LAUSR

Abstract The origin and evolution of terrestrial volatiles is a fundamental question in Earth and planetary sciences. To better address this, it is necessary to estimate the present-day isotopic and elemental composition of volatiles in the mantle, which is difficult because degassing during magma transport to the surface changes the initial elemental composition. However, some basalts are considered to be relatively un-degassed and therefore can be used to determine the concentrations of C, N, H 2 O and noble gases in the mantle. This is the case of the so-called popping rock sample 2 π D43, found near 14°N on the Mid-Atlantic Ridge in 1985, which has been considered as a reference for upper mantle isotopic and elemental volatile compositions due to its high gas content. Such gas-rich samples have been recovered in only a few places on mid-ocean ridges but are crucial to characterizing the heterogeneities in mantle composition and therefore the cycling of volatiles on Earth. In order to better understand the occurrence of popping rocks, mantle heterogeneities and volatile cycling on Earth, cruise AT33-03 on R / V Atlantis was conducted in 2016 on the Mid-Atlantic Ridge near 14°N. Popping rocks were found in the same area as sample 2 π D43, using the submersible Alvin , and their noble gas compositions were determined by step-crushing and laser ablation. A new anaerobic sampling protocol was tested in an attempt to minimize atmospheric contamination, which is ubiquitous in basaltic glass samples used for noble gas analyses, and to evaluate laser extraction techniques. Preliminary step-crushing results suggest that the new anaerobic sampling protocol reduces atmospheric influence on the noble gas measurements. Comparison with laser ablation data shows that there is still some atmospheric contamination, suggesting that contamination begins on the seafloor. Two groups of samples have been identified based on their noble gas compositions: one group with their highest measured ratios similar to those of sample 2 π D43 ( 20 Ne/ 22 Ne of ∼12.5, 40 Ar/ 36 Ar of ∼27,000), and another group with lower maximum noble gas ratios ( 20 Ne/ 22 Ne of 12.2, 40 Ar/ 36 Ar of 16,100) and higher argon concentrations. Based on these results, along with Cl/K and H 2 O/Ce ratios and preliminary lead, strontium and neodymium isotope data, it is suggested that these two groups of samples reflect separate eruptive events that originated from distinct mantle sources. This provides evidence for upper mantle heterogeneity at small scales and could imply that the mantle source with lower neon and argon isotopic ratios has been more influenced by recycling of atmospheric noble gases.