V31D-2183
The Hydration State of Aqueous Silica at High P and T
The hydration state of monomeric aqueous silica was investigated by measuring the solubility of quartz in H2O-CO2 fluids at 800 °C and 1.0 GPa in piston-cylinder apparatus with NaCl pressure medium. Ground and polished single crystal fragments were weighed before and after a dissolution experiment on a Mettler UMX2 ultramicrobalance (1σ = 0.2 μg). Crystals and fluid ingredients (water + Ag oxalate or HM-buffered hydrous oxalic acid) were sealed in Pt envelopes by arc-welding without significant loss of volatiles, as confirmed by puncture-and-drying weight losses of quenched capsules. Run times were 24 hr. Twelve solubility determinations demonstrate that quartz solubility decreases with increasing CO2 content of the fluid phase, from XSiO2 = 2.22x10-2 in pure H2O to 1.89x10-4 at XH2O = 0.254. H2O activities (a) were calculated from Aranovich and Newton (Amer. Mineral. 84, 1319, 1999) and SiO2 activities from Newton and Manning (GCA 66, 4165, 2002). Calculated logaSiO2 decreases linearly with logaH2O, from -2.222 to -3.761 at logaH2O = 0 to -0.389. The dissolution of quartz can be expressed by: SiO2 (quartz) + nH2O = Si(OH)4·(n-2)H2O (solute monomer) where n is the total hydration state of the monomer, including 4OH- and n-2 hydrogen-bonded H2O. A plot of logaSiO2 versus logaH2O (Walther and Orville, Amer. Mineral. 68, 731, 1983) gives a slope of 4.00 (R2 = 0.9987) and an intercept of logK = -2.21, which yields ΔG° = 45.41 kJ, where ΔG° is the standard Gibbs free energy of formation of the monomeric complex from the oxides at 800 °C and 1.0 GPa. The monomeric complex thus has exactly two hydrogen-bonded H2O per formula at 800 °C, 1 GPa. This result agrees with earlier conjectures and suggests that the nature of hydrogen bonding of H2O with dissolved silica is essentially invariant over the entire range of conditions attending subduction-zone and crustal metamorphism, independent of the extent of silica oligomerization.
V31D-2184
Solubility of CePO4 and YPO4 in H2O, H2O-NaCl, H2O-NaF and H2O-Albite Fluids at 800°C and 1 GPa: Implications for REE Transport During Subduction-Zone Metasomatism
Monazite (CePO4) and xenotime (YPO4) are important accessory minerals in subducted metasediments because they host significant REE and are useful for geochronology. It is therefore essential to understand their behavior during metasomatic processes that attend subduction. In order to set constraints on the solubility of CePO4 and YPO4 at high pressure and temperature, we carried out a preliminary series of weight-loss experiments on synthetic single crystals of each phosphate in a piston- cylinder apparatus at 800°C and 1 GPa. The CePO4 and YPO4 crystals were grown from a Pb-free flux at a temperature of 1280°C for 15 hr, which was then cooled down to 870°C with a cooling rate of 3°C per hour. For each experiment, a single crystal was placed in a Pt envelope, which was added to a 3.5 mm OD Pt capsule with ultrapure H2O, H2O-NaCl, H2O-NaF or H2O-albite. Results indicate that the concentration (m, mol/kg H2O) of CePO4 (mz) and YPO4 (xt) dissolved in pure H2O is very low (mz, 0.0006 m; xt, 0.0002 m) but mz solubility increases strongly with increasing NaCl to 0.0059 m at 30 mol% NaCl. Solubility of xt only increases moderately to 0.0033 m at the same NaCl concentration, thus indicating a different solubility behavior of the two phosphates. In 10 wt% albite the solubility of both phosphates increases slightly, similar to that in a 10 mol% NaCl solution (mz, 0.0014 m; xt, 0.0011 m). Preliminary data on the solubility in a 20 mol% NaF solution indicate an enormous increase in solubility of both phosphates: mt, 0.70 m; xt, 0.75 m. These correspond to enhancements by factors of 1100 and 3000, respectively, supporting previous indications that REE mobility is strongly promoted by complexing with F in the aqueous phase. Because F is readily partitioned into subduction-zone fluids, even small amounts of this element in metasediments will yield at least local REE metasomatism during devolatilization reactions associated with subduction-zone metamorphism.
