Determination of iron in seawater by electrothermal atomic absorption spectrometry and atomic fluorescence spectrometry: a comparative study.

Two methods available for direct determination of total Fe in seawater at low concentration level have been examined: electrothermal atomization atomic absorption spectrometry (ETAAS) and electrothermal atomization laser excited atomic fluorescence spectrometry (ETA-LEAFS). In a first part, we have optimized experimental conditions of ETAAS (electrothermal program, matrix chemical modification) for the determination of Fe in seawater by minimizing the chemical interference effects and the magnitude of the simultaneous background absorption signal. By using the best experimental conditions, a detection limit of 80 ng L(-1) (20 microL, 3sigma) for total Fe concentration was obtained by ETAAS. Using similar experimental conditions (electrothermal program, chemical modification), we have optimized experimental conditions for the determination of Fe by LEAFS. The selected experimental conditions for ETA-LEAFS: excitation wavelength (296.69 nm), noise attenuation and adequate background correction led to a detection limit (3sigma) of 3 ng L(-1) (i.e. 54 pM) for total Fe concentration with the use a 20 microL seawater sample. For the two methods, concentration values obtained for the analysis of Fe in a NASS-5 (0.2 microg L(-1)) seawater sample were in good agreement with the certified values.

Direct fluorescence monitoring of coal organic matter released in seawater.

Whenever immersed in seawater after a collier accident, a fossil fuel such as coal could become a source of pollution to the marine environment. To study the effect of such a contamination, four coal samples from different origins were used. A first analysis on those coals enabled us to determine the content of polycyclic aromatic hydrocarbons. Seawater was then mixed with coal to study the organic matter released from coal into seawater. Fluorescence was used for its sensitivity to aromatic compounds, with the additional purpose of evaluating the relevance of using an immersed fluorescence probe to monitor water pollution. Excitation-emission matrices were recorded and the excitation-emission wavelength range corresponding to the highest fluorescence intensity was 230 nm/[370 nm; 420 nm]. The samples with coal happened to fluoresce more than the coal-free samples, the difference depending on the coal origin. The fluorescence intensity increased with coal mass, up to some limit. The particle size also influenced the fluorescence intensity, the finest particles releasing more fluorescing substances, due to their higher exchange surface. When seawater percolated through coal, the samples fluoresced highly at the beginning, and then the fluorescence intensity decreased and reached the seawater level. However, even with a 10 ns acquisition time shift, the fluorescence spectra were not specific enough to show the presence of PAHs in the samples, which were too diluted to be detected, whenever released from coal into seawater. The lifetimes of the seawater and of the coal samples were respectively 4.7 and 3.8 ns, indicating that the substances released from coal mainly consisted of short-lived fluorescing substances, such as natural humic or fulvic substances. Consequently, the presence of coal does not seem to be too detrimental to the marine environment, and a direct fluorescence probe could be used to monitor the seawater organic charge increase due to the immersion of coal in seawater.

Improvement of direct determination of Cu and Mn in seawater by GFAAS and total elimination of the saline matrix with the use of hydrofluoric acid.

Hydrofluoric acid, added to seawater, can assist in the removal of chloride in the drying step by precipitating fluoride salts, thus suppressing the chloride interference effects induced on the atomization signals of Cu and Mn. By adding HF to seawater before the analysis, MgF(2) and CaF(2) are precipitated at the bottom of the sampling flask, without precipitating Cu and Mn, and are consequently not introduced into the graphite furnace. Because sodium salts are eliminated at the pretreatment step, the whole seawater matrix is eliminated before the atomization of Cu or Mn. Therefore, the analyzed volume of seawater can be increased by using the multi-injection procedure without degradation of the limit of detection and risks of spectral interferences. The limit of detection obtained for Cu and Mn are 0.05 and 0.01mugL(-1), respectively, for a 50muL analyzed seawater volume.

Speciation of major arsenic species in seawater by flow injection hydride generation atomic absorption spectrometry.

Arsenic present at 1 microg L(-1) concentrations in seawater can exist as the following species: As(III), As(V), monomethylarsenic, dimethylarsenic and unknown organic compounds. The potential of the continuous flow injection hydride generation technique coupled to atomic absorption spectrometry (AAS) was investigated for the speciation of these major arsenic species in seawater. Two different techniques were used. After hydride generation and collection in a graphite tube coated with iridium, arsenic was determined by AAS. By selecting different experimental hydride generation conditions, it was possible to determine As(III), total arsenic, hydride reactive arsenic and by difference non-hydride reactive arsenic. On the other hand, by cryogenically trapping hydride reactive species on a chromatographic phase, followed by their sequential release and AAS in a heated quartz cell, inorganic As, MMA and DMA could be determined. By combining these two techniques, an experimental protocol for the speciation of As(III), As(V), MMA, DMA and nonhydride reactive arsenic species in seawater was proposed. The method was applied to seawater sampled at a Mediterranean site and at an Atlantic coastal site. Evidence for the biotransformation of arsenic in seawater was clearly shown.

Effects of various salts on the determination of arsenic by graphite furnace atomic absorption spectrometry. Direct determination in seawater.

The effect of Na, Mg, Ca and Sr as their nitrate, chloride and sulfate salts and seasalt, with and without the use of palladium, on the determination of arsenic by electrothermal atomic absorption spectrometry was investigated. In the absence of any stabilizing agent, arsenic was partially lost as molecular species at low temperatures. The effect of salts on the shape of the atomization signal, the integrated absorbance and the stabilizing effect were highly dependent both on their nature and mass. By trapping arsenic, oxide species resulting from the decomposition of nitrate salts induced a high stabilization effect depending on their vaporization temperatures: MgO approximately CaO>SrO>Na2O. The stabilization effect of chlorides occurred about 200 degrees C lower and depended on mass, volatility and hydrolytic properties: SrCl2>CaCl2>MgCl2 approximately NaCl. The effect of sulfates was mainly dependent on their decomposition/vaporization mechanisms, and in the presence of Na2SO4 or CaSO4 a strong chemical interference effect was observed. Palladium stabilized arsenic in the presence of nitrates, chlorides or even sulfates, leading to a similar delaying effect, signal shape and integrated absorbance. Seasalt induced also important modifications to the atomization signal of As. Moreover, an interference effect was observed, which could probably be attributed to the simultaneous vaporization of sulfate in seasalt. In seawater, Pd suppressed this interference effect and permitted to use a high pyrolysis temperature up to 1400 C to remove the major part of the seawater matrix before atomization. Under optimized conditions, the detection limit for As obtained in unmodified seawater in the presence of Pd was 0.34 microg L(-1) for a 10 microl sample.

Determination of 1 ng/l. Levels of mercury in water by electrothermal atomization atomic-absorption spectrometry after solvent extraction.

Optimization of the furnace parameters for electrothermal atomization of mercury leads to a characteristic mass of 20 pg in aqueous solution and 30 pg in chloroform extracts (with Zeeman correction). With a single-step solvent extraction from 100 ml of sample, performed in the sampling vessel, and direct injection of 400 microl of the extract into the furnace, a characteristic concentration of approximately 0.8 ng/l. is reached.