Atmospheric noble gases in Mid-Ocean Ridge Basalts: Identification of atmospheric contamination processes

Noble gases in oceanic basalts always show the presence in variable proportions of a component having elemental and isotopic compositions that are similar to those of the atmosphere and distinct from the mantle composition. Although this component could be mantle-derived (e.g. subduction of air or seawater-derived noble gases trapped in altered oceanic crust and sediments), it is most often suggested that this air component is added after sample collection and probably during storage at ambient air, although the mechanism remains unknown. In an attempt to reduce this atmospheric component observed in MORBs, four experimental protocols have been followed in this study. These protocols are based on the hypothesis that air can be removed from the samples, as it appears to be sheltered in distinct vesicles compared to those filled with mantle gases. All of the protocols involve a glove box filled with nitrogen, and in certain cases, the samples are stored under primary vacuum (lower than 10−2 mbar) to pump air out or, alternatively, under high pressure of N2 to expel atmospheric noble gases. In all protocols, three components are observed: atmospheric, fractionated atmospheric and magmatic. The fractionated air component seems to be derived from the non-vitreous part of the pillow-lava, which has cooled more slowly. This component is enriched in Ne relative to Ar, reflecting a diffusive process. This contaminant has already been observed in other studies and thus seems to be relatively common. Although it is less visible, unfractionated air has also been detected in some crushing steps, which tends to indicate that despite the experiments, air is still present in the vesicles. This result is surprising, since studies have demonstrated that atmospheric contamination could be limited if samples were stored under nitrogen quickly after their recovery from the seafloor. Thus, the failure of the protocols could be explained by the insufficient duration of these protocols or by the inaccessibility of vesicles filled with air as assessed by (Ballentine and Barfod, 2000).