COMPRESSIBILITY AND THERMAL EXPANSION OF GARNETS WITH COMPOSITIONS TYPICAL OF INCLUSIONS IN DIAMONDS

Transcription

1 COMPRESSIBILITY AND THERMAL EXPANSION OF GARNETS WITH COMPOSITIONS TYPICAL OF INCLUSIONS IN DIAMONDS Ph.D. candidate: SULA MILANI, II course Supervisor: Prof. FABRIZIO NESTOLA Cycle: XXVII Abstract Garnet is the most abundant inclusion found in diamonds and is one of the most important mineral of the Earth s upper mantle. In order to obtain information about the depth of formation of diamond-inclusion pair precise and accurate thermoelastic parameters for the diamond and the inclusion are strongly necessary. The aim of this Ph. D. project is to determine the bulk modulus and thermal expansion of a series of garnets with compositions analogue to those found in diamonds. The applied methodology is in situ high-pressure X-ray diffraction with a diamond anvil cell device and in situ high-temperature X-ray diffraction using a microfurnace. During this second year the Isothermal Equation of State of a synthetic single-crystal of pyrope was determined and a series of multi-anvil experiments were carried out to synthesize grossular and uvarovite endmembers and solid solutions. Introduction Most of the diamonds formed in the sub-cratonic lithospheric mantle, in the so called diamond window (Fig. 1; Stachel and Harris, 2008), but only a tiny amount of them contain mineral, fluid and melt inclusions which have been trapped during diamond formation. These inclusions represent a direct window on the Earth s upper mantle which is otherwise inaccessible to direct observations. Study of inclusions in diamonds, therefore, is important to better constrain the chemical and physical conditions which determine the processes of diamond formation. Garnets are one of the most abundant mineral inclusions found in diamonds. They have a quite complex chemical composition being mainly solid solutions of pyrope (Mg 3 Al 2 Si 3 O 12 ), almandine (Fe 3 Al 2 Si 3 O 12 ), grossular (Ca 3 Al 2 Si 3 O 12 ), and uvarovite (Ca 3 Cr 2 Si 3 O 12 ). Fig 1. - Schematic vertical section of the Earth s crust and part of the upper mantle (after Stachel and Harris, 2008). 1

2 In literature the pressure and temperature at which the diamond-inclusion pairs grown have been calculated applying classical geothermobarometric methods based on the chemical exchange between minerals. However, these geothermobarometers can be applied only when the appropriate set of mineral assemblages is present in the same inclusion. Recently, a method based on the pressure exerted by the diamond on the mineral inclusions and on the stresses in the surrounding diamond has been applied to obtain the pressure of the diamond-inclusion source (e.g. Nestola et al., 2011, Howell et al., 2012). This method can potentially be applied to any single mineral inclusion, but requires accurate knowledge of the thermoelastic parameters (bulk modulus, thermal expansion, shear modulus, variation of the bulk modulus as a function of temperature) of both the diamond and the inclusion. These properties are well constrained for diamond, but accurate and precise data for the minerals typically found as inclusions are still lacking. The aim of this study is, therefore, to determine precise and accurate thermoelastic parameters for garnets with complex composition, similar to those found as inclusions in diamonds. Experimental methods a) In situ high-pressure experiments The Equations of State (Eos) of pure almandine and the solid solution Py60-Al40 were determined in the first PhD year. During the second year the Equation of State (EoS) of a synthetic single crystal of pyrope was determined. The experiment was carried out at the Department of Geosciences, University of Padua. A crystal of 40 μm thickness and 70 μm largeness was loaded in an ETH-type Diamond Anvil Cell (DAC; Miletich et al., 2000), using a steel gasket, preindented to a thickness of 100 μm and with a 300 μm diameter hole. As an internal pressure standard a crystal of quartz was used (Angel et al., 1997) and as pressure transmitting medium a mixture 4:1 methanol-ethanol was used (such mixture transmits hydrostatically the pressure to maximum 10 GPa, Angel et al., 2007). Unit-cell parameters were determined at 12 different pressure steps up to 8.46 GPa using a STOE STADI IV four-circle diffractometer working with MoKα operating at 50 kv and 40 ma equipped with a point detector and controlled by the software SINGLE (Angel and Finger, 2011). During the centering procedure, the effect of crystal offsets and diffractometer aberration were eliminated from refined peak positions by eight-position centering method (Angel et al., 2000; King and Finger, 1979). As can be observed in Fig. 2a, 2b it was not necessary to measure other intermediate almandinepyrope solid solution, because both the bulk modulus and the unit-cell volume are linearly correlated. a) b) Fig. 2 Diagrams showing the linear correlation between (a) composition (mol% of pyrope) and bulk-modulus and (b) composition and unit-cell volume. The error bars in (b) are smaller than the symbol size. 2

