MELT RE-FERTILIZATION IN HESS DEEP PERIDOTITES

Abstract

Abyssal peridotites drilled from Hess Deep, at the intersection of the East-Pacific Rise and the Galapagos spreading center, are mostly strongly depleted, with high Cr# in spinel, low modal abundances of CPX and very depleted pyroxene mineral compositions. They have very low Ti, Na and Al in both OPX and CPX relative to most abyssal peridotites and have very low rare earth element abundances. Modal data for this suite show that they are harzburgites, with olivine 80 ±3, OPX 18 ±2, CPX 2±1, and spinel ~1 volume per cent. (Dick and Natland, 1996). A subset of peridotites from the Hess Deep have been re-fertilized in some melt components, resulting in large enrichments in Na in CPX, and these samples, as well as some very depleted ‘typical’ Hess Deep peridotites, have been analyzed for trace element abundances in CPX by ion probe. Enrichment factors in the re-fertilized samples range up to 35x for Na, 5x for Ti, greater than 140x for Ce, 80x for Sm, 2.5x for Yb, >10 for Sr and 40x for Zr. All of our samples are spinel harzburgite, with no plagioclase or alteration after plagioclase recognized, though parts of these cores do contain plagioclasebearing dunite and harzburgite. Enrichments are extremely localized, with multiple thin sections taken from one rock sample showing different degrees of enrichment. The incompatible trace element enrichments are accompanied by only minor shifts in Mg# in pyroxenes and olivines, suggesting that these rocks interacted with melts much more enriched than the last melts that had been extracted from them during fractional melting, with relatively minor outright trapping of melt into these rocks. Rare earth element ratios, such as (Ce/Sm)n and (Sm/Yb)n, vary by a factor of ~30. Strong enrichments in light and medium rare earth elements (REEs) are accompanied by minor shifts in heavy REEs, producing the steep trend seen in Figure 1. Cr-spinels in most Hess Deep peridotites have Cr#’s at 51 to 55, but one of our samples with metasomatic enrichments has been shifted to a Cr # of 46.6. Ti in spinel in these samples shows only minor effects from the metasomatic interactions. Modal abundances of CPX remain quite low in the metasomatized samples from this suite, ranging from ~1 to ~3.5 %, with the highest value in the most enriched sample. Melt-rock reactions and melt-trapping clearly have led to substantial enrichments in various elements in some samples from Hess Deep, but it remains unclear to what extent these processes have effected abyssal peridotites worldwide. Figure 2 shows Na2O versus TiO2 in abyssal peridotites, along with a fractional melting curve, and it is clear that abyssal peridotite CPX compositions in the whole suite are not consistent with melting as the sole process that has determined their compositions. Incorporating dynamic melting models into the calculation does not change this conclusion. Nor does incorporating more complex functions for changing distribution coefficients with pressure. Models that attempt to explain the overall geochemistry of abyssal peridotites include porous reactive flow (Asimow, 1999) or melt trapping into more extensively melted peridotite (Elthon, 1992), but no models have yet fully explained the geochemistry of these rocks.