In this paper, a novel and miniaturised flow-through dialysis-based potentiometric detector is proposed for the determination of chloride in soil samples. The outstanding feature of the designed unit is the integration of analyte isolation from matrix constituents via membrane separation with differential potentiometric detection. Two identical tubular all-solid-state Ag/AgCl ion selective electrodes (ISEs) were assembled respectively at the inlet and outlet of the acceptor channel. Thus, as a consequence of the continuous forward flow of solutions through the microdialyser the outlet tube becomes the indicator electrode for the analyte diffusate while the nested tube at the entrance serves as reference electrode. The effect of physical and chemical parameters on the mass transfer efficiency is discussed in detail and compared with conventional configurations involving downstream detection. The membrane morphology for optimum dialysis performance is also thoroughly evaluated in terms of thickness, porosity and molecular weight cut-off. Higher dialysis efficiency and reduction of dilution factors up to a value of 5 were attained by halting the recipient stream temporarily. Under the optimised conditions, a dynamic working range of 5–5000 mg l−1 chloride with a linear interval between 10 and 5000 mg l−1
(for 1 min stopped-flow and 200 µl sample volume), a repeatability better than 3.0% and a 3 σblank detection limit of 1.2 mg l−1 chloride were the analytical figures of merit of the devised configuration. The potentiometric dialysis sensor features extreme tolerance to high molecular weight interfering matrix compounds (>1000 mg l−1 humic acid), which makes it specially suited for the interference-free potentiometric determination of chloride in soils containing high levels of organic matter. The miniature size, low-reagent consumption and high analytical throughput (25–40 h−1) also warrant its applicability to in-field monitoring or screening schemes. The accuracy of the measurements was assessed using ion-chromatography as an external reference method. A mean t-test showed no statistical differences between both methodologies at the 95% confidence level.
You have access to this article
Please wait while we load your content...
Something went wrong. Try again?