The potentials of the bead injection scheme integrated into the sequential injection (SI) lab-on-valve (LOV) format, using reversed-phase materials for on-line matrix removal and analyte preconcentration, and interfaced with detection by electrothermal atomic absorption spectrometry (ETAAS), are exploited for the ultratrace determination of cadmium. The proposed approach relies upon the renewable surface concept and involves the incorporation of hydrophobic beads into channels of the LOV microconduit, thus serving as transient microcolumn packed reactors. After on-line analyte derivatization with diethyldithiophosphate (DDPA) in 1.3%
(v/v) nitric acid medium, the resulting non-charged chelate is adsorbed onto the surface of the column material. The analyte loaded beads are afterwards eluted with 30 µl of ethanol which, via air segmentation, is subsequently introduced into the graphite tube for quantification. The used beads are then backflushed and delivered to waste. The ETAAS measurement is synchronized with the separation/preconcentration steps of the ensuing sample.
Two reversed-phase sorbents, that is, poly(styrene-divinylbenzene) copolymer alkylated with octadecyl groups (C18-PS/DVB) and poly(tetrafluoroethylene)
(PTFE), were selected in view of their different morphology and physical properties. Although C18-PS/DVB beads fulfil the basic requirements to be handled readily in a renewable fashion in the LOV microconduits, i.e., bead size homogeneity and perfect spherical shape, yielding improved reproducibility as compared to the PTFE beads (3.4%
versus 4.3%), the latter material results in better retention efficiency (74%
versus 28%), enrichment factor (17.2 versus 7.4), linear dynamic range (0.05–1 µg l−1 Cd versus 0.2–1.5 µg l−1 Cd), and detection limit (5 ng l−1versus 135 ng l−1) for a sample loading volume of 1.25 ml. The designed procedure, which handles 12 samples h−1, was validated by determination of the cadmium content in environmental CRM and NIST standard reference materials and by cadmium spike recoveries of a human urine sample. Statistical comparisons between the proposed procedure and certified values revealed no significant differences at the 95% confidence level.