A comparison of in situ vs. ex situ filtration methods on the assessment of dissolved and particulate metals at hydrothermal vents

The objective of this study was to assess the impact of the filtration method (in situ vs. ex situ) on the dissolved/particulate partitioning of 12 elements in hydrothermal samples collected from the Lucky Strike vent field (Mid-Atlantic Ridge; MAR). To do so, dissolved ( <0.45 mu m) and particulate Mg, Li, Mn, U, V, As, Ba, Fe, Zn, Cd, Pb and Cu were measured using different techniques (HR-ICP-MS, ICP-AES and CCSA). Using in situ filtration as a baseline, we showed that ex situ filtration (on-board and on shore after freezing) resulted in an underestimation of the dissolved pool, which was counterbalanced by an overestimation of the particulate pool for almost all the elements studied. We also showed that on-board filtration was acceptable for the assessment of dissolved and particulate Mn, Mg, Li and U for which the measurement bias for the dissolved fraction did not exceed 3%. However, in situ filtration appeared necessary for the accurate assessment of the dissolved and particulate concentrations of V, As, Fe, Zn, Ba, Cd, Pb and Cu. In the case of Fe, on-board filtration underestimated the dissolved pool by up to 96%. Laboratory filtration (after freezing) resulted in a large bias in the dissolved and particulate concentrations, unambiguously discounting this filtration method for deep-sea chemical speciation studies. We discuss our results in light of the precipitation processes that can potentially affect the accuracy of ex situ filtration methods.

Metal measurement in aquatic environments by passive sampling methods: Lessons learning from an in situ intercomparison exercise

Passive sampling devices (PS) are widely used for pollutant monitoring in water, but estimation of measurement uncertainties by PS has seldom been undertaken. The aim of this work was to identify key parameters governing PS measurements of metals and their dispersion. We report the results of an in situ intercomparison exercise on diffusive gradient in thin films (DGT) in surface waters. Interlaboratory uncertainties of time-weighted average (TWA) concentrations were satisfactory (from 28% to 112%) given the number of participating laboratories (10) and ultra-trace metal concentrations involved. Data dispersion of TWA concentrations was mainly explained by uncertainties generated during DGT handling and analytical procedure steps. We highlight that DGT handling is critical for metals such as Cd, Cr and Zn, implying that DGT assembly/dismantling should be performed in very clean conditions. Using a unique dataset, we demonstrated that DGT markedly lowered the LOQ in comparison to spot sampling and stressed the need for accurate data calculation.