Optimization and evaluation of different chemical and electrochemical hydride generation systems for the determination of arsenic by microwave plasma torch optical emission spectrometry

The determination of trace concentrations of As and its species in water and sediment samples by the use of microwave plasma torch optical emission spectrometry (MPT-OES) and chemical (CHG) as well as different electrochemical hydride generation (EcHG) systems was studied, when using Ar and He as working gases for the microwave plasmas. Under optimized conditions and with He as working gas the detection limits (3sigma) for As (228.82 nm) were found to be 21 and 13 mug/l for chemical and electrochemical hydride generation, respectively. When Ar is used as working gas, the detection limits are higher, i.e., 60 and 48 mug/l for chemical and electrochemical hydride generation, respectively. Several miniaturized electrochemical hydride generation cells, among which some use glassy carbon foam and carbon fiber for the cathode, were used and the detection limits with these systems were found to be by a factor of 3-5 higher than in the conventional electrochemical hydride generation cell. The effects of Ca, Fe, Bi, Se, etc., on the determination of As with chemical and miniaturized electrochemical hydride generation systems were studied, and it was found that the interferences in electrochemical hydride generation were lower than in chemical hydride generation. The efficiency of the generation of AsH3 in chemical hydride generation and all electrochemical hydride generation systems, as determined by a coulometric titration of the remaining As(III) in the waste solutions of the gas-liquid separator, was found to be below 18% to 90%, depending on the cells. A modified graphite furnace (GF) unit was coupled to the hydride generation svstem for hot-trapping of the hydride forming elements. When trapping the AsH3 produced in a miniaturized electrochemical hydride generation system on Pd in a graphite furnace and sweeping the As into the He microwave plasma torch, the detection limit for As could be improved to 1.7 mug/l (improvement by a factor of 14). The procedure without trapping could be used for the deten-nination of As in a standard reference water (SR-M 1643d) containing 56.02 +/- 0.73 mg/l of total As within an experimental error of 8%. With the miniaturized electrochemical hydride generation and microwave plasma torch emission spectrometry in the case of trapping the total As could be determined in Saxony river sediment samples and in Hungarian spring water samples at the 10-30 and 50-360 mug/l levels, respectively. (C) 2004 Elsevier RV All rights reserved.