Disequilibrium degassing model determination of the 3He concentration and 3He/22Ne of the MORB and OIB mantle sources

Abstract

Models of the dynamics of eruptive degassing provide a unique insight into the pre-eruptive concentrations of the major volatiles (CO2, H2O), noble gases, and the mantle reservoirs supplying them: a fundamental component in describing the accretion and evolution of the Earth. We investigate and develop a disequilibrium degassing model, exploring the parameters required to reproduce noble gas compositions observed in two ocean island basalt (OIB) and one mid-ocean ridge basalt (MORB) sample suites from the East Pacific Rise (EPR), Loihi Seamount, and Iceland. The original model assumed an identical loss of major volatile components for each degassing step (Gonnermann and Mukhopadhyay, 2007). We recalculate the major volatile vapor phase composition for each degassing step, taking account of the degassing history over previous steps. Final noble gas elemental ratios, using the same eruption parameters, can differ by orders of magnitude from the original model's calculations. We further adapt our model variant to take into account both decompression-driven degassing during magma ascent and degassing at constant pressure during sample quenching. Our results show that elemental ratios and noble gas concentrations can be effectively decoupled by the different degassing stages of an eruption. Ascent rate determines whether degassing during magma ascent is modeled as predominantly closed or open system degassing. The closed system conditions that dominate the slowly ascending MORB result in a less dramatic decrease in degassed elemental ratios than for the OIB models, consistent with the observed data. The MORB model also constrains the initial MORB source melt 3He/22Ne ratio, allowing that the MORB mantle could have a ratio as low as the OIB ratio, a feature required by steady state mantle models. The two OIB sample suites show very similar noble gas ratios and concentrations despite a large difference in final eruption pressures. We propose that this can fit our model if the effect of gas loss by quenching is negligible. Our model then shows that only a small amount of degassing takes place during magma ascent, meaning that initial OIB He melt concentrations must be similar to or lower than MORB source melt concentrations, although greater initial CO2 concentrations are required. This result is entirely consistent with mantle models which see a significant recycled component in the OIB-source mantle: such material would be expected to have higher CO2 but lower 3He concentrations than the depleted mantle.