GSA Annual Meeting in Seattle, Washington, USA - 2017

Paper No. 207-6
Presentation Time: 9:25 AM

EXPERIMENTALLY CONSTRAINED H AND O ISOTOPE FRACTIONATION BETWEEN WATER AND VOLCANIC GLASS


HUDAK, Michael R.1, BINDEMAN, Ilya N.1 and GUAN, Yunbin2, (1)Earth Sciences, University of Oregon, Eugene, OR 97403, (2)Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA 91125, mhudak2@uoregon.edu

Hydrogen isotopes in hydrous volcanic glass are a promising paleoclimate proxy. The majority of water, and therefore hydrogen, is from secondary hydration by meteoric water. During and immediately following an eruption, the glass degasses and become nearly anhydrous. Meteoric water H and O isotopes are a function of climate and are imparted to the glass as hydration progresses. The relative mass difference between H and D is the greatest among isotopes and may result in large kinetic fractionation during geologic processes. Constraining the H isotope fractionation between water and volcanic glass will improve the accuracy and utility of this paleoclimate proxy. Isothermal vapor hydration experiments were conducted with isotopically labeled water (+57‰ δD, -0.17‰ δ18O) and obsidian from the Big Obsidian Flow at Newberry volcano in Oregon (0.07wt% H2O, -110‰ δD) between 175°C and 375°C in a muffle furnace. The starting glass particles were analyzed on a particle size analyzer for the grain size distribution. Sample were suspended in Ag capsules above 0.36 mL of water (effectively an infinite reservoir) in airtight stainless steel vessels. Run products from hour to month long experiments were analyzed on a thermal conversion elemental analyzer (TCEA) for bulk water content, δD, and δ18O of water in glass. Time series of experimentally hydrated glass show rapid initial increases in δD that slow and asymptotically approach the δD value of the water. Total water concentrations increase more gradually at a nearly constant rate through time. Relative water concentrations across the ~30 μm of the hydration rind were imaged with nanoSIMS and used for mass balance calculations using a spherical approximation of the average particle size along with bulk water and δD measurements from the TCEA analyses. These methods test if the fractionation between water and glass with respect to δD and δ18O (rapid extraction) is equilibrium, involving atom for atom exchange, or kinetic and controlled by diffusion. As D diffuses into more of the glass volume, the observed bulk fractionation is reduced, but hydration is still incomplete and experiments are ongoing. If fractionation is kinetic, volcanic glass may be hydrated rapidly at hydrothermal T and may record the unfractionated composition of the meteoric waters at the time of the eruption.