Online Resource Supplement
Experimental and analytical methods
Experiments in the Fe3O4–Fe2SiO4 system employed several mixtures of magnetite (mt) and fayalite, or pre-synthesised spinel. Magnetite was produced by reacting Fe2O3 powder in a one atmosphere gas-mixing furnace at 1300°C and log ƒO2 = -5.5; conditions which should yield essentially stoichiometric magnetite (Dieckmann, 1982). This was confirmed by analysis of the powder X-ray diffraction pattern which revealed single-phase magnetite with a unit-cell parameter of ao= 8.3966(6) Å. Fayalite was also produced in a gas-mixing furnace at 1100°C and a log ƒO2 = -13. Several repeated grinding and firing cycles were necessary to produce single-phase fayalite, as monitored by powder X-ray diffraction. For experiments using a pre-synthesised spinel with a composition of 85 % magnetite -15 % Fe2SiO4, the spinel was produced from a mixture of magnetite and fayalite packed in a Ag capsule and run at 6 GPa and 1100 °C in a belt apparatus (see Woodland &Angel 2000 for a description of this press). This synthesis yielded a single-phase spinel with a unit cell of 8.3762(8) Å, which is in excellent agreement with that expected for such a composition (Woodland Angel 2000).
For compositions in the Fe3O4–Fe2SiO4–Mg2SiO4 system a similar approach was taken except that various Fe2SiO4–Mg2SiO4 olivine solid solutions were employed together with magnetite in the starting mixtures. Details of these starting materials are given in Koch et al. (2004). The Cr-bearing experiments employed two different Fe3O4-FeCr2O4 solid solutions, with 20 and 50 mol % chromite component. These two solid solutions were synthesized at 1 bar from high-purity Fe2O3 and Cr2O3 at a log ƒO2 = -6.5 in a gas-mixing furnace at 1300 °C. Repeated sintering and grinding cycles were performed until homogeneity was achieved. The unit-cell parameters for 80 % magnetite – 20 % chromite and 50 % magnetite – 50 % chromite are 8.3856(8) Å and 8.3973(8) Å, respectively, which are in very good agreement with data from Robbins et al. (1971) for these two compositions.
High-pressure experiments in the P-T range of 9 – 15 GPa and 1000 – 1300 °C were performed at both the Bayerisches Geoinstitut, Bayreuth and at the University of Frankfurt. At the Bayerisches Geoinstitut three multi-anvil presses were employed: HYMAG MA-6/81000 tsplit-sphere-type press, Zwick 5000 t press and a Voggenreiter Walker-type press.These presses have been cross-calibrated over awide range in temperature (at least 800-1800°C, see Keppler and Frost 2005). The experiments in Frankfurt were performed in an 800 t Walker-type multi-anvil press. The high-temperature calibration of this multi-anvil apparatus was based upon bracketing the positions of the αFe2SiO4–Fe2SiO4 transition (Yagi et al., 1987), coesite – stishovite transition (Zhang et al., 1996) and the αMg2SiO4–Mg2SiO4 transition at 1200°C and 13.6 GPa (Morishima et al. 1994).
Sample mixtures were packed in Ag capsules that were hammered shut, forming a cold weld. Cr2O3-doped MgO pressure assemblies with either 18 mm or 14 mm edge length were employed. At Frankfurt, heating was provided by a Re-foil furnace with the temperature monitored by a W5/Re95 – W26/Re74 thermocouple with the emf uncorrected for pressure.The capsule was surrounded by a thin MgO sleeveto avoid metal-metal contact with the furnace. The thermocouple, separated by a 0.2 mm thin layer of ceramic cement, was placed directly above the sample. Below the sample, a MgO spacer ensures the central position of the capsule. Further technical details are provided in Hanrahan et al. (2009). At Bayreuth, a LaCrO3 furnace and W3/Re97 – W25/Re75 thermocouple were used. Experimental details of these experiments are provided in Koch et al. (2004). The experiments were terminated by turning off the power to the furnace. The experimental conditions and the run duration are summarized in Tables 1, 2 and 3.
