Article révisé par les pairs
Résumé : The present study evaluates several critical issues related to precision and accuracy of Cu and Zn isotopic measurements with application to estuarine particulate materials. Calibration of reference materials (such as the IRMM 3702 Zn) against the JMC Zn and NIST Cu reference materials were performed in wet and/or dry plasma modes (Aridus I and DSN-100) on a Nu Plasma MC-ICP-MS. Different mass bias correction methods were compared. More than 100 analyses of certified reference materials suggested that the sample-calibrator bracketing correction and the empirical external normalisation methods provide the most reliable corrections, with long term external precisions of 0.06 and 0.07‰ (2SD), respectively. Investigation of the effect of variable analyte to spike concentration ratios on Zn and Cu isotopic determinations indicated that the accuracy of Cu measurements in dry plasma is very sensitive to the relative Cu and Zn concentrations, with deviations of δ65Cu from -0.4‰ (Cu/Zn = 4) to +0.4‰ (Cu/Zn = 0.2). A quantitative assessment (with instrumental mass bias corrections) of spectral and non-spectral interferences (Ti, Cr, Co, Fe, Ca, Mg, Na) was performed. Titanium and Cr were the most severe interfering constituents, contributing to inaccuracies of -5.1‰ and +0.60‰ on δ68/64Zn, respectively (for 500 μg l-1Cu and Zn standard solutions spiked with 1000 μg l-1 of Ti or Cr). Preliminary isotopic results were obtained on contrasting sediment matrices from the Scheldt estuary. Significant isotopic fractionation of zinc (from 0.21‰ to 1.13‰ for δ66Zn) and copper (from -0.38‰ to 0.23‰ for δ65Cu), suggest a control by physical mixing of continental and marine water masses, characterized by distinct Cu and Zn isotopic signatures. These results provide a stepping-stone to further evaluate the use of Cu and Zn isotopes as biogeochemical tracers in estuarine environments. © 2008 The Authors. Journal compilation © 2008 International Association of Geoanalysts.