Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
4700249 | Chemical Geology | 2009 | 11 Pages |
The currently available data set of S–Cl–F abundances in volcanic gas plumes and high-temperature fumarolic gas samples from basaltic volcanism is reviewed here in the attempt to derive constraints on the modes of halogen degassing from mafic silicate melts. Apart from large volcano-to-volcano variations, reflecting remarkable differences in volatile abundances in the source magmas, each of the explored volcanoes displays large changes of SO2/HCl and SO2/HF ratios with the style of volcanic activity, with HCl/HF staying fairly constant. Halogen abundances are low and SO2/HCl and SO2/HF are high when fresh (volatile-rich) magmas sustain degassing, as during explosive eruptions, at the onset of eruptive cycles, or shortly before paroxysmal events. Low SO2/HCl and SO2/HF ratios are instead characteristic of late stages of volcanic degassing, typically being observed in the concluding stages of basaltic eruptions, or during periods of reduced magma supply at persistently degassing volcanoes. These observations are taken as evidence of halogens being less keen to enter the gas phase (relative to S) during degassing of basaltic magmas; and quantitatively interpreted in light of a Rayleigh-type open-system degassing model. The model, though simple, quantitatively reproduces the range of volcanic gas compositions observed at basaltic volcanoes worldwide, and allows prediction of vapour/melt partitioning contrasts of factors ~ 9 and ~ 36 for the volatile couples S–Cl and S–F, respectively. These predictions require validation from appropriately designed experiments of halogen partitioning between magmatic vapours and silicate melts over a range of P–T–X conditions.