HR: 09:00h
AN: V31G-05    [Abstracts]
TI: 40Ar/39Ar Dating of Volcanic Glass
AU: * Morgan, L E
EM: lmorgan@berkeley.edu
AF: University of California, Berkeley, Department of Earth & Planetary Science 307 McCone Hall #4767, Berkeley, CA 94720, United States
AU: Renne, P R
EM: prenne@bgc.org
AF: University of California, Berkeley, Department of Earth & Planetary Science 307 McCone Hall #4767, Berkeley, CA 94720, United States
AU: Renne, P R
EM: prenne@bgc.org
AF: Berkeley Geochronology Center, 2455 Ridge Road, Berkeley, CA 94709, United States
AU: Watkins, J M
EM: jwatkins@berkeley.edu
AF: University of California, Berkeley, Department of Earth & Planetary Science 307 McCone Hall #4767, Berkeley, CA 94720, United States
AB: Application of the ⁴⁰Ar/³⁹Ar method to volcanic glasses has been somewhat stigmatized following several studies demonstrating secondary mobility of K and Ar. Much of the stigma is unwarranted, however, since most studies only impugned the reliability of the K-Ar and ⁴⁰Ar/³⁹Ar techniques when applied to glass shards rather than obsidian clasts with low surface area to volume ratios. We provide further evidence for problematic K loss and/or ³⁹Ar recoil ejection from glass shards in ⁴⁰Ar/³⁹Ar step heating results for comagmatic feldspars and shards. In an extreme case, the plateau age of the feldspars (0.17 ± 0.03 Ma at 2σ) is significantly younger than the plateau age of the glass (0.85 ± 0.05 Ma at 2σ). If the feldspar age is reasonably interpreted as the eruption age of the ash, it is likely that the glass shards experienced K and/or ³⁹Ar loss. Electron microprobe analyses of the glass shards have low totals (~93%) and no systematic lateral variability (i.e., diffusion gradients) in K, suggesting that the lengthscale of the glass shards is smaller than the lengthscale of K diffusion. Obsidian clasts should not be as susceptible to K loss since any hydrated (K-depleted) volume represents a small fraction of the total material and can often be physically removed prior to analysis. Samples described here are detrital obsidian clasts from the Afar region of Ethiopia. Evidence from Fourier Transform Infrared Spectroscopy (FTIR), and previous work by Anovitz (1999), confirm that the scale of water and potassium mobility are often small in comparison to the size of obsidian clasts but large enough to effect the bulk composition of glass shards. This expectation is confirmed in another tuff wherein comagmatic obsidian clasts and sanidine phenocrysts yield indistinguishable ⁴⁰Ar/³⁹Ar ages of 4.4 Ma High abundances of non-radiogenic ⁴⁰Ar, and kinetic fractionation of Ar isotopes during quenching and/or laboratory degassing resulting in incomplete equilibration between atmospheric and magmatic argon, may also hinder accurate ⁴⁰Ar/³⁹Ar geochronology of volcanic glasses. Clasts derived from single flows (as determined by extrusion age and trace element geochemistry) display variations as much as 3-4 orders of magnitude in atmospheric ⁴⁰Ar concentrations. The clasts were likely sourced from different parts of the flows with varying proximity to the surface and thus differing thermal and atmospheric uptake histories. Because radiogenic and non-radiogenic components of ⁴⁰Ar are energetically indistinct in glass, most samples fail to yield isochrons due to limited range in ⁴⁰Ar: ³⁹Ar: ³⁶Ar. Most yield plateau ages, whose validity rests on the assumption of atmospheric initial ⁴⁰Ar/³⁶Ar. Some samples yield inverse isochrons with sub-atmospheric ⁴⁰Ar/³⁶Ar intercepts; unirradiated subsamples of these same samples also have sub-atmospheric ³⁸Ar/³⁶Ar ratios that are too high to be explained by mass fractionation. While the effects of non-radiogenic ⁴⁰Ar remain poorly understood, this sample set yielded 32 out of 41 clasts having plateau and isochron ages within 2σ error of each other, and potentially problematic samples are generally identified with non-atmospheric isochron ⁴⁰Ar/³⁶Ar intercepts.
DE: 1065 Major and trace element geochemistry
DE: 1105 Quaternary geochronology
DE: 1115 Radioisotope geochronology
DE: 1194 Instruments and techniques
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2007 Fall Meeting