V31D-3049
Revised Mossbauer Calibration for Fe3+/FeT of XANES Basalt Standards: Implications for MORB

Wednesday, 16 December 2015
Poster Hall (Moscone South)
Marc M Hirschmann1, Hongluo Zhang1 and Elizabeth Cottrell2, (1)Univ Minnesota, Minneapolis, MN, United States, (2)Smithsonian, NMNH, Washington, DC, United States
Abstract:
Among techniques for determining Fe3+/FeT of natural glasses, XANES affords high precision, spatial resolution, and sample throughput and consequently has become widely used. However, because XANES determinations depend on standardization against materials of known Fe3+/FeT, they are only as accurate as the methods used for calibration. In many cases, calibration is performed with Mossbauer spectroscopy. Accurate determination of Fe3+/FeT by Mossbauer spectroscopy is the subject of a long-standing controversy, in part owing to debate as to the influence of recoilless fraction on the area ratios of room temperature (RT) Mossbauer absorption doublets associated with paramagnetic Fe2+ and Fe3+ in silicate glasses. Recoilless fraction effects for glasses are comparatively subtle, and so characterization efforts have not always produced statistically resolvable effects, in part because glasses produce broadened line shapes that degrade analytical precision, but both theoretical considerations of bond strengths and abundant evidence from minerals demonstrate that RT Mossbauer analyses will overestimate Fe3+/FeTof Fe-bearing silicates.

Cottrell & Kelley (2011) used the basalt XANES calibration of Cottrell et al. (2009) to show that the average Fe3+/FeT of MORB glasses is close to 0.16, but this calibration depends chiefly on RT Mossbauer spectra of a suite of standard glasses. New cryogenic (10 K) Mossbauer spectra of these same glasses suggests a correction factor, C, of 1.1, where [Fe3+/FeT(corrected)]=[Fe3+/FeT (RT)]/([Fe3+/FeT (RT)]+C(1-[Fe3+/FeT(RT)]). If this correction is applied to the XANES data, the median of MORB glasses is ~0.15. Because recoilless fractions are not exactly unity even at 10 K, this correction represents a minimum; however the 10 K data are in good agreement with Fe3+/FeT predicted for the Mossbauer standards by the wet-chemistry-based model of Kress and Carmichael (1991) based on the synthesis fO2. The precision of XANES for the determination of Fe3+/FeT is ±0.005; the absolute accuracy is tied to the accuracy of the technique used to calibrate. For Mossbauer, accuracy is likely closer to ±0.02