Key points are not available for this paper at this time.
We analysed forty-nine, hand-picked, mid-ocean ridge basalt (MORB) glasses for their Mg isotopic compositions using a high-precision, critical mixture double-spiking approach. Additionally, the Fe isotopic compositions of a subset (twenty-two) of these samples were measured. Samples from the East Pacific, Mid-Atlantic, South East Indian and East Scotia ridges have average δ26Mg ∼ −0.21 ‰ and δ57Fe ∼0.10 ‰, while those from the ultra-slow spreading Gakkel and South West Indian ridges have lighter Mg isotope compositions, δ26Mg as low as ∼ −0.32 ‰, and heavier Fe isotope compositions, δ57Fe up to 0.25 ‰. Overall, the samples show a striking negative correlation between δ26Mg and δ57Fe. Few MORB have δ26Mg as high as predicted by equilibrium models of melting and differentiation; in complementary fashion, measured δ57Fe are typically higher than modelled, equilibrium values. Furthermore, we show that the slope of 26Mg/24Mg and 25Mg/24Mg covariations in MORB is consistent with kinetic not equilibrium fractionation. These observations identify an important diffusive control on both Mg and Fe isotopic compositions of MORB. Fractional transport of melt beneath ridges juxtaposes low Mg/Fe fractional melts from depth with high Mg/Fe residues at the top of the melting column. We argue that diffusively limited, partial Mg-Fe exchange between melt and peridotite through which it migrates leads to melts becoming isotopically lighter in Mg and heavier in Fe. Low melt rock ratios from sub-adiabatic melting beneath the slowest spreading ridges result in the greatest diffusive exchange between peridotite and melt en route to surface.
Liu et al. (Sun,) studied this question.