Fluorine and Chlorine in mantle: are they useful volatile tracers?

Fluorine and chlorine are commonly considered together with hydrogen, carbon, and sulfur as volatile elements in volcanic system because molecular species of these elements are found in volcanic gasses. However, fluorine and chlorine do not correlate with H₂O and CO₂among degassed volcanic glasses, suggesting these halogens are inefficiently degassed, and fractionated during magma ascent. Because Cl is highly soluble in aqueous fluid, and F commonly substitutes in the OH site of hydroxyl-bearing minerals (such as amphiboles, micas, humites, and apatite), they can be ideal geochemical tracers recording fractionation due to aqueous fluid, melt, and/or supercritical fluid in the mantle.

In this contribution, we summarize recent experimental results, natural observations, and illustrate prospect of these light halogen as geochemical tracers. Notably, recent partitioning experiments report complex behavior of F. For example, F is incompatible in anhydrous mantle minerals (olivine, pyroxene, and garnet) against aqueous fluid and silicate melts, while F can be compatible in amphibole, and mica against aqueous fluid and/or silicate melt. Furthermore, some studies indicate that F fractionation between liquid and solid is strongly influenced by the nature of the liquid. For all conditions mentioned here, Cl is systematically more incompatible than F. These data provide powerful bases to interpret geochemical correlations found in least degassed basaltic glasses, mainly from olivine-hosted inclusions, of arc, MORB and OIB. In some cases we have a strong indication showing amphibole-aqueous fluid fractionation beneath arc. Lastly, F measurements in oceanic crust and Alpine ophiolites suggest lithology-P-T dependent intake/release of these elements, adding a layer of complexity to global halogen cycles.