Abstract: The membrane extraction-pre-concentration (MEP) cell consisted of two channels separated by a Celgard 2500 membrane (0.025 mm thick, 45% porosity, 0.04 µm effective pore size). Oil samples diluted with hexane (1:4) were propelled through one channel and the acceptor solution was propelled through the other channel. The acceptor solution was held stationary during the extraction-pre-concentration period while the diluted oil sample was circulated continuously. The MEP cell was coupled online to FIA or HPLC systems for the determination of triazines in vegetable oils. The FIA system used 0.1 M HClO4 as the acceptor solution. At the end of the 5-15 min MEP period, the acceptor solution was propelled to the diode-array spectrophotometer where the total triazines were determined at 220 nm. The calibration graphs were linear (range not given) with detection limits of 0.91 ppm. RSD (n = 10) were 6%. The HPLC system used 5 mM HClO4 in methanol/H2O (9:1) as the acceptor solution. After MEP, the acceptor solution was analyzed on a 5 µm Spheri-5-RP 18 column (25 cm x 4.6 mm i.d.) with 50% aqueous acetonitrile as mobile phase (1.25 ml/min) and detection at 220 nm. Calibration graphs for the determination of atrazine, ametryne, prometryne and terbutryne in corn, sunflower and olive oil were linear for 0.05-20 ppm and the detection limits were ~e;0.1 ppm. RSD (n = 10) were 10%.

Abstract: Anti-terbutryne mAb were immobilized to polystyrene and glass beads, and the different immobilization techniques were compared for their applicability in FIA. The immobilization methods used were (i) covalent immobilization to glass beads using acetic acid/periodate with subsequent coupling of biotinylated antibodies; (ii) adsorption of avidin to polystyrene beads and coupling of biotinylated antibodies and (iii) covalent immobilization of antibodies to glass beads with glutaraldehyde. Methods (ii) and (iii) were the most effective, and (iii) was found most effective for FIA. The antibodies could be regenerated at least 60 times. The calibration graph was linear from 0.01 µgl (detection limit) to 5 µg/l. The method could be used for monitoring drinking water sources.