Determination of adsorbable organic halogens in surface water samples by combustion–microcoulometry versus combustion–ion chromatography titration
Résumé
Adsorbable Organic Halogen (AOX) is an analytical parameter of considerable interest since it allows to evaluate the amount of organohalogen disinfection by-products (OXBPs) present in a water sample. Halogen speciation of AOX into adsorbable organic chlorine, bromine and iodine, respectively AOCl, AOBr and AOI, is extremely important since it has been shown that iodinated and brominated organic by-products tend to be more toxic than their chlorinated analogues. Chemical speciation of AOX can be performed by combustion-ion chromatography (C-IC). In the present work, the effectiveness of the nitrate wash according to ISO 9562 standard method protocol to eliminate halide ions interferences was firstly examined. False positive AOX values were observed when chloride concentration exceeded 100 ppm. The improvements made to the washing protocol have eliminated chloride interference for concentrations up to 1000 ppm. A C-IC method for chemical speciation of AOX into AOCl, AOBr, and AOI has been developed and validated. The most important analytical parameters were investigated. The following optimal conditions were established: an aqueous solution containing 2.4 mM sodium bicarbon-ate/2.0 mM sodium carbonate, and 2% acetone (v/v) as mobile phase, 2 mL of aqueous sodium thiosulfate (500 ppm) as absorption solution, 0.2 mL min −1 as water inlet flow rate for hydropyrolysis, and 10 min as post-combustion time. The method was validated according to NF T90-210 standard method. Calibration curves fitted through a quadratic equation show coefficients of determination (r 2) greater than 0.9998, and RSD less than 5%. The LOQs were 0.9, 4.3, and 5.7 g L −1 Cl for AOCl, AOBr, and AOI, respectively. The accuracy, in terms of relative error, was within a ± 10% interval. The applicability of the validated method was demonstrated by the analysis of twenty four water samples from three rivers in France. The measurements reveals AOX amounts above 10 g L −1 Cl in all untreated samples, suggesting the presence of organohalogen compounds in the sampled rivers. On weight concentration basis, AOCl accounted for 77-100% of AOX in the treated water samples. A good agreement between the conventional AOX method and the developed C-IC method was found.