V43C-3161
SIMS Calibration of Nitrogen in Silicate Glasses and Applications to Melt Inclusions

Thursday, 17 December 2015
Poster Hall (Moscone South)
Margo Elaine Regier1, Richard L Hervig1, Paul J Wallace2, Madison Myers2 and Colin J N Wilson3, (1)Arizona State University, Tempe, AZ, United States, (2)University of Oregon, Eugene, OR, United States, (3)Victoria University of Wellington, Wellington, New Zealand
Abstract:
Previous attempts to constrain N fluxes between the mantle, crust, and atmospheric reservoirs over geologic time have been inhibited due to the difficulty of detecting trace amounts of nitrogen, especially as triply-bonded N2 in oxidized magmas. Secondary ion mass spectrometry (SIMS) can detect nitrogen as N+, and as the molecular ions SiN-, NO-, and CN-. However, there are few matrix-matched, bulk-analyzed standards for N, which makes quantification of its signal challenging. Here, we use the implantation of known amounts of nitrogen into rhyolitic glasses containing a range of H2O, and subsequent depth-profile analysis of these samples to derive N calibration factors1,2. We use a primary beam of O- and detection of N+ ions. Results demonstrate that the useful yield (ions detected per atom sputtered) of N increases with H2O content, whereas the useful yield of the measured matrix ion (30Si) decreases with higher H2O. Quantification is also complicated by variable N background that scales inversely with observed higher sputtering rates in hydrated glasses. The former issue can be resolved by fitting a curve to the H2O vs. calibration factor plot and the latter by varying the primary beam density on low-H2O materials. Together, these approaches allow us to quantify the N content in variably hydrated rhyolitic glasses. Application of these calibrations to quartz-hosted melt inclusions from the Bishop3, 4 and Huckleberry Ridge Tuffs4 reveals un-degassed N contents of melts at depth. We show that N can be used in concert with other volatiles to tease out magmatic processes, such as recharge events and magma mixing. We conclude that unless these inclusions do not represent the bulk concentration of N in the melt, large silicic eruptions have not released enough N to significantly impact the atmospheric reservoir over geologic time.

1Burnett DS et al. (2015) Geostand Geoanalytical Res 39:265-276; 2Wilson RG et al. (1989) Secondary ion mass spectrometry. Wiley (New York); 3Dunbar NW, Hervig RL (1992) J Geophys Res 97:15129-15150; 4Myers ML et al. (2015) AGU Fall Mtg V018 presentation.