Previous page     Contents     Next page

3.2 f. Phase equilibria of hydrous phases at high pressure and temperature (D.J. Frost)

The presence of relatively small concentrations of H2O can have large effects on the properties of minerals and melts that form earth's mantle. A first step in quantifying these effects is to identify the likely phases that may host H2O at high pressure and temperature. In addition, the stability of hydrated minerals at conditions applicable to subducting slabs has important implications for the recycling of H2O into the mantle.

There is no shortage of proposed hosts for H2O in Earth's interior. Many minerals that comprise the bulk of the mantle are known to partition significant concentrations of H2O. The high pressure (MgFe)2SiO4 polymorph wadsleyite, for example, is capable of containing up to 3.3 wt% H2O. H2O may also be stored in hydrous potassic phases such as phlogopite or K-richterite. Water-rich experiments performed at high pressure and temperature have led to the identification of a whole family of dense hydrous magnesium silicate phases (DHMS). However, little is known about the stability fields of such phases within bulk compositions applicable to the mantle.

The first aim of this project is to identify whether DHMS are important in the storage of H2O in the earth. High pressure and temperature multianvil experiments are being performed in order to examine the maximum thermal stability of DHMS phases within plausible mantle bulk compositions. Starting materials comprise natural orthopyroxene and olivine combined with either brucite or presynthesised hydrous phases (superhydrous phase B and phase E). The resulting harzburgitic compositions contain approximately 4 wt% H2O. Such a composition could be produced in subducting slabs by the breakdown of antigorite at 6 GPa.

Starting materials are welded into platinum capsules with inner Re-foil sleeves to minimise Fe-loss. Multianvil experiments are performed between 10 and 25 GPa and between 900 and 1500° C. Phases in the quenched experimental charges are examined using micro-focus X-ray diffraction, Raman spectroscopy and electron microprobe analysis.

A simple summary of results on the stability of DHMS in a harzburgitic composition is shown in Fig. 3.2-11. Superhydrous phase B shows the highest temperature stability of the DHMS

Fig. 3.2-11: Ternary MSH diagrams showing the composition of run products from multianvil experiments performed on a hydrated harzburgite composition between 10 and 25 GPa. MgO and FeO are on the left-hand apices, H2O at the top and SiO2 on the right. Solid circles denote hydrous phases, grey symbols show the starting composition and white circles are nominally anhydrous phases. Arrows denote the presence of a fluid or melt, the composition of which was not analysed. Ternary diagrams are shaded grey where hydrous phases were produced. Positions of the - , - and -Pv+MW transformations for (Mg1.8Fe0.2)SiO4 are shown in grey. An average mantle adiabat (ama), an estimate of the minimum temperature in a subducting slab (cs), and the antigorite breakdown curve (ant) are also shown.

phases, being stable to 1350°C at 22 GPa. This phase is a potential host for H2O in subducting slabs at the base of the transition zone. At lower mantle conditions phase D is stable to approximately 1200°C. Microprobe analyses provide information on the partitioning of Al2O3 and FeO into DHMS phases. Phase D is more Al2O3- and FeO-rich than coexisting magnesium silicate perovskite, and superhydrous phase B is FeO-rich in comparison to ringwoodite. Such partitioning may stabilise these phases to higher temperatures in more fertile peridotite compositions. Experiments are also being performed on basaltic compositions containing 1-4 wt% H2O at transition zone and lower mantle conditions. This will help to constrain the depth to which H2O could be recycled into the mantle in the mafic component of subducting slabs.

The presence of H2O in the bulk mantle or in subducting slabs could influence the position and width of transitions that result in seismic discontinuities. A second aim of this project is to examine experimentally the effect of H2O on such phase transformations. The determination of the stability fields of wadsleyite, ringwoodite and perovskite in wet systems can also provide information on the thermodynamics of water solubility in nominally anhydrous phases. Comparative experiments are performed where dry and wet samples are run in the same experiment. Two welded platinum capsules are placed symmetrically about a thermocouple inserted horizontally through the wall of the furnace. Experiments have been performed on Mg2SiO4 and San Carlos olivine with between 1 and 4 wt% H2O.

Preliminary results show that the stability field of wadsleyite is extended to higher pressure by approximately 1 GPa in the presence of 4 wt% H2O. The ringwoodite to perovskite plus MgO transition, on the other hand, appears to be insensitive to the presence of H2O.

Bayerisches Geoinstitut, Universität Bayreuth, 95440 Bayreuth, Deutschland
Tel: +49-(0) 921 55 3700 / 3766, Fax: +49-(0) 921 55 3769, E-mail: bayerisches.geoinstitut(at)