Experimental Determination of Hydrogen Partitioning in an Apatite-Haplobasalt Andesite Melt System at 1200°C and 1 GPa

Igneous apatite crystals contain a record of the evolving concentrations of magmatic volatile elements such as F, H, and Cl during crystallization and degassing. In order to develop apatite as a quantitative barometer of magmatic volatile species, improved understanding of apatite-melt volatile equilibria is needed. We conducted apatite crystallization experiments using a piston cylinder apparatus at 1,200°C and 1 GPa. Apatite crystals were grown from an Fe- and Cl-free, haplobasaltic-andesite melt of constant composition. Experiments were quenched to below 700°C in 15 seconds, producing homogeneous glass plus apatite crystals ranging from 5 to 30 microns in diameter. Apatite and coexisting glass were analyzed using the Cameca ims 6f SIMS, with an O- beam and positive secondary ions, following the methods of Boyce and Hervig (2008, 2009), but using the inversion-based calibration routines of Boyce and Eiler (this meeting). These measurements provide the first direct constraint on the partitioning of hydrogen between apatite and melt. For the exchange reaction F-(melt) + OHapatite = OH-(melt) + Fapatite, the Keq was calculated from the measured data to be ~25, consistent with the results of Mathez and Webster (2004) and Webster et al. (2008), despite our experiments being nominally Cl- and Fe-free. Significant non-Henrian behavior is observed for hydrogen, but trace F behaves effectively Henrian at low concentrations, again consistent with previous work. These preliminary results reinforce the assertion that simple Henry's or Raoult's Law models of volatile partitioning between apatite and melt are rarely - if ever - appropriate, and that both F and H concentrations must be considered in order to use hydrogen concentrations in apatite to determine concentrations of hydrogen in the coexisting melt.