V31D-2185
Al-Si Complexing in High-Pressure Aqueous Fluids: Implications for Al Transport During Subduction-Zone Metasomatism
Low solubility of Al in pure H2O suggests that Al is immobile in metasomatic environments such as subduction zones. However, occurrences of Al-rich minerals in veins in the same settings imply significant Al mobility. To address this paradox, we investigated the solubility of corundum in fluids of varying initial SiO2/H2O ratios at 800°C and 1 GPa. We also investigated the pressure effect on the solubility of corundum in SiO2-bearing fluids at a fixed SiO2/H2O ratio. Starting materials of synthetic corundum, natural high-purity quartz, and ultrapure H2O were equilibrated in a piston-cylinder apparatus using a double-capsule method. Initial SiO2 concentrations were varied between zero and quartz saturation by adding small quartz crystals. Solubility was determined by the weight loss from the corundum crystals. The solubility of Al in pure water was 0.0028 molal, consistent with previous work. We find that at 800°C, 1.0 GPa, the concentration of Al increases with increasing dissolved SiO2, to a maximum of 0.0152 molal at quartz saturation. The data are described by the equation mAl = 0.0065(mSiO2)2 - 0.0004(mSiO2) + 0.0032, where m is the molality of the subscripted solute. The increase in Al concentration with increasing Si requires Al-Si complexing, consistent with recent predictions of Al-Si oligomerization in high P-T fluids. The concave upward trend in the isothermal isobaric data indicates that on average, Al/Si in the oligomers is <1. In addition, Al solubility rises with increasing pressure at constant T and mSiO2 = 0.675 ± 0.017 molal, according to the equation mAl = - 0.0037(P)2 + 0.0175(P) - 0.0069. The Al solubility is ~2X higher than that of corundum in pure H2O at each P. These results suggest that Al-Si complexing is a simple mechanism for enhancing Al solubility during subduction-zone metasomatism, as well as in other metamorphic settings.
V31D-2186
Peralkaline, Silica-rich Aqueous Fluids in Equilibrium with Albite, Jadeite and Quartz at 600°C, 15.5, 17.5 and 20 kbar: Implications for Subduction-Zone Metasomatism
High pressure, low temperature conditions prevail at the surface of a subducted slab, where aqueous fluids released during dehydration reactions play an important role in mass transfer. To model the composition of aqueous fluids produced near the blueschist-to-eclogite transition, we investigated the solubility of albite+quartz, jadeite+quartz in H2O at 600°C and 15.5, 17.5 and 20 kbar. Starting materials of natural, high-purity albite+quartz or jadeite+quartz were equilibrated with H2O and reagent NaSi3O6.5 (NaS3) in a piston-cylinder apparatus using a double capsule technique. Fluid compositions were determined by weight loss and phase-equilibrium bracketing methods. Both albite+quartz and jadeite+quartz dissolve incongruently in H2O leaving residual paragonite. NaS3 was added incrementally until no paragonite was observed, which allowed determination of the composition of the fluid in equilibrium with both mineral assemblages. Invariant fluid compositions (in wt%) with jadeite+quartz+paragonite at 20 kbar are 0.141 Na, 0.073 Al and 0.787 Si; at 17.5 kbar, the composition is 0.131 Na, 0.089 and 0.780 Si. The fluid composition with albite+quartz+paragonite at 15.5 kbar is 0.156 Na, 0.104 Al and 0.740 Si. Fluid compositions are very silica rich, and molar Si/(Na+Al) ratios increase with pressure from 1.84 to 2.22. Solute compositions normalized to Si=1 are Na0.329Al0.121Si at 20 kbar, Na0.330Al0.136Si at 17.5 kbar and Na0.377Al0.164Si at 15.5 kbar, illustrating that fluids are also sodium rich, and that molar Na/Al rises with increasing pressures at constant temperature. The very high solubility of albite+quartz and jadeite+quartz, and the relatively low Al in the fluid, indicate that aqueous fluids in very high P, low T metamorphism may be highly reactive, peralkaline Na-silicate solutions, with potential for substantial of mass transfer.
V31D-2187
An Experimental Investigation of the Reaction: Glaucophane = Talc + 2 Jadeite
The upper stability of blueschist facies metamorphism is constrained by the reaction: glaucophane = talc + 2 jadeite, where the breakdown of glaucophane represents the transition from blueschist facies metamorphism to a higher pressure assemblage. A previous study done by Carman and Gilbert (1983) placed the boundary at 700 °C, 3.32 GPa with a positive dP/dT slope of 2.5 * 10-3 GPa per °C, where they used gel-oxide mixtures as starting materials. Using thermodynamic data from Holland and Powell (1998) the reaction of interest has a negative slope instead, where the reaction boundary is placed at 3.61 GPa, 700 °C with a dP/Dt slope of – 8.0*10-4 per °C. Using thermodynamic data reported by Corona and Jenkins (2007) the slope of the reaction of interest is also negative, placing it at 3.76 GPa, 700 °C with a dP/dT slope of -1.0 * 10-3 per °C. To address these differences in the placement of this important boundary in P-T space, we experimentally investigated this reaction using a multi-anvil apparatus. Synthetic starting materials were use, where the starting compositions for the three phases were: talc, Mg3Si4O10(OH)2; jadeite, NaAlSi2O6; and glaucophane, (Na1.92Mg0.08)(Mg3.22Al1.78)(Al0.17Si7.83)O22(OH)2. Experiments ranged in conditions from 5.5 GPa to 2.5 GPa at 600 and 750 °C. A half bracket was observed at 3.8 GPa where the growth of jadeite and talc occurred at the expense of glaucophane. No net reaction was observed at 3.5 GPa, and we are currently defining the maximum pressure where glaucophane forms, which at this time is greater than 2.5 GPa and 750 °C via Jenkins and Corona (2006). Microprobe data indicates that both talc and jadeite deviate from their starting compositions to an average composition of (Na0.1Mg2.8)(Al0.1Si3.9)O10(OH)2 and (Mg0.10Na0.82)(Mg0.10Al0.90)Si2O6, respectively. The average composition of glaucophane deviates to: Na0.14(Na1.64Mg0.36)(Mg3.23Al1.77)(Al0.28Si7.72)O22(OH) 2. Adjustments for solid solution in the talc, pyroxene, and amphibole would suggest the ideal boundary lies at 3.63 GPa, 700 °C which is in agreement with the boundary derived with Holland and Powell's values. Current research continues to define the slope of this boundary.
V31D-2188
The estimation of the age of serpentinization by the SHRIMP dating of zircon from the rodingite vein and zoisitite in serpentinite of the Nagasaki metamorphic rocks
Serpentinization does not generally produce minerals suitable for direct isotopic dating. Therefore the age of serpentinization is usually estimated on the basis of age of other rocks associated with serpentinites. In this study, we report the ion probe dating of zircon in metasomatic lithologies such as rodingite veins and zoisite- rich rocks, which are considered to have formed at the same time as serpentinization. Zoisitites and rodingites from the Nagasaki Metamorphic Rocks, western Japan, are high-pressure-low-temperature metasomatic rocks that occur as tectonic inclusions in a serpentinite-matrix melange. The reaction zone is present between rodingite and serpentinite, with sharp contacts. The serpentinite mainly consists of antigorite with minor chrysotile, diopside, spinel, pyrrhotite, and chlorite. Chrysotile occurs near the reaction zone and also occurs as veins in serpentinite. Zoisitite occurs as blocks several meters across, consisting of granular clinozoisite, diopside, chlorite, apatite, titanite, and zircon. Rodingite occurs as veins in serpentinite. A large number of rodingite dykes intrude the serpentinite. The rodingite veins are centimeters to tens of centimeters wide and several meters in length. Rodingite is pinkish, mostly consisting of garnet, vesuvianite, diopside, apatite, titanite, and zircon. On the basis of the field observation and petrology, we infer that zoisitite, rodingite and antigorite-serpentinite formed at the same time. The reaction zone formed during the latter period of serpentinization when the chrysotile was produced. To characterize mass transfer during the serpentinization, major and trace element bulk analysis was performed. The zoisitite and rodingite have similar geochemical characteristics. We infer that elements released by serpentinization are Fe, Ni, Zn, V, Cr, Mg, Sc, Co, and Cu. On the other hand, Ca, Al, Mn, Ti, Sr, Zr, Nb, LREE, Bi, Th, and U were enriched in serpentinite by the fluid. Zircon grains are found in the zoisitite, rodingite, and the reaction zone. Two types of zircon were found in these rocks. Zircons 10-160 microns long are characterized by irregular margins and magmatic cores. Other grains <10 microns across are anhedral, mostly with homogeneous, high cathodoluminescence, although some grains have complex textures. The zircon grains are interpreted as follows: 1. zircon with magmatic features, such as oscillatory zoning, formed in protoliths, probably gabbroic; 2. zircon with patches and crack-fillings of high-CL were modified from magmatic zircon during the early stages of serpentinization and rodingite - zoisitite formation, by partial dissolution and recrystallization; 3. new sub-10 micron grains, together with overgrowths of homogeneous high-CL zircon, grew along fractures from hydrous fluids. Few magmatic grains with recrystallized areas and high-CL overgrowths could be dated by SHRIMP at NIPR. The weighted mean of common-Pb corrected U-Pb ages from 3 grains in zoisitite is 109.2±2.2Ma, with no resolvable age difference between magmatic cores and high-CL overgrowths. Age data from 4 grains in rodingite are scattered between 107 and 97Ma, with younger ages from spots overlapping porous annealed cracks. Age data are tentative but suggest that serpentinization occurred only shortly after crystallization of the gabbroic protolith.
V31D-2189
Rehydration Metamorphism of the Iratsu Eclogite Mass in the Sambagawa Belt, Japan
The Iratsu eclogite mass in the Sambagawa metamorphic belt, central Shikoku, Japan, underwent pervasive rehydration metamorphism so that most of all retrogressed to amphibolite. The Iratsu eclogite mass is originated from gabbro and basalt, and is interpreted as a relic of a fossil subducted slab. Gradual stages of rehydration retrogression from eclogite to low-grade greenschist are preserved by disequilibrium textures of incomplete reactions. Hence, the Iratsu eclogite mass provide an exceptional opportunity to study the process of rehydration reactions and exhumation of a subducted oceanic crust. The petrological investigations showed that the retrogressed eclogites can be classified into two types by reaction textures during rehydration stage: One is the perfectly equilibrated type, which is a rock almost perfectly retrogressed into well-foliated epidote-amphibolite consists of amphibole + epidote + chlorite + plagioclase + quartz +/- garnet, mainly sampled from the rim of the Iratsu body. The other is the locally equilibrated type, which is a rock partially retrogressed rock often preserved garnet and omphacite and weakly foliated sampled from the central part of the Iratsu body. Garnet grains are rimmed by amphibole (pargasite) + epidote layers, whereas omphacite grains broke into symplectites of amphibole (hornblende) + plagioclase. Such clear distinction of the mineral assemblages and mineral compositions between two parts around the dissolving minerals demonstrates that the system was not equilibrated as a whole but locally equilibrated in each part around the dissolving minerals. Partial pseudomorphs, which are frequently observed in rocks of the locally equilibrated type, are the most reliable evidence of a reaction, since the replaced mineral and the solid products can be directly observed. In this study, we developed the new methodology based on a simple mass-balance relation, by improving Gresens 1967 and Godard & Mabit 1998. Our method can easily determine the stoichiometric relation of the reaction, the reaction progress and the material transfer including H2O from the EPMA compositional images of a partially pseudomorph. By applying this method to pseudomorphs after garnet and omphacite in the whole area of the Iratsu eclogite mass, we found that the extent of the rehydration reactions, the associated material transfer and the extent of deformation increase gradually from the center to the rim of the mass. These results clearly indicate that the rehydration metamorphism was caused by the infiltration of external fluid into the Iratsu eclogite mass.
V31D-2190
Did the Himalayan eclogitization occur before or after the India-Asia collision?
Eclogites exposed in the Kaghan Valley of Pakistan are divided petrologically into two groups as high- pressure (HP) and ultrahigh-pressure (UHP) eclogites. Group I (HP) eclogites are composed of garnet + clinopyroxene + quartz + amphibole ± phengite ± epidote/allanite + rutile ± ilmenite with abundant zircon grains of large size (> 250 µm), considered as derived from a gabbroic protolith. Group II (UHP) eclogites are composed of garnet + clinopyroxene + coesite + epidote + phengite + amphibole + rutile/ilmenite with rare zircon grains of small size (< 50 µm), considered as the products of tholeiitic basalts. Whole-rock multi-isotope systems indicate considerable trace element mobility or fluid infiltration during subduction of the Indian plate under the Asian plate. Due to this mobility or fluid infiltration, the Sm-Nd, and Lu-Hf isotope systems do not define a geologically meaningful age. The very low Sm/Nd (< 0.6) and Lu/Hf (< 0.4) ratios of garnet and the variable two- or three-point Sm-Nd and Lu-Hf isochrons/mixing lines indicate mineral disequilibrium. Two generation of zircons are found in these eclogites. First generation zircons are of magmatic origin. They are large and concentrically zoned. Second generation zircons are of metamorphic origin. They are small and or occur as overgrowths along the magmatic portions. The U-Pb concordant age of 267 ± 2 Ma, from the core portion of zircon (magmatic) from Group I eclogites, is assumed to be the eclogite protolith age, whereas, the 71 ± 6 Ma from the rim portion of zircon (metamorphic overgrowth) provides evidence that the eclogite protolith subducted before the India-Asia collision, and the UHP event took place afterwards when these rocks reached to depth more than 90 km.
V31D-2191
Trace Element Mobility and Nd-Sr-Pb Isotopes in the High-Grade Metamorphic Rocks of the Franciscan Subduction Complex: Evidence for an Arc Protolith
Geochemical and isotopic data from our previous study for one high-grade Franciscan locality (Ring Mountain) indicate a nascent island-arc basalt origin for these high-grade rocks, similar to that proposed for the Coast Range Ophiolite (CRO). In this study we report multiple trace elements and Sr-Nd-Pb isotopes of high-grade coherent sheets and tectonic blocks, and low-grade coherent sheets from 13 localities of the Franciscan Complex along a length of ~250 Km. The high-grade rocks, the oldest metamorphic rocks of the Franciscan (154-169 Ma metamorphic ages) show conspicuously flat REE similar to island-arcs although some of these rocks show slight depletions in LREE. The low-grade rocks (metamorphosed at or after 135 Ma) show prominent LREE depletions similar to that of MORB. The overall trace element and REE patterns of these rocks are distinctly different from Ocean Island Basalts and continental crust. High-grade rocks of this study also show high Ba and Pb concentrations and high Ba/Rb, Ba/Th, U/Th, U/Nb, La/Nb ratios. A strong negative Nb anomaly is observed and Ce/Pb ratios are low. In comparison, low-grade rocks only mildly exhibit some of the typical arc-like trace element patterns mentioned above and clearly display a different set of trace element concentrations and ratios compared to the high-grade blocks. There is a marked difference between the high-grade and low-grade rocks in their Nd-isotopic ratios that strongly indicates different protoliths for these two rock groups. The low-grade rocks have distinctly higher εNd(I) (>10) values similar to MORB while the high-grade rocks have lower εNd(I) (4 to 9.8, majority ~6) values similar to global arc tholeiites. Our data show that none of these rocks underwent geochemical exchange with fluids sourced from continentally derived sediments, so the observed geochemical signatures reflect the protoliths rather than being an artifact of metasomatism and metamorphism. The Franciscan Pb-isotopic data of this study are similar to the Izu-Bonin and Mariana arcs even though they also fall within the field of Pacific MORB. All these are similar to those of the Western Pacific arc. The low- grade coherent schists plot closest to the Northern Hemispheric Reference Line with Pb isotopes similar to both Pacific MORB and Izu-Bonin Arc. Our isotopic and trace element data thus indicate an arc tholeiite origin for the high-grade rocks that were subducted early in Franciscan history versus a MORB origin for the low-grade rocks that were subducted later. There is no evidence of subducted ridge components in our geochemical data of the high-grade Franciscan rocks. Because coherent high-grade rocks of the Franciscan must have been derived from the subducting plate, the arc tholeiite protoliths for the first-subducted Franciscan metamorphic rocks demand the existence of a pre-Franciscan subduction zone to have generated them. The geochemical similarity between the high- grade rocks and the unmetamorphosed CRO that structurally overlies the Franciscan suggests that the CRO and protoliths for the high-grade rocks formed over the same subduction zone. Franciscan subduction began shortly afterwards within this arc crust.
V31D-2192
Constraints on early Franciscan subduction rates from 2-D thermal modeling
Mesozoic subduction along the western margin of North America in California jumped west in mid-Jurassic time to form the Franciscan accretionary wedge outboard of the Coast Range ophiolite and the Great Valley forearc basin. Overall rates of subduction are likely to have been of the order of 100 mm/yr, based on Mesozoic plate motion analysis for the Pacific region. New and previously published Lu-Hf garnet growth ages on high-grade blocks in the Franciscan, however, suggest that early Franciscan metamorphism took place in a slowly cooling environment, characteristic of slow subduction rates. In order to assess rates of early Franciscan subduction, 2-D thermal modeling was carried out based on the following assumptions: (1) Subduction was initiated at 170 Ma. This is based on the age of the oldest tectonic block (a garnet amphibolite dated at 168.7 ± 0.8 Ma), and on the age of the Coast Range ophiolite, believed to be a supra-subduction zone ophiolite formed in a forearc setting behind the newly initiated subduction zone. (2) At the start of subduction the hangingwall mantle wedge was at asthenospheric temperatures, consistent with initiation of subduction at the ophiolite spreading center. (3) The earliest subducted lithosphere was 10 m.y. old at the start of subduction. (4) New thermobarometric data from the dated blocks suggest that they were metamorphosed at pressures between 1200 and 1800 MPa, but uncertainties are large. We therefore assume that they were metamorphosed at an average pressure of 1400 MPa, corresponding to a depth of 42 km beneath mantle rocks. We investigated subduction rates between 1 and 100 mm/yr, and found that the best fit between the ages and estimated metamorphic temperatures of the blocks falls between 5 and 7 mm/yr. We cannot explain the oldest (garnet amphibolite) block by this mechanism, however, as its age coincides with our assumed time of subduction initiation. We suggest that the garnet amphibolite blocks in the Franciscan may have formed from crystallizing basaltic melts at depth within the hangingwall mantle wedge during formation of the Coast Range ophiolite, and were subsequently partly re-equilibrated at lower temperature during subduction.