HR: 0800h
AN: T41E-1355    [Abstracts]
TI: Are Ultramafic `Assimilants' Modifying MORB?
AU: * Collier, M L
EM: collier@ldeo.columbia.edu
AF: Lamont-Doherty Earth Observatory, 61 Route 9W, Palisades, NY 10964 United States
AU: Kelemen, P B
EM: peterk@ldeo.columbia.edu
AF: Lamont-Doherty Earth Observatory, 61 Route 9W, Palisades, NY 10964 United States
AB: Observed systematics in fractionation-corrected major element compositions of Mid-Ocean Ridge Basalt (MORB) samples should, in theory, provide information about the composition of the mantle as well as the suite of melt generation, migration and evolution processes that are active below mid-ocean ridges. Na8 (Na2O extrapolated to 8 wt percent MgO to account for low-pressure fractionation, [1]) is negatively correlated with Fe8 in the global dataset, although this `global trend' shows significant variance. Datasets limited to specific MOR segments plot as one of several characteristic `local trends' which often differ from the global trend. The full variability of Na8 in the global distribution occurs at slow spreading ridges, raising the question of whether Na8 is influenced by factors related to the spreading rate, such as the thickness of the thermal boundary layer beneath the ridge. Previous work has sought to understand these observations as effects of lateral variations in mantle potential temperature, differences within local melting regimes, chemical heterogeneity in the parent mantle, reaction between ascending melt and residual peridotite, and systematic errors in the fractionation correction [e.g. 2-4]. Inspired by observations of plagioclase lherzolite produced by 0.6 GPa reaction between melt and peridotite along the Mid-Atlantic Ridge [5], we explore the possibility that the major element composition of ascending, primitive MORB is significantly affected by reactions with residual peridotite, and quantify the expected `signature' of such reactions. In particular, we consider the fate of a basaltic magma that is transported through the melting region in unreactive dunites or cracks [6,7], but reacts with peridotite near the base of the thermal boundary layer due to low permeability caused by crystallization during conductive cooling [e.g. 8,9]. Using the pMELTS program [10], we simulate this scenario by calculating the equilibrium composition of cooling basaltic melt reacting with an ultramafic `assimilant'. We study the effect of varying assimilation `rates' (gm per deg C), pressures, initial magma compositions, and assimilant compositions on melt compositions produced by our scenario, and compare these to observed trends for fractionation corrected MORB. Although this is an oversimplified model, we find that a single initial melt reacting with a given peridotite, over a range of assimilation rates and pressures, can generate the full range of Na8 and Fe8 values in the global dataset, including a trend that mirrors the global Na-Fe negative correlation. 1. Klein and Langmuir JGR87 2. Langmuir et al AGU92 3. Shen and Forsyth JGR95 4. Gaetani et al JGR95 5. Kelemen et al ODP Leg 209 IR04 6. Nicolas JPet86 7. Kelemen et al Nature95 8. Sparks and Parmentier EPSL91 9. Kelemen and Aharonov AGU98 10. Ghiorso et al G3 02
DE: 1009 Geochemical modeling (3610, 8410)
DE: 1032 Mid-oceanic ridge processes (3614, 8416)
DE: 1065 Major and trace element geochemistry
SC: Tectonophysics [T]
MN: Fall Meeting 2005