Author(s): Shan Gao, Eiichi Takahashi, Toshihiro Suzuki
In order to understand the melting processes that occur within recycled oceanic crust and mantle in a heterogeneous plume (e.g., that beneath the Hawaiian Islands), a series of high-pressure-high-temperature layered experiments were performed at 2.9 GPa, 5 GPa, and 8 GPa, from 1300°C to 1650°C, using a fertile peridotite KLB-1 and N-MORB. Our experiments at conditions below the dry peridotite solidus produced melt compositions that ranged from basaltic andesite to tholeiite. An Opx reaction band formed between eclogite and peridotite layers, likely via chemical reaction between a silica-rich eclogite-derived partial melt and olivine in the peridotite matrix. At temperatures at or above the dry peridotite solidus, substantial melting occurred in both basalt and peridotite layers, and fully molten basalt melt and melt pockets from the peridotite layer combined. In our layered experiments, major and minor element contents in reacted melts closely matched those of Hawaiian tholeiite and picrite, except for Fe. Partial melts of anhydrous run products had ~55 – 57 wt% SiO2 at low temperature (i.e., were andesitic) and had ~50 – 53 wt% SiO2 at high temperatures, slightly below the dry peridotite solidus (i.e., were tholeiitic, and similar to those that occur during the Hawaii shield-building stage). Based on the Fe- and LREE-enriched signature in Hawaiian tholeiites, we propose that recycled components in the Hawaiian plume are not modern N-MORB, but are Fe-rich tholeiite; a lithology that was common in the Archaean and early Proterozoic. We have demonstrated that the entire compositional spectrum of Hawaiian tholeiites (basalt to picrite) can be formed by basalt-peridotite reactive melting near the dry solidus of peridotite. Based on these results, we propose that the potential temperature of the sub-Hawaiian plume may be much lower than previously estimated.
See also: Comments to Paper