3 b) Syntheses During this second year several syntheses were carried out at the Bayerisches Geoinsitut (BGI), University of Bayreuth, Bayreuth, Germany, worldwide famous laboratory for High-Pressure experiments. In this laboratory I tried to synthesize uvarovite and grossular garnets and some solid solutions (Gr 50 -Uv 50, Gr 75 -Uv 25 ). Synthesizing single crystals of garnets is very challenging, for this reason more than one experiment have been done in order to find the optimal conditions (e.g. starting compositions, pressure, temperature, temperature path) for the syntheses. All the experiments were done using a Multi-Anvil apparatus reaching the pressure of 6 GPa and temperatures in the range C. The syntheses were done using a 18/11 assembly (Keppler and Frost, 2005), with two platinum capsules and a graphite heater. The temperature was measured during the experiments by a type D thermocouple. The starting materials used were different stoichiometric composition of SiO 2, CaO, Cr 2 O 3 and Al(OH) 3 oxides intimately grinded in an agate mortar for ca 1 hour under ethanol. In order to decrease the melting point of the oxides mixture and enhance the crystal growth water was used as a flux. For what concern the grossular garnets and the solid solutions the water used was given by the Al(OH) 3 component, while for the uvarovite garnets 2 3 μl of molecular water was added before closing the sample capsules. Results and discussion a) High-pressure experiment The estimation for the best Equation of State for the measured P-V data of pyrope was made on the basis of the Fe-f plot. In the diagram is evident that the data fit on a positively inclined straight line, indicating that the Birch-Murnaghan equation of state, truncated at the third-order (BM3-EoS; Birch, 1947), gives the best fit of the data (Fig. 3). The determination of all the EoS parameters was made using the software EoS-FIT 5.2 (Angel, 2000) that permit to refine simultaneously all the parameters of the equation. Comparing the data obtained for the pyrope with the data obtained for the almandine and the Al 40 -Py 60 solid solution (Table 1, Fig. 3) we observe that the values for the volume and the isothermal bulk moduli decrease with increasing the pyrope content, thus the Mg 2+ content in the garnet. This means that the pyrope is more compressible than the almandine; the reason for this behavior is not yet well understood but a possible explanation could be the size of the cation Mg 2+, that it is too small for the big garnet dodecahedral site, allowing a larger compression. Fig.3 Equation of State (EoS) of almandine, pyrope and of Al 40 Py 60. BM3 Birch-Murnaghan of 3 rd order. 3

