Alkali basalts and related rocks (basanites, nephelinites) are common products of magmatic processes beneath oceanic islands and seamounts, but their origin is poorly understood. Because such rocks hold key information about geochemical and geodynamical processes in Earth's mantle, constraints on their genesis are critical to our broader understanding of the interior of the planet. Presently, there remains considerable debate and uncertainty regarding the thermal vigor of the mantle sources of oceanic island basalts (OIB) and the possibility that recycled materials - ancient crust that has been returned to the mantle via subduction, as well as volatile components such as H2O and CO2 - was involved in their genesis. Resolving these debates requires constraints on the relationship between the compositions of magmas and the temperatures, pressures, and compositions of their sources.
This proposal will support an extension of recent experimental work by the team that has shown that small-degree partial melts of 'typical' garnet peridotite - the material thought to constitute most of the upper mantle - cannot account for the compositions of oceanic island basalts. This means either that experiments have not been conducted under the right conditions or that an additional (pyroxenitic or metasomatized peridotitic) lithology is required in the source of OIB. They will continue to use an iterative experimental method ('MISE'), developed and improved over the last funding period, to determine the composition of small degree partial melts of garnet lherzolite and to explore the effects of pressure, melt fraction, and source enrichment in CO2, K2O, H2O and FeO on the compositions of near-solidus partial melts. These new experiments will allow for a better determination of the compositions and proportions of liquids generated during incipient partial melting of garnet peridotite. If alkali basalts originate from small-degree melting of garnet peridotite, and if the temperature and pressure of melting can be constrained by major element chemistry of plausible parental liquids, the experiments will relate the petrologic character of OIB to the depth of melting and mantle potential temperature in their source. A second set of experiments will be aimed at understanding the sources and conditions of melting of basalts from oceanic island based on their trace element compositions. In particular, first row transition elements (FRTE) have emerged as important probes of the origins of OIB. Concentrations of Ni and Ti as well as Mn/Fe and Zn/Fe ratios are distinct in OIB as compared to MORB and may require contributions from pyroxenite or from compositionally modified peridotite. Consequently, it is proposed to measure partition coefficients between garnet lherzolite minerals (olivine, pyroxenes, and garnet) of FRTE from small degree partial melting experiments and from crystallization experiments at 3-5 GPa.
This project used experiments at high pressure and temperature to understand the origin of basalt lavas that erupt at oceanic islands such as Hawaii, Iceland, or other localities around the world. It was already known that such magmas originate by partial melting of rocks at depths in the mantle of approximately 100 km, but the types of rocks in these deep sources was not well constrained. Whereas most of the rocks in these regions are believed to be garnet peridotite, considerable evidence suggested that the places that partially melt to form such basalts are in part comprised of other less-common rock types. however, the data to constrain this hypothesis was not available because experiments creating partial melts of garnet peridotite had not been performed at the right conditions. Most importantly, the melts that likely form in such regions are *incipient* partial melts, meaning only a very small fraction of the source rock is turned to melt. Previously, there were no good methods for determining the compositions of such incipient partial melts. We applied a specialized technique developed in our lab that allowed us to do so, and thereby tested the hypothesis that partial melts of "normal" mantle - garnet peridotite - create magmas similar to those that erupt at oceanic islands. We found that there are many similarities between the experimental melts and the compositions of natural lavas, but that there are also crucical differences that seemingly require the participation of more exotic source rocks. These exotic source rocks originate by processes in the mantle that include the large-scale plate tectonic mixing of ancient surface rocks, and so are clues to the long-term evoultion and history of Earth's interior.
