Modes of planetary-scale Fe isotope fractionation

A comprehensive set of high-precision Fe isotope data for the principle meteorite types and silicate reservoirs of the Earth is used to investigate iron isotope fractionation at inter- and intra-planetary scales. 14 chondrite analyses yield a homogeneous Fe isotope composition with an average delta Fe-56/Fe-54 value of -0.015 +/- 0.020 parts per thousand (2 SE) relative to the international iron standard IRMM-014. Eight non-cumulate and polymict eucrite meteorites that sample the silicate portion of the HED (howardite-eucrite-diogenite) parent body yield an average delta Fe-56/Fe-54 value of -0.001 +/- 0.017 parts per thousand, indistinguishable to the chondritic Fe isotope composition. Fe isotope ratios that are indistinguishable to the chondritic value have also been published for SNC meteorites. This inner-solar system homogeneity in Fe isotopes suggests that planetary accretion itself did not significantly fractionate iron. Nine mantle xenoliths yield a 2 sigma envelope of -0.13 parts per thousand to +0.09 parts per thousand in delta Fe-56/Fe-54. Using this range as proxy for the bulk silicate Earth in a mass balance model places the Fe isotope composition of the outer liquid core that contains ca. 83% of Earth's total iron to within +/- 0.020 parts per thousand of the chondritic delta Fe-56/Fe-54 value. These calculations allow to interprete magmatic iron meteorites (delta Fe-56/Fe-54 = + 0.047 +/- 0.016 parts per thousand; N = 8) to be representative for the Earth's inner metallic core. Eight terrestrial basalt samples yield a homogeneous Fe isotope composition with an average delta Fe-56/Fe-54 value of +0.072 +/- 0.016 parts per thousand. The observation that terrestrial basalts appear to be slightly heavier than mantle xenoliths and that thus partial mantle melting preferentially transfers heavy iron into the melt [S. Weyer, A.D. Anbar, G.P. Brey, C. Munker, K. Mezger and A.B. Woodland, Iron isotope fractionation during planetary differentiation, Earth and Planetary Science Letters 240(2), 251-264, 2005.] is intriguing, but also raises some important questions: first it is questionable whether the Fe isotope composition of lithospheric mantle xenoliths are representative for an undisturbed melt source, and second, HED and SNC meteorites, representing melting products of 4Vesta and Mars silicate mantles would be expected to show a similar fractionation towards heavy isotope compositions. This is not observed. Four international granitoid standards with SiO2 contents between 60 and 70 wt.% yield delta Fe-56/Fe-54 values between 0.118 parts per thousand and 0.132 parts per thousand. An investigation of the alpine Bergell igneous rock suite revealed a positive correlation between Fe isotope compositions and SiO2 contents - from gabbros and tonalites (delta Fe-56/Fe-54 approximate to 0.03 to 0.09 parts per thousand) to granodiorites and silicic dykes (delta Fe-56/Fe-54 approximate to 0.14 to 0.23 parts per thousand). Although in this suite delta Fe-56/Fe-54 correlates with delta O-18 values and radiogenic isotopes, open-system behavior to explain the heavy iron is not undisputed. This is because an obvious assimilant with the required heavy Fe isotope composition has so far not been identified. Alternatively, the relatively heavy granite compositions might be obtained by fractional crystallisation of the melt. Ultimately, further detailed studies on natural rocks an the experimental determination of mineral/melt fractionation factors at magmatic conditions are required to unravel whether or not iron isotope fractionation takes place during partial mantle melting and crystal fractionation. (c) 2006 Elsevier B.V. All rights reserved.