OXYGEN-ISOTOPE FRACTIONATION IN THE MANTLE

Citation
Jm. Rosenbaum et al., OXYGEN-ISOTOPE FRACTIONATION IN THE MANTLE, Geochimica et cosmochimica acta, 58(21), 1994, pp. 4767-4777
Citations number
64
Categorie Soggetti
Geosciences, Interdisciplinary
ISSN journal
00167037
Volume
58
Issue
21
Year of publication
1994
Pages
4767 - 4777
Database
ISI
SICI code
0016-7037(1994)58:21<4767:OFITM>2.0.ZU;2-Q
Abstract
A three-component, diffusion-limited, open-system exchange model for p yroxene, olivine, and fluid is presented that reproduces the range of oxygen isotope compositions of coexisting silicates from mantle-derive d samples. Closed-system exchange cannot reproduce observed disequilib rium fractionations. Measured oxygen self-diffusion data were used to constrain fluid-silicate reaction rates. The magnitudes of fluid-miner al fractionations for typical mantle fluids are presented and show tha t equilibrated fluid will always be more O-18-rich than coexisting sil icates. Three conditions are necessary to generate the disequilibrium pyroxene-olivine oxygen isotope fractionations observed in the mantle dataset. There must be at least an order of magnitude difference betwe en the fluid-pyroxene and fluid-olivine reaction rates, fluid and soli d phases must be out of equilibrium prior to reaction, and fluid must be moving several orders of magnitude faster than it is reacting with the silicates. Assuming initial isotopic equilibrium between pyroxene and olivine and reasonable porosities, the following constraints apply . If pyroxene is reacting faster than olivine, to be expected when pyr oxene and olivine grain sizes are subequal, a fluid flow rate of at le ast 5 X 10(3) times the pyroxene/fluid reaction rate is needed to gene rate the most extreme negative pyroxene-olivine fractionations seen in natural samples. If olivine reacts faster, the case when its effectiv e grain size is <1% that of pyroxene, flow rates of 1 x 10(4) to 5 X 1 0(4) times the fluid/pyroxene reaction rate are required to have the s ame effect. The composition of the fluid phase and starting mantle nec essary to reproduce the mantle dataset are dependent on which silicate is the faster reactant. Using self-diffusion data to constrain reacti on rates, flow rates >0.2 cm/y at 1200 degrees C and >3 cm/y at 1400 d egrees C are needed when pyroxene is the faster reactant. Flow rates > 1 cm/y at 1200 degrees C and >15 cm/y at 1400 degrees C are necessary in the olivine dominated system. Regardless of which silicate reacts f aster, large amounts of fluid are required, with fluid/rock ratios app roaching unity, implying a fluid-focusing mechanism operating in the m antle source region of the xenoliths. Measurable, negative, disequilib rium pyroxene-olivine 18(O) fractionations are very short-lived at man tle conditions, surviving <10(5) y at temperatures above 1200 degrees C.