doi:10.1016/j.pepi.2006.02.004
Copyright © 2006 Elsevier B.V. All rights reserved.
Phase transformations between garnet and perovskite phases in the Earth’s mantle: A theoretical study
References and further reading may be available for this article. To view references and further reading you must
purchase this article.
L. Vitosa, b, c, 
, 
, B. Magyari-Köped, R. Ahujac, J. Kollárb, G. Grimvalld and B. Johanssona, c
aApplied Materials Physics, Department of Materials Science and Engineering, Royal Institute of Technology, SE-10044 Stockholm, Sweden
bResearch Institute for Solid State Physics and Optics, H-1525 Budapest, P.O. Box 49, Hungary
cCondensed Matter Theory Group, Physics Department, Uppsala University, SE-75121 Uppsala, Sweden
dTheory of Materials, Physics Department, Royal Institute of Technology, Stockholm Center for Physics, Astronomy and Biotechnology, SE-106 91, Stockholm, Sweden
Received 19 October 2005;
revised 30 January 2006;
accepted 3 February 2006.
Available online 24 March 2006.
Abstract
Using first-principles theories, we show that the stability of garnet and perovskite phases in an Al-free system is strongly influenced by both pressure and temperature, giving rise to a sequence of phase changes. Around 17±3 GPa pressure, the (Mg1−y Cay)3(MgSi)Si3 O12 majorite garnet dissociates into Ca- and Mg-perovskites. This divariant transition is associated with structural, density and elastic changes, and for y≈0.13 it has a width of 
0.6 GPa. In CaSiO3 plus MgSiO3 aggregate, a (Mg, Ca)SiO3 solid solution with an intermediate orthorhombic perovskite structure can be formed. The (Mg1−x Cax)SiO3 solid solution with x≈0.04–0.06 is calculated to be stable at the transition zone base and uppermost lower mantle conditions, and with increasing pressure it separates into perovskite end-members. The pressure–temperature stability limit for the perovskite solid solutions is close to the mantle geotherms, suggesting the appearance of structural and chemical inhomogeneities driven by temperature anomalies within the Earth’s lower mantle.
Keywords: Earth’s lower mantle; Majorite garnet; Ca-perovskite; Mg-perovskite; Perovskite solid solutions; Seismic inhomogeneity; Density Functional calculation
Fig. 1. The enthalpy difference between distorted and cubic CaSiO3 perovskite phases as a function of pressure. For the tetragonal phase we considered the I4/mmm,P4/mmm,P4/mbm and I4/mcm space groups (Shim et al., 2002), and for the orthorhombic phase the Pbnm, Cmcm and Imma space groups (Magyari-Kope et al., 2002).
Fig. 2. Equations of states for Mg and Ca silicates. Filled color symbols represent the present theoretical P−V data for MgSiO3 (blue), CaSiO3 (green), (Mg0.9 Ca0.1)SiO3 (red) perovskites and (Mg0.87 Ca0.13)3(MgSi)Si3 O12 (brown) garnet, respectively. Black open symbols are the observed pressure-volume data for MgSiO3 perovskite (squares) (Stixrude and Cohen, 1993), CaSiO3 perovskite (triangles) Shim et al., 2002 and Mao et al., 1989, and (Mg0.84 Ca0.16)3(MgSi)Si3 O12 garnet (diamond) (Hazen et al., 1994). Experimental P−V data of the phases in MORB transformed at mantle conditions (Funamori et al., 2000) are shown by open color symbols. The inset shows the calculated pressure dependencies of the average degree of octahedral tilts in the perovskite structures together with experimental data for MgSiO3 perovskite (open squares) (Kudoh et al., 1987).
Fig. 3. The Gibbs energy of formation for the MgCa-pv and MgCa-gt solid solutions. The energies are plotted as functions of pressure for 0 and 2000 K (inset). The legend shows the amount of substitutional Ca considered in EMTO and PAW calculations. Filled symbols mark the results from PAW simulations, while continuous lines correspond to the results from EMTO calculations. Note that in inset, the scale factor for the gt structure has been divided by 40.
Fig. 4. Theoretical pressure–temperature phase relations of the MgSiO3-CaSiO3 perovskite system. To the left of the Ptr(T) solid curves the (Mg, Ca)SiO3 perovskite solid solution is stable, while Mg and Ca form their own perovskite phases on the right hand side. The uncertainties in the transition pressure are around ±5%, as estimated from the difference between the two sets of theoretical data. The average octahedral tilts are shown in parentheses. The dashed lines (LMC) mark the low mantle pressure–temperature conditions. Approximate depths are shown in the upper part of the figure. Black squares (CMA) correspond to the orthorhombic perovskite solid solution observed in MORB experiments (Funamori et al., 2000).
Corresponding author. Tel.: +46 8 790 8356; fax: +46 8 100 411.
Abstract
Purchase PDF (357 K)
E-mail Article
Add to my Quick Links
Cited By in Scopus (2)
Purchase PDF (218 K)
