A new hydrous Al-bearing pyroxene as a water carrier in subduction zones Mauro Gemmi a, b, ⁎, Johannes Fischer a , Marco Merlini a , Stefano Poli a , Patrizia Fumagalli a , Enrico Mugnaioli c , Ute Kolb c a Dipartimento di Scienze della Terra ‘A. Desio’, Università degli Studi di Milano, Italy b Center for Nanotechnology Innovation@NEST, Istituto Italiano di Tecnologia, Pisa, Italy c Institut für Physikalische Chemie, Johannes Gutenberg-Universität, Mainz, Germany abstract article info Article history: Received 15 April 2011 Received in revised form 19 July 2011 Accepted 11 August 2011 Available online xxxx Editor: L. Stixrude Keywords: subduction hydrous pyroxene precession electron diffraction electron diffraction tomography A new Hydrous Al-bearing PYroxene (HAPY) phase has been synthesized at 5.4 GPa, 720 °C in the MgO– Al 2 O 3 –SiO 2 –H 2 O model system. It has the composition Mg 2.1 Al 0.9 (OH) 2 Al 0.9 Si 1.1 O 6 , a C-centered monoclinic cell with a = 9.8827(2), b = 11.6254(2) c = 5.0828(1) Å and β = 111.07(1)°. The calculated density is 3.175 g/cm 3 and the water content is 6.9% H 2 O by weight. Its structure has been solved in space group C2/c by the recently developed automated electron diffraction tomography method and refined by synchro- tron X-ray powder diffraction. HAPY is a single chain inosilicate very similar to pyroxenes but with three instead of two cations in the octahedral layer, bonded to four oxygens and two hydroxyl groups. The Si tetrahedra are half occupied by Al and cation ordering appears in the octahedral layer with two sites occupied by Mg and one by Al. The stability of such previously unknown hydrous silicate beyond the chlorite pressure breakdown may significantly promote the H 2 O transport in the subduction channel to depths exceeding 150 km. © 2011 Elsevier B.V. All rights reserved. 1. Introduction H 2 O-transfer from the subducted slab to the overlying mantle wedge has been largely recognized as one of the major processes driving rheological modifications of the upper mantle, affecting its petrological variability and promoting magma genesis. There is an in- creasing evidence from the geological record that H 2 O flux pathways through the slab-mantle interface is not a simple gravity driven trans- fer of H 2 O from mafic rocks of the oceanic crust to ultramafic mantle wedge compositions. Profound alteration of the oceanic crust, defor- mation of hundreds of meters thick sedimentary cover, delamination and/or erosion of the hanging wall of the mantle wedge lead to the formation of large melange zones, characterized by the generation of “hybrid” rocks, often characterized by large amounts of Mg- phyllosilicates and aluminum-enriched bulk compositions (Bebout, 2007; Spandler et al., 2008). The chemical model system MgO–Al 2 O 3 –SiO 2 –H 2 O (MASH) is therefore a useful compositional space to investigate high pressure — low temperature phase relationships, not only in hydrated perido- tites but also in a variety of slab-mantle interface lithologies ranging from ultramafic schists to metasedimentary materials. A complete systematic survey of the subsystem MSH is nowadays available (Frost, 1999; Ulmer and Trommsdorff, 1999), providing constraints on transformations in pure harzburgitic compositions. On the contrary, despite its relevance, the Al-bearing space is still incom- pletely characterized and most data available refer to the portion of the system MASH with MgO:SiO 2 molar ratio ≤ 1 (see Supplementary material), involving Mg-staurolite (Fockenberg, 1998a; Schreyer and Seifert, 1969), yoderite (Schreyer and Seifert, 1969), Mg-chloritoid and Mg-carpholite (Chopin and Schreyer, 1983), MgMgAl- pumpellyite (Artioli et al., 1999; Fockenberg, 1998b), later identified as Mg-sursassite (Bromiley and Pawley, 2002; Gottschalk et al., 2000), kornerupine (Wegge and Schreyer, 1994), pyrope + water (Fockenberg, 2008). The most relevant phase at MgO:SiO 2 N 1 is chlorite, but experimental studies to date focus mainly on chlorite stability in the presence of quartz (Massonne, 1989) and of enstatite (Pawley, 2003; Staudigel and Schreyer, 1977) on bulk compositions, near or above the intersections of the tie-lines chlorite + enstatite = pyrope + forsterite, and chlorite + quartz = talc + kyanite + H 2 O. A limited dataset is available on the breakdown of chlorite in the pyrope stability field, to pressures of 3.5 GPa (Staudigel and Schreyer, 1977) and higher (Fockenberg, 1995), and a coherent arrangement of reactions involving chlorite at high pressure on the basis of the available experiments is given by Ulmer and Trommsdorff (1999). However, knowledge of the uppermost pressure stability of chlorite entirely relies on a brief summary of experimental results provided by Fockenberg (1995). In a pilot study planned to explore the stability of chlorite (Fischer et al., 2010), we encountered a phase of unknown chemical composi- tion and powder diffraction pattern in a run charge synthesized at Earth and Planetary Science Letters 310 (2011) 422–428 ⁎ Corresponding author at: Center for Nanotechnology Innovation@NEST, Istituto Italiano di Tecnologia, Pisa, Italy. Tel.: +39 050 509791; fax: +39 050 509417. E-mail addresses: mauro.gemmi@iit.it, mauro.gemmi@gmail.com (M. Gemmi). 0012-821X/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.epsl.2011.08.019 Contents lists available at SciVerse ScienceDirect Earth and Planetary Science Letters journal homepage: www.elsevier.com/locate/epsl