An ion imprinted amino-functionalized mesoporous sorbent for the selective minicolumn preconcentration of cadmium ions and determination by GFAAS

摘要

An amino-functionalized mesoporous sorbent with double imprinting of Cd(II) and surfactant micelles was prepared via the sol–gel co-condensation method. The material was employed for the highly selective preconcentration of cadmium ions in minicolumns (MCs) and their determination at ultratrace levels by graphite furnace atomic adsorption spectroscopy (GFAAS). For comparative purposes ion-imprinted (II) and non-imprinted (NI) solids were characterized by SEM, FTIR and nitrogen gas adsorption–desorption. Batch experiments including pH influence, sorption capacity and adsorption kinetics were carried out in order to optimize the adsorption–elution process. II showed a greater absorption capacity than the NI sorbent being 122 mg g?1 and 67 mg g?1 respectively, consistent with the Langmuir isotherm equation. Sample and elution flow rates, volumes of the sample and eluent and the choice of the most suitable eluent were tested and optimized under MC dynamic conditions for both sorbents. Under optimized conditions, the II filling showed a preconcentration factor (PCF) of 50 whilst that of NI was 10. The higher selectivity of filler II was revealed when comparing the maximum tolerable limit (MTL) of interfering cations and anions commonly found in water samples, the MTL for II being 100 to 200 times higher than that for NI for all the tested concomitants. The main figures of merit found for solid II are: a limit of detection of 0.0011 ng mL?1 (3Sb), a linear range of 0.01–20 ng mL?1, and RSD% of 2 (n = 6; 0.05 ng mL?1). Additionally, the operation under dynamic conditions together with the employment of low volumes of the sample and eluent allowed a large lifetime of sorbent II of more than 700 cycles with no loss of sensitivity or need for refilling. The method was successfully applied to the determination of traces of Cd(II) in osmosis and tap water with recoveries of 98.8–101.3%. A full discussion will be provided, including a comparison with other methods already reported in the literature.