to the determination of traces of Cr (> or =0.18 micro g g(-1)) and Mn (> or =0.6 micro g g(-1)) in mussel samples, and Mn (> or =0.37 micro g g(-1)) in non-fat milk powder.