V12A-01
Near-Solidus Solubility of Albite+Quartz in H2O at 1 GPa: Implications for Crustal Magmas and Fluids
The solubility and melting of albite (Ab, NaAlSi3O8) ± quartz (Qz, SiO2) in H2O has been used to gain insight into the nature of magmas and metamorphic fluids in the continental crust. For example, the Ab+H2O and Ab+Qz+H2O solidi are widely used as guides to crustal melting. However, some studies suggest that Ab dissolves incongruently in H2O at near-solidus conditions, which implies that these solidi are not likely to be encountered in nature. To address this problem, we carried out experiments on the stability and solubility of the assemblage Ab+Qz in H2O at 1 GPa, 580-640°C. Starting materials were natural Amelia Ab and Brazilian Qz. Ab crystals were placed in a 1.6 mm OD Pt capsule, which was lightly crimped and then loaded with ultrapure H2O and a quartz crystal into a 3.5 mm OD Pt capsule, which was sealed by arc welding. Experiments were conducted for 24 h in a piston-cylinder apparatus using 1" NaCl- graphite furnaces. Quenched run products were evaluated to assess stable assemblages, and solubilities were determined from crystal weight changes. Hydrous melt (L) became stable at 630±10°C. Paragonite (Pg, NaAl3Si3O10(OH)2) was present in all run products, indicating that Ab+Qz and L+Qz dissolve incongruently in H2O, and that Ab+Qz+L+H2O is metastable at 1 GPa. The bulk solubility of the assemblage Ab+Pg+Qz in H2O doubles from 4.0 to 8.0 wt% as the solidus is approached (580- 630°C), with corresponding increases Si, Al, and Na molalities. Over the same T range, Na/Al and Na/Si (molar) respectively decrease (2.1 to 1.5) and increase (0.153 to 0.175), indicating that fluids are Na- and Si-rich peralkaline solutions far from Ab stoichiometry. At