Multivariate optimization of solid-phase extraction applied to iron determination in finished waters.

In this work, Amberlite XAD-4 resin functionalized with salicylic acid was synthetized, characterized and applied as a new packing material for an on-line system to iron determination in aqueous samples. The detection method is based on the sorption of Fe(III) ions in a minicolumn containing the synthesized resin, followed by a desorption step using an acid solution and measurement of iron by vis-spectrophotometry (CAS method). The optimization of the solid-phase extraction system was performed using factorial design and Doehlert matrix considering six variables: sample percolation rate (0.5-9 ml min(-1)), sample metal concentration (20-200 microg l(-1)), flow-through sample volume (0-5 ml) (all three directly linked to the extraction step), elution flow-rate (0.5-9 ml min(-1)), concentration and volume of eluent (HCl 0.1-0.5M) (all three directly linked to the elution step). The aim of this study was to obtain a set of operating ranges for the six variables tested in order to obtain--by means of a mathematical function allowing maximisation of each response (desirability function)--at least 90% of iron recovery rates. Using the experimental conditions defined in the optimization, the method allowed iron determination with achieved detection limit of 2.3 microgl(-1) and precision (assessed as the relative standard deviation) of 9.3-2.8% for iron solutions of 10.0-150 microgl(-1). Real samples (coming from a water treatment unit) were used successfully when evaluating potentialities of the developed SPE procedure coupled to a spectrophotometric determination.