Mercury is one of the most harmful pollutant present on Earth and shows some peculiar characteristics which make this element relevant to study. Hg undergoes to complex biogeochemical transformations in the environment and shows different isotopic fractionation phenomena of both mass dependant (MDF) and mass independent (MIF) nature. The study of the isotopic mass fractionation, seems to be a powerful tool to trace and identify the sources of pollution and the behavior of this element in natural systems and organisms. In order to determine the isotope ratios of mercury an analytical procedure was set up by using a multicollector ICP/MS coupled with an on-line reduction apparatus as sample introduction system. Instrumental mass bias was corrected by using the 205Tl/203Tl isotopic reference coupled to a standard-sample bracketing approach to compensate any instrumental drift. Due to the high mass of the element, the fractionation phenomena are very small thus, the measure precision must be good enough to be able to significantly detect the differences in the isotopic composition. Because of the high number of instrumental parameters which must be set and which may influence the measurements, the instrumental conditions were optimized in order to obtain the suitable internal precision even for sample with low amount of mercury. The optimization step of the instrumental determination was conducted by means of the most advanced and innovative chemometric techniques of Design of Experiments (DoE). The multivariate approach is suitable for optimization purposes and in particular when many parameters must be varied at the same time such in this case. In facts, this method allows obtaining at the same time the maximum useful information with the minor number of experiments. Moreover, the multilinear regression models built on DoE data allow not only to determine the influence of the parameters on the finale responses but also the interactions among the parameters themselves. For the determination of the mercury isotope ratios five parameters were chosen for the optimization: the number of integration cycles, the integration time per cycle, the concentration of mercury in the samples, the concentration of thallium, the internal standard, and the flux of the solutions of mercury and the reduction agents through the reduction system. The responses to be optimized were the internal precisions for each measurement, expressed as RSE%. By means of the models built on the DoE data it was possible to study the influence of the parameters on the measurement precisions and to predict the optimum parameters in different instrumental conditions. The method was applied to sediment samples collected along an ideal transect ranging from the Idrija Hg mine district, Slovenia, to the Aussa-Corno river. Results showed that the isotopic approach can be used as an useful tracing tool in natural environment.
metadata only access