| Résumé | A Beenakker microwave-induced plasma (MIP) and a furnace atomization plasma excitation spectrometry (FAPES) source were compared with respect to performance capabilities for the analysis of derivatised mercury species when coupled with high resolution GC for sample introduction. Natural gas condensate was used as a test material. Emission was monitored at the 253.6 nm mercury line. For standard solutions, comparable LODs were estimated for dimethyl-, methylbutyl- and dibutylmercury for both sources that were between 0.5 and 4.1 pg. In diluted condensate (5%) detection limits were between 1.5 and 4.7 pg. However, measurement of dimethylmercury was not possible with the MIP as the plasma was extinguished due to solvent vapour overload. Precision of replicate measurement of dimethyl-, methylbutyl- and dibutylmercury was better than 2% RSD for 2.5 µl injections to the FAPES at concentrations well above the LOD. The linear range spans 3 decades in both sources. Molecular emission from NO bands gives rise to background in both sources. In the FAPES source, hydrocarbon effluents reduce NO concentrations giving rise to background fluctuations, whereas in the MIP background fluctuations arise mainly from quenching of the plasma. |
|---|
