Subduction recycling of continental sediments and the origin of geochemically enriched reservoirs in the deep mantle: experimental constraints at 16-23 GPa

Robert P. Rapp 1,2 , Tetsuo Irifune 2 , Nobu Shimizu 3
1 Research School of Earth Sciences, Australian National University, Canberra, ACT 0200, Australia
2 Geodynamics Research Center, Ehime University, Matsuyama, 790-8577, Japan
2 Department of Geology and Geophysics, Woods Hole Oceanographic Institution, Woods Hole, MA, 02543, USA

Since Earth's formation more than 4.5 billion years ago, the continental crust has been progressively extracted from the silicate mantle by island arc magmatism associated with subduction of oceanic lithosphere along convergent plate boundaries. As a consequence of arc magmatism and continent formation, elements that are generally incompatible in mantle minerals (e.g., K, Rb, Sr, Ba, Th and U) during partial melting tend to be highly concentrated in the continental crust, and a complementary depletion in these elements (referred to as "large-ion lithophile elements", or LILEs) is observed in the upper mantle. Over geologic time, the continents have thus come to represent a geochemically enriched reservoir relative to the "depleted" upper mantle. Erosion of the continental masses deposits terrigenous sediments in the deep ocean basins, material that is eventually returned to the deep mantle by subduction recycling. Isotope and trace element geochemical features of ocean island basalts (OIBs) infer the presence of long-lived (~1-2 Ga old) compositional heterogeneities in their plume source in the deep mantle, and these are usually attributed to the recycling of continent-derived sediments via subduction. But a number of questions regarding crustal recycling in subduction zones remain unanswered: how deeply can continental sediments be subducted? do sediments retain their unique geochemical signature during transport through the upper mantle? What mineral phases control the distribution of LILEs in continental sediments at very high pressures? Are sedimentary rocks at high pressure more or less dense than ambient mantle (i.e., are they buoyant, or do they sink?)? In an effort to answer these questions, we have undertaken a high-pressure experimental study of the physical and chemical properties of sedimentary rocks at pressures corresponding to the lowermost upper mantle and uppermost lower mantle (i.e., at depths corresponding to the mantle transition zone, ~500-700 km). Phase equilibria experiments on natural metasedimentary rocks at 15-23 GPa indicate that the high pressure phase assemblage in continental sediments consists of stishovite, garnet, kyanite, corundum, and K-hollandite (KAlSi 3 O 8 ), and that K-hollandite retains most if not all of the bulk rock's budget of LILEs and heat-producing elements (i.e., K, U, Th) (Rapp et al., 2007, submitted). The bulk thermoelastic properties and equation-of-state parameters of K-hollandite have also been determined using in-situ synchrotron radiation at the Spring-8 light source in Japan (Nishiyama et al., 2005). From these measurements, we have estimated the density of continental sediments through the mantle transition zone and into the lower mantle, and our results suggest that terrigenous sediments achieve positive buoyancy at the top of the lower mantle (see Fig. 1). This further implies that OIBs with the isotopic and geochemical signature of recycled terrigenous sediments (e.g., EM-1 type) cannot originate from depths greater than approximately 800 km.

Figure 1. Density profile for model sediment composition (equal parts stishovite, K-hollandite, and garnet) based on high-pressure, in-situ equation of state measurements on K-hollandite, and literature values for stishovite and Mj-garnet. Shown for comparison is a 1D seismologically determined density profile for model mantle composition (ak135) from Kennett et al. (1995)

References: Kennett, B.L., Engdahl, E.R., and Buland, R. (1995) Constraints on seismic velocities in the Earth from travel times. Phys. Earth Planet. Int. 121 , 85-102.

Nishiyama,, N., Rapp, R.P., Irifune, T., Sanehira, T., Yamazaki, D., and Funakoshi, K. (2005) Stability and P-V-T equation of state of KAlSi3O3-hollandite determined by inn situ X-ray observations and implications for dynamics of subducted continental crust material. Physics Chem. Minerals 32 , 627-637.

Rapp, R.P., Irifune, T., Shimizu, N., Nishiyama, N., Norman, M.D., and Inoue, T. (2007) Subduction recycling of continental sediments and the origin of geochemically enriched reservoirs in the deep mantle. Earth Planet. Sci. Lett. (submitted).