S–Cl–F degassing pattern of water-rich alkali basalt: Modelling and relationship with eruption styles on Mount Etna volcano

Abstract

Listen

Translate article

Our knowledge of the degassing pattern of sulphur, chlorine and fluorine during ascent and eruption of basaltic magmas is still fragmental and mainly limited to water-poor basalts. Here we model and discuss the pressure-related degassing behaviour of S, Cl and F during ascent, differentiation and extrusion of H 2O–CO 2-rich alkali basalt on Mount Etna (Sicily) as a function of eruptive styles. Our modelling is based on published and new melt inclusion data for dissolved volatiles (CO 2, H 2O, S, Cl, F) in quenched explosive products from both central conduit (1989–2001) and lateral dyke (2001 and 2002) eruptions. Pressures are obtained from the dissolved H 2O and CO 2 concentrations, and vapour–melt partition coefficients of S, Cl and F are derived from best fitting of melt inclusion data for each step of magma evolution. This allows us to compute the compositional evolution of the gas phase during either open or closed system degassing and to compare it with the measured composition of emitted gases. We find that sulphur, chlorine and fluorine begin to exsolve at respective pressures of ∼140 MPa, ∼100 MPa and ≤ 10 MPa during Etna basalt ascent and are respectively degassed at > 95%, 22–55%, and ∼15% upon eruption. Pure open system degassing fails to explain gas compositions measured during either lateral dyke or central conduit eruptions. Instead, closed-system ascent and eruption of the volatile-rich basaltic melt well accounts for the time-averaged gas composition measured during 2002-type lateral dyke eruptions (S/Cl molar ratio of 5 ± 1, 35% bulk Cl loss). Extensive magma fragmentation during the most energetic fountaining phases enhances Cl release (55%) and produces a lower S/Cl ratio of 3.7, as actually measured. Comparatively slower magma rise in the central conduits of Etna favours both sulphide saturation of the melt and greater chlorine release (55%), resulting in a distinct S/Cl evolution path and final ratio in eruptive gas. In both eruption types, any previous bubble–melt separation at depth leads to increased S/Cl and S/F ratios in emitted gas. High S/Cl ratios measured during some discrete eruptive events can thus be explained by transitions from closed (deep) to open (shallow) system degassing, with differential gas transfer extending down to ∼2 km depth below the vents. This depth coincides with the base of the volcanic pile where structural discontinuities and the high magma vesicularity (60%) may favour separate gas flow. Finally, the excess S–Cl–F gas discharge through Etna summit craters during non-eruptive periods requires a mixed supply from shallow magma degassing in the volcanic conduits and deeper-derived SO 2-rich bubbles from the sub-volcano plumbing system. Our modelling provides a useful reference framework for interpreting the monitored variations of S, Cl and F in Mount Etna gas emissions as a function of volcanic activity. More broadly, the observations made for S, Cl and F degassing on Etna may apply to other basaltic volcanoes with water-rich magmas, such as in arcs.