4 The EoS coefficient determined for the pyrope clearly shows significant differences with existing literature data (e.g. Leger et al., 1990, Zhang et al., 1998, Nobes et al., 2000, Sinogeikin et al., 2000). V 0 (Å 3 ) K T0 (GPa) K almandine (10) 172.5(1.5) 5.8(5) Al 40 Py (13) 167.2(1.8) 5.6(5) pyrope (17) 163.7(1.7) 6.4(4) Table 1 EoS coefficients obtained for the almandine, Al40Py60 and pyrope single crystals. Research plan for next year During the third year the high pressure and high temperature experiments on the last synthetic garnets will be ultimate. In this way it will be possible to cover the compositional range of garnets found as inclusions in diamonds and to verify how the thermoelastic parameters change with the chemical composition. All the obtained data will be used to model the relationships between the thermoelastic parameters of garnet and its composition. At last, high pressure and high temperature experiment will be carried out also on natural or synthetic garnets, of eclogitic composition, to verify the reliability of the model obtained in this study. References ANGEL, R.J. et al. (1997). J. Appl. Crystallogr., 30, ANGEL, R.J. (2000). In: Hazen, R.M., Downs, R.T. (Eds), Rev. in Mineral. And Geochem., 41, pp ANGEL, R.J et al. (2000). In: Hazen, R.M., Downs, R.T. (Eds), Rev. in Mineral. And Geochem., 41, pp ANGEL, R.J. et al. (2007). J. Appl. Crystallogr., 40, ANGEL, R.J. et al. (2011). J. Appl. Crystallogr., 44, BIRCH, F. (1947). Phys. Rev., 71, HOWELL, D. et al. (2012). Eur. J. Mineral., 24, KEPPLER, H. and FROST, D.J. (2005). EMU Notes in Mineralogy, 7, KING, H.E. et al. (1979). J. Appl. Crystallogr., 12, LEGER, J.M. et al. (1990). Phys. Chem. Mineral., 17, MILETICH, R. et al. (2000). In: Hazen, R.M., Downs, R.T. (Eds), Rev. in Mineral. And Geochem., 41, pp NESTOLA, F. et al. (2011). Earth Planet. Sci. Lett., 305, SINOGEIKIN, S.V. (2000). Phys. Earth Planet. Inter., 120, STACHEL, T. and HARRIS, J.W. (2008). Ore Geol. Rev., 34, ZHANG, L. et al (1998). Phys. Chem. Min., 25, SUMMARY OF LAST YEAR S ACTIVITIES I have spent 6 months (1 st August st January 2014) at the Bayerisches Geoinsitut (Bayreuth, Germany), where I have attended the courses of Prof. D.J. Frost and H. Keppler (see the list below). Courses: - WHIGHAM C.: Academic English Writing-Skills, 08 th Jan. 28 th Feb. 2013, Università di Padova. - ANGEL R.J.: Elasticity as the key to Material and Earth Sciences, 22 nd Jan. 21 st Feb. 2013, Università di Padova. - BATTISTUTTA R.: Cristallografia e Biocristallografia, 04 th March 15 th June 2013, Università di Padova, 6 CFU. - RANALLI G.: Some problems of Geodynamics, 15 th 19 th 2013, Università di Padova. - DALCONI M.C.: Corso teorico-pratico sull analisi dati in diffrazione di raggi X da polveri con il metodo Rietveld, 10 th 12 th June 2013, Università di Padova. - FROST D.J.: Planetary Geology, 17 th Oct th Jan. 2014, Bayerisches Geoinsitut, University of Bayreuth, Bayreuth, Germany. 4

5 - KEPPLER H.: Spectroscopic Methods in Mineralogy, 16 th Oct th Jan. 2014, Bayerisches Geoinsitut, University of Bayreuth, Bayreuth, Germany. Teaching activities: Teaching assistant: 25 hours, Mineralogia, Laurea triennale in Scienze geologiche (2011/2012). Communications: MILANI S., MAZZUCCHELLI M., NESTOLA F., ALVARO M., ANGEL R.J., GEIGER C.A and DOMENEGHETTI C The P-T conditions of garnet inclusion formation in diamond: thermal expansion of synthetic end-member pyrope. EGU 2013 meeting, April 7 th -12 th, Vienna, Austria. NESTOLA F., MILANI S., ANGEL R.J., PASQUAL D. and GEIGER C.A Pressure-volume equation of state for pyrope-almandine solid solutions. EGU 2013 meeting, April 7 th -12 th, Vienna, Austria. NESTOLA F., NIMIS P., MILANI S., ANGEL R.J., BRUNO M. and HARRIS J.W Crystallographic Relationships between Diamond and its Olivine Inclusions. An Update. Goldschmidt 2013 meeting, August 25 th 30 th, Firenze, Italy. Publications: NESTOLA F., NIMIS P., ANGEL R.J., MILANI S., BRUNO M., PRENCIPE M. and HARRIS J.W Syngenesis or Protogenesis? New constraints on diamond formation from olivine inclusions. Submitted to Nature Geoscience. 5