The chemical compositions of the product phaseswere measured by electron microprobe at the Universität Heidelberg for the Mg-bearing experiments and at Universität Frankfurt for the Si-bearing and Cr-bearing experiments. Operating conditions were 15 kV and 20 nA and all measurements were made in wavelength dispersive mode. AtHeidelberg, wollastonite, Fe2O3 and MgO were used as standards (see also Koch et al. 2004). In Frankfurt, fayalite, wollastonite and Cr2O3 were used as standards. Phase compositions in terms of cations per formula unit are provided in Tables 1, 2 and 3 and are averages of between 3 to 20 points; in most cases more than 5 points were averaged. The quoted uncertainties represent 1 errors.
Unit-cell parameters and molar volumes were determined from powder X-ray diffraction patterns. For the Mg-bearing samples, patterns were collected from 30 to 120° 2 with a Stoe STADI-P diffractometer located at the Bayerisches Geoinstitut, Bayreuth operating in transmission mode with monochromatic Co Ka1 radiation and a small-aperture linear PSD. For the other samples, patterns were obtained from 20 to 120 2 either in Frankfurt or Bayreuth with a Philips X´Pert PRO diffractometer. In Bayreuth, monochromatic Co Ka1 radiation was employed and in Frankfurt, monochromatic Cu Ka1 radiation was used. Each sample was ground under acetone together with a small amount of Si (ao = 5.43088 Å), which served as an internal standard to correct the 2scale. Such corrections were generally less than 0.10° 2andallowcell parameters determined on different instruments to be reliably compared. Unit-cell parameters were then obtained from the corrected diffraction patterns by full-pattern Rietveld refinements using the GSAS software package (Larson & von Dreele 1988) and the EXPGUI user interface (Toby 2001).
References
Dieckmann, R. (1982) Defects and cation diffusion in magnetite (IV)-nonstoichiometry and point-defect structure of magnetite Fe3-δO4. Berichte der Bunsen-Gesellschaft-Physical Chemistry Chemical Physics, 86, 112–118
Hanrahan M, Brey G, Woodland A, Altherr R, Seitz H-M (2009) Towards a Li barometer for bimineralic eclogites. Contrib Mineral Petrol 158: 169-183. DOI 10.1007/s00410-009-0376-7
Keppler H, Frost DJ (2005) Introduction to minerals under extreme conditions. In Miletich R (ed) Mineral behaviour at extreme conditions. Eur Mineralogical Union Notes, vol 7, Eötvös University Press, Budapest, 1-30
Koch, M., Woodland, A.B., Angel, R.J. (2004): Stability of spinelloid phases in the system Fe3O4–Fe2SiO4–Mg2SiO4 at 1100°C and up to 10.5 GPa. Phys. Earth Planet. Interiors, 143-144, 171-183.
Larson, A.C. von Dreele, R.B. (1988) GSAS manual. Los Alamos National Laboratory, report LAUR, 86-748.
Morishima H, Kato T, Suto M, Ohtani E, Urakawa S, Utsumi W, Shimomura O, Kikegawa T (1994) The phase boundary between a-and b-Mg2SiO4 determined by in situ X-ray observation. Science 265:1202–1203. doi:10.1126/science.265.5176.1202
Toby BH (2001) EXPGUI, a graphical user interface for GSAS. J Appl Cryst 34: 210–213
Woodland AB, Angel RJ (2000) Phase relations in the system fayalite-magnetite at high pressures and temperatures. Contrib Mineral Petrol 139, 734-747.
Yagi T, Akaogi M, Shimomura O, Suzuki T, Akimoto S (1987) In Situ Observation of the Olivine-Spinel Phase Transformation in Fe2SiO4 Using Synchrotron Radiation. J Geophys Res 92: 6207-6213
Zhang J, Li B, Utsumi W, Liebermann RC (1996) In situ X-ray observations of coesite-stishovite transition: reversed phase boundary and kinetics. Phys Chem Mineral 23: 11-16