V14C-07
Experimental determination of dissolved CO2 content in nominally anhydrous andesitic melts at graphite/diamond saturation – Remobilization of deeply subducted reduced carbon via partial melts of MORB-like eclogite

Monday, 14 December 2015: 17:30
103 (Moscone South)
James Eguchi and Rajdeep Dasgupta, Rice University, Houston, TX, United States
Abstract:
Experimental phase relations of carbonated lithologies [1] and geochemistry of deep diamonds [2] suggest that deep recycling of carbon has likely been efficient for a significant portion of Earth’s history. Both carbonates and organic carbon subduct into the mantle, but with gradual decrease of fO2 with depth [3] most carbon in deep mantle rocks including eclogite could be diamond/graphite [4]. Previous studies investigated the transfer of CO2 from subducted eclogite to the ambient mantle by partial melting in the presence of carbonates, i.e., by generation of carbonate-rich melts [5]. However, the transfer of carbon from subducted eclogite to the mantle can also happen, perhaps more commonly, by extraction of silicate partial melt in the presence of reduced carbon; yet, CO2 solubility in eclogite-derived andesitic melt at graphite/diamond saturation remains unconstrained. CO2content of eclogite melts is also critical as geochemistry of many ocean island basalts suggest the presence of C and eclogite in their source regions [6].

In the present study we determine CO2 concentration in a model andesitic melt [7] at graphite/diamond saturation at conditions relevant for partial melting of eclogite in the convecting upper mantle. Piston cylinder and multi anvil experiments were conducted at 1-6 GPa and 1375-1550 °C using Pt/Gr double capsules. Oxygen fugacity was monitored with Pt-Fe sensors in the starting mix. Completed experiments at 1-3 GPa show that CO2 concentration increases with increasing P, T, and fO2 up to ~0.3 wt%. Results were used to develop empirical and thermodynamic models to predict CO2 concentration in partial melts of graphite saturated eclogite. This allowed us to quantify the extent to which CO2 can mobilize from eclogitic heterogeneities at graphite/diamond saturated conditions. With estimates of eclogite contribution to erupted basaltic lavas, the models developed here allow us to put constraints on the flux of COto mantle source regions coming from subducted crust and investigate the possible role this process may play in the deep carbon cycle.

[1] Dasgupta (2013) RiMG. [2] Shirey, et al. (2013) RiMG. [3] Frost & McCammon (2008) Ann Rev Earth Plan Sci. [4] Stagno, et al. (2015) CMP. [5] Kiseeva, et al. (2012) JPet. [6] Mallik & Dasgupta (2014) G3. [7] Spandler, et al. (2008) JPet.