Search

The Ca-Sr fractionation between zoisite and, respectively, lawsonite and an aqueous fluid has been determined by synthesis experiments in the presence of a 1 M (Ca,Sr)Cl2 aqueous fluid at 2.0 GPa/550, 600, and 700 °C and 4.0 GPa/800 °C for zoisite and 2.0 GPa/400 °C and 4.0 GPa/600 °C for lawsonite. Solid run products were characterized by EMP, SEM, and XRD with Rietveld refinement and fluids were analyzed by ICP-OES. Zoisite exhibits notable intracrystalline Ca-Sr fractionation between the A1 and A2 sites and calculated intracrystalline exchange coefficients KD(Sr-Ca)A1-A2 = 1.5 to 26 show strong preference of Sr over Ca for the slightly larger A2 site. Calculated individual site-dependent zoisite/aqueous fluid (af, in superscripts)-exchange coefficients for the studied 1 M (Ca,Sr)Cl2 aqueous fluids are K(Sr-Ca)zo A1-af = 3.38 to 41.08 for the A1 site and K(Sr-Ca)zo A1-af = 0.45 to 6.51 for the A2 site. Assuming γCaaf = γSraf and a symmetric mixing model, the thermodynamic evaluation of the site-dependent exchange reactions Ca2+(af) + SrA1(M2+)A2 Al3[Si3O11(O/OH)] = Sr2+(af) + CaA1 (M2+)A2Al3[Si3O11(O/OH)] and Ca2+(af) + (M2+)A1 SrA2 Al3[Si3O11(O/OH)] = Sr2+(af) + (M2+)A1 CaA2 Al3[Si3O11(O/OH)] yields Δµ0 = --29 kJ/mol and WSr-Cazo A1 = 5.5 kJ/mol for the A1 site and Δµ0 = -1.1 kJ/mol and WSr-Cazo A2 = 0 kJ/mol for the A2 site at P and T of the experiments. The data indicates ideal Ca-Sr substitution on the A2 site. Lawsonite formed in both the orthorhombic Cmcm and the monoclinic P21/m form. Calculated lawsonite-aqueous fluid-exchange coefficients indicate overall preference of Ca over Sr in the solid and are KD(Sr-Ca)law Cm -cm-af = 1.12 to 11.32 for orthorhombic and KDD(Sr-Ca)law Cm -cm-af = 1.67 to 4.34 for monoclinic lawsonite. Thermodynamic evaluation of the exchange reaction Ca2+(af) + SrAl2Si2O7(OH)2·H2O = Sr2+(af) + CaAl2Si2O7(OH)2·H2O assuming γCaaf = γSraf and a symmetric mixing model yields similar values of Δµ0 = --9 kJ/mol and WSr-Calaw Cmcm-af = 10 kJ/mol for orthorhombic and Δµ0 = -10 kJ/mol and WWSr-Calaw P21/m = 11 kJ/mol for monoclinic lawsonite. Calculated Nernst distribution coefficients for the studied 1 M (Ca,Sr)Cl2 aqueous fluids are DSrzo-af = 2.8 ± 0.7 for zoisite at 2 GPa/600 °C and DSrlaw Cmcm-af = 0.6 ± 0.2 for orthorhombic lawsonite at 4 GPa/600 °C and show Sr to be compatible in zoisite but incompatible in lawsonite. This opposite mineral-aqueous fluid-fractionation behavior of Sr with respect to zoisite and lawsonite on the one hand and the ideal Ca-Sr substitution on the zoisite A2 site in combination with the strong intracrystalline Ca-Sr fractionation in zoisite on the other hand, make Sr a potential tracer for fluid-rock interactions in zoisite- and lawsonite-bearing rocks. For low Sr-concentrations, xSrzo directly reflects xSraf and allows us to calculate Sr-concentrations in a metamorphic aqueous fluid. During high-pressure aqueous fluid-rock interactions in subduction zone settings the opposite mineral-aqueous fluid-fractionation behavior of Sr results in different aqueous fluid characteristics for lawsonite- vs. zoisite-bearing rocks. Ultimately, subduction zone magmas may trace these different aqueous fluid characteristics and allow distinguishing between cold, lawsonite-bearing vs. warm, zoisite-bearing thermal regimes of the underlying subduction zone. [ABSTRACT FROM AUTHOR]

We performed hydration experiments of pure and Nb-, Cr-, and V-doped synthetic dry (H2O < 3 ppm) single rutile crystals. They were equilibrated with pure H2O in hydrothermal experiments at constant conditions of 600 °C, 400 MPa, and fO2 near the Ni-NiO buffer, run time between ~25 min and 14 days. Slabs cut parallel to (110) of the reacted single crystals (1 to 2 mm³) were analyzed for H+ by FTIR. Hydration occurs almost spontaneously and the H2O-equivalent is uniformly distributed in the samples, but depends extremely on trace element contents. In pure rutile, the average H2O-content is 314 ± 50 ppm, the saturation level at these conditions. Rutile doped with 500 ppm Nb has a lower average H2O content of ~235 ppm, rutile with 2000 ppm Cr has ~900 ppm H2O, and rutile with 2000 ppm V does not incorporate H2O. During stepwise heating at atmospheric pressure of a reacted Nb-doped rutile, H+ is quickly released between 450 and 550 °C. UV-VIS spectra of unreacted colorless and reacted deep blue pure rutile show the rutile-characteristic sharp absorption edge in the UV spectra. The reacted rutile has a broad absorption band at 6500 cm-1 wavenumber attributed to intervalence charge transfer transition Ti3++Ti4+ → Ti4++Ti3+. The reduction of Ti4+ to Ti3+ is charge balanced by the incorporation of H+. The Nb-doped rutile changed its color from light greenish-blue (untreated) to deep blue. In the untreated Nb rutile, the UV-VIS absorption band at 6500 cm-1 indicates that Nb5+ is charge balanced by Ti3+. In the reacted Nb-rutile the absorption band is more intense, but compared with the pure rutile, H+ incorporation is lower by the equivalent of Ti3+ reduced in the untreated rutile. Reacted Cr-rutile almost retains its brownish-orange color, but the spectrum shows a prominent Ti3+/Ti4+ IVCT band at ~6400 cm-1 with moderate intensity considering the high-H2O contents of ~900 ppm. The high-H+ contents are best explained by the reduction of Cr4+ to Cr2+. The UV-VIS spectra of the dark-blue to opaque V-doped rutile show a very strong absorption toward low energies, which is likely caused by reduction of Ti4+ to Ti3+ for charge balance of V5+. This forms a deep narrow window of transmittance in the range 25 000--20 000 cm-1, which causes the dark-blue color. To explore the possible use of H-in-rutile as a geohygrometer, geothermobarometer, and oxybarometer, we measured the H+ content in a natural rutile crystal from a retrograded eclogite with a zoned trace element (Fe, Nb, and Zr) content. The crystal reveals a slight correlation between the variable H2O (~200 to 900 ppm) and its trace element concentrations. The observations indicate that the preservation of H+ contents in this natural rutile is a complicated interplay of diffusive reequlibration of fast H+, slower Fe and very slow other trace elements. An interpretation of the H2O contents of the natural crystal in terms of fO2 or aH2O is not possible. [ABSTRACT FROM AUTHOR]