Aqueous fluids and hydrous melts in high-pressure and ultra-high pressure rocks: Implications for element transfer in subduction zones

High-pressure (HP) and ultra-high pressure (URP) terranes are excellent natural laboratories to study subduction-zone processes. In this paper we give a brief theoretical background and we review experimental data and observations in natural rocks that constrain the nature and composition of the fluid phase present in HP and UHP rocks. We argue that a fluid buffered by a solid residue is compositionally well defined and is either an aqueous fluid (total amount of dissolved solids < 30 wt.%) or a hydrous melt (H2O < 35 wt.%). There is only a small temperature range of approximately 50-100 degrees C, where transitional solute-rich fluids exist. A review of available experimental data suggest that in felsic rocks the second critical endpoint is situated at 25-35 kbar and similar to 700 degrees C and hence must be considered in the study of UHP rocks. Despite this, the nature of the fluid phase can be constrained by relating the peak metamorphic conditions of rocks to the position of the wet solidus even if the peak pressure exceeds the pressure where the wet solidus terminates at the second critical endpoint. Transitional solute-rich fluids are expected in URP terrains (P > 30 kbar) with peak temperatures of about 700 +/- 50 degrees C. At higher temperatures, hydrous granitic melts occur whereas at lower temperatures aqueous fluids coexists with eclogite-facies minerals. This argument is complemented by evidence on the nature of the fluid phase from high-pressure terrains. We show that in the diamond-bearing, high-temperature UHP rocks from the Kokchetav Massif there are not only hydrous felsic melts, but probably also carbonate and sulfide melts present. Hydrous quartzo-feldspathic melts are mainly produced in high temperature UHP rocks and their composition is relatively well constrained from experiments and natural rocks. In contrast, constraining the composition of aqueous fluids is more problematic. The combined evidence from experiments and natural rocks indicates that aqueous fluids liberated at the blueschist to eclogite facies transition are dilute. They contain only moderate amounts of LILE, Sr and Pb and do not transport significant amounts of key trace elements such as LREE, U and Th. This indicates that there is a decoupling of water and trace element release in subducted oceanic crust and that aqueous fluids are unable to enrich the mantle wedge significantly. Instead we propose that fluid-present melting in the sediments on top of the slab is required to transfer significant amounts of trace elements from the slab to the mantle wedge. For such a process to be efficient, top slab temperature must be at least 700-750 degrees C at sub-are depth. Slab melting is likely to be triggered by fluids that derive from dehydration of mafic and ultramafic rocks in colder (deeper) portions of the slab. (c) 2006 Elsevier B.V. All rights reserved.