High precision osmium elemental and isotope measurements of North Atlantic seawater

International audienceThis study presents new Os isotope and elemental data of seawater from the North Atlantic Ocean. Two techniques have been optimised and used to achieve tracer–seawater equilibration: (i) heating of a seawater–tracer mixture at 100 °C with Br2, CrO3 and H2SO4 in Teflon bombs and solvent extraction of OsO4 with Br2; and (ii) heating of a seawater–tracer mixture to 300 °C with CrO3 and H2SO4 in a sealed glass tube (HPA-S-s) at 125 bars and back extraction with Br2. Both techniques yield similar total procedural blanks in the range of 14–18 fg, and have a total Os yield that ranges from 86 to 92%. For the 200 fg Os standard, the effective ionisation efficiency is between 6 and 10% demonstrating that single filament ionization is capable of achieving significant Os sensitivity. To assess the external reproducibility on the 187Os/188Os ratio, both methods have been systematically applied to a single sample from a water depth of 2000 m and yield an indistinguishable 187Os/188Os ratio of 1.026 ± 0.017 (n = 11) and 1.020 ± 0.013 (n = 3), for techniques (i) and (ii), respectively. The external reproducibility on the 187Os/188Os ratio and Os concentration is 1.6% and 3.7% respectively, demonstrating the capability of both methods for analysing Os in natural water. Both procedures have been applied to a seawater profile from the North East Atlantic Ocean (IB13: 60°29.4′ N, 19°59.2′ W, 2650 m water depth). The results show little measurable variability in either Os concentration or isotope composition with depth and the overall average for the whole profile is 1.024 ± 0.031 (3%) for the 187Os/188Os ratio and 9.68 ± 0.48 (4.9%) pg kg−1 for the Os concentration. Therefore, for this locality at least, Os appears to behave conservatively