Eleonora Paris (paris@camserv.unicam.it)1, Gabriele Giuli1, Claudia Romano2, Don Dingwell3 & Ivan Davoli4
1 Dip. di Scienze della Terra and INFM, Università di Camerino, I-62032 Camerino, Italy
2 Dip. di Scienze Geologiche, Terza Università di Roma, Roma, Italy
3 Bayerisches Geoinstitut, Universitaet Bayreuth, Bayreuth, Germany
4 Dip. di Fisica and INFM, Università di Roma Tor Vergata, Roma, Italy
In the melt structure Nb, Ta and W are considered as octahedrally coordinated and acting as network-modifiers. However the absence of structural data on these elements in glasses and melts of geological interest is the reason of this investigation, since the presence of different polyhedral geometries can affects the polymerization of the melt and its structure-related properties. Here we present the results obtained via Nb, Ta and W XAS spectra of aluminosilicate glasses of different composition. The spectra were collected at ESRF (GILDA beamline) in fluorescence mode. The glasses were synthesized at high temperatures in the system SiO2-Al2O3-Na2O-K2O with different Al2O3 versus alkali ratios and with variable Nb2O5 (Ta2O5, WO2) concentrations (0.1-10 wt%). Nb XANES spectra show variations in the intensity of the pre-edge region peak as a function of the bulk chemical composition of the glasses and in particular of the Al2O3/Na2O+K2O ratio. EXAFS spectra of glasses of basaltic and granitic composition, show that the Nb-O average distances are 1.96(2) Å and Nb is in tetrahedral coordination (CN=4). In a peralkaline composition niobium CN changes from [4] to [7] with the formation of a pentagonal bypiramid of oxygens around Nb (Nb-O=2.15(2) Å 5x and Nb-O=2.52(2) Å 2x). The interpretation of these data suggest that: -it is possible to analyze these elements in spite of the low concentrations; -the assumption of octahedral Nb in melts is wrong; -the stabilization of CN higher than 4 is related to the bulk composition of the glass and is driven by the alkalis content A comparison between the structural environment and geometrical configuration in glasses of similar starting composition of Ta and W is also shown. These may prove useful in rationalizing the geochemical behavior of high-field elements in fractionating igneous systems and in better relating the structural and physical properties of the melts.
Georg M. Partzsch (partzsch@classic.min.uni-heidelberg.de) & Dominique Lattard (dlattard@classic.min.uni-heidelberg.de)
Mineralogische Institut, Universität Heidelberg, INF 236, Germany
Classical interpretations of layered intrusions (Skaergaard, Kiglapait) suggest that the strong iron enrichment during the differentiation (Fenner trend) is related to fractional crystallization under conditions closed to oxygen. To test this hypothesis, we have performed equilibrium crystallization experiments at 1 bar in a closed, dry, ferro-basaltic system. The starting material was a synthetic eight-component ferro-basaltic glass similar in composition to liquids proposed to have been parental to the Skaergaard intrusion. In a first step, the ferrous-ferric ratio of batches of the starting material were set at different values through a treatment at super-liquidus temperatures (1180°C) under oxygen fugacities fixed by CO/CO2 mixtures in the range FMQ+1 to FMQ-2 (starting fO2). The samples were subsequently sealed in evacuated silica glass ampoules, and held at sub-liquidus temperatures (1090-1170°C) to achieve equilibrium crystallization under conditions closed to oxygen. The order of crystallization with decreasing temperature is similar to that in a system open to oxygen, i.e. plagioclase, olivine, Ca-rich clinopyroxene and, depending on the starting fO2, either ilmenite (FMQ-2), or ilmenite followed by titanomagnetite (FMQ-1), or the contrary (*FMQ). Compared to conditions open to oxygen, magnetite crystallizes at higher temperatures, which indicates increasing fO2 during crystallization of the silicates in closed systems. Preliminary estimates with the magnetite-ilmenite oxibarometer point to an increase of at least 1 log unit over a temperature decrease of 80°C. The FeO(tot) content of the melt keeps on rising after the onset of magnetite crystallization, as in the Fenner trend. This is related to the small proportions of magnetite at its liquidus. However, even with low starting fO2, the maximum FeO(tot) contents of the melts are similar to those in conditions open to oxygen (~ 18 wt.%). The present results do not suggest the strong Fe-enrichment of the residual liquid predicted for the Skaergaard Intrusion.
A. R. Pawley (alison.pawley@man.ac.uk)1 & D. R. Allan (dra@ph.ed.ac.uk)2
1 Department of Earth Sciences, University of Manchester, Manchester M13 9PL, UK
2 Department of Physics and Astronomy, University of Edinburgh, Edinburgh EH9 3JZ, UK
Lawsonite, CaAl2Si2O7(OH)2.H2O, is a common mineral in blueschist-facies metabasalts. Because of its high H2O content (11.5 wt%) and high-pressure stability (~12 GPa at 900°C), it could be important for transporting H2O deep into the Earth's mantle at subduction zones. Recent compressibility studies of lawsonite using X-ray diffraction and Raman spectroscopy (Daniel et al., 1998) and infrared spectroscopy (Scott and Williams, 1999) have revealed a structural discontinuity near 9 GPa, but the high-pressure structure was not determined.
We have investigated the structure of lawsonite during compression to 16 GPa using a diamond-anvil cell and angle-dispersive X-ray powder diffraction, using the image-plate detector system at the Synchrotron Radiation Source at the Daresbury Laboratory, UK. The diamond-anvil cell was loaded with a 4:1 methanol-ethanol pressure transmitting medium and ruby chip for pressure calibration. The lawsonite was a natural sample, of close to end-member composition. Up to 10 GPa lawsonite is orthorhombic, with space group Cmcm. Its bulk modulus, obtained from a fit to the data of the Murnaghan equation with K' = 4, is K = 125.0(2) GPa. At 10 GPa the diffraction pattern shows the onset of peak splitting, indicating a phase transition. From 10 to 11 GPa the sample is mixed phase, while at 12 GPa and above it is single phase. The high-pressure phase is monoclinic, with space group P21/m, and the phase transition is displacive, reversible and probably first order.
In another experiment we heated lawsonite in an externally-heated diamond-anvil cell to 200°C at high pressures. The results suggest that the dP/dT slope of the phase transition is shallow, and therefore that lawsonite in subducting slab might undergo this phase transition before breaking down.
Daniel I, Fiquet G, Gillet P, Schmidt M & Hanfland M, Terra Nova Abs. Supp, 10, 10, (1998).
Scott HP & Williams Q, Phys. Chem. Minerals, 26, 437-445, (1999).
Index of EMPG VIII Volume
Further EMPG VIII Information
Index of the Journal of Conference Abstracts
Cambridge Publications Home Page