Origin of primitive andesites by melt-rock reaction in the sub-arc mantle (Invited)
Abstract
The genetic relationship between primitive granitoids, including high-Mg andesites (HMAs) and bajaites, and primary granitoids, or "pristine" adakites, has been vigorously debated since Defant and Drummond (1991; henceforth D&D) first applied the term "adakite" to refer to Cenozoic arc magmas (andesites and dacites) "associated with young subducting lithosphere", with low Y and Yb, low high-field strength elements (HFSEs), high Sr, and high Sr/Y and (La/Yb)N ratios "relative to island arc andesite-dacite-rhyolite". These characteristics were attributed to an origin for adakites by partial melting of basaltic crust within the subducting slab (hence "slab melts"). That such a process can produce melts with the characteristics described by D&D has since been largely confirmed by dehydration melting experiments on hydrous metabasalt at ~1-4 GPa. Attention was also drawn to the geochemical similarities between "adakites" and large-ion lithophile element (LILE)-enriched, high-field strength element (HFSE) depleted magnesian andesites (HMAs) from Adak Island in the western Aleutians, first described by Kay (1978), implying a genetic relationship between primary granitoid (adakites) formed by partial melting of basaltic ocean crust in the subducting slab, transformed to garnet-amphibolite or eclogite, and primitive magnesian andesites (HMAs) with high Mg-numbers (Mg# = molar Mg/(Mg+Fe)x100) and high concentrations of Ni and Cr. What then is the true origin of these enigmatic arc magmas, with both crustal and mantle, derivative and primitive, geochemical signatures? Kay (1978) suggested a "hybrid" model, in which "hydrous melting of eclogite (slab melting) results in a small volume of dacitic melt ("pristine adakite"), which rises into the hotter overlying peridotite wedge and equilibrates with olivine and orthopyroxene, reacting with olivine until it becomes andesitic". In this paper, I will discuss the results of melt-rock reaction experiments modelling this peridotite assimilation-melt metasomatism process, in which "pristine" adakite melts are reacted with depleted peridotite over a range of temperatures (1100-1300°C), pressures (1.6-3.8 GPa), and melt:rock ratios. All natural starting materials were used, and the major- and trace-element characterisitics of the resulting mantle-hybridized melts and crystalline reaction residues have been determined by electron and laser-ablation ICPMS microprobes. In many cases, melts comparable to HMAs are formed, with crustal "adakitic" trace element signatures overprinted by "primitive", mantle-derived signatures (Mg# >60; high Ni and Cr). However, these "primitive" granitoids are not in equilibrium with an olivine-bearing, peridotitic reaction assemblage, but an olivine-free, pyroxenitic (garnet clinopyroxenite or garnet websterite) assemblage. As such, HMAs cannot be said to be crustal or mantle in origin, but of hybrid pedigree, with their trace element signature adakite-derived, but their primitive charcter mantle-derived. REFERENCES Defant, M.J., & Drummond, M.S. (1991) Derivation of some modern arc magmas by melting of young subducted lithosphere. Nature347, 662-665 Kay, R.W. (1978) Aleutian magnesian andesites: melts from subducted Pacific ocean crust. Jour. Volcan. Geotherm. Res. 4, 117-132.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2009
- Bibcode:
- 2009AGUFM.V34A..05R
- Keywords:
-
- 1012 GEOCHEMISTRY / Reactions and phase equilibria;
- 1020 GEOCHEMISTRY / Composition of the continental crust;
- 1031 GEOCHEMISTRY / Subduction zone processes;
- 3630 MINERALOGY AND PETROLOGY / Experimental mineralogy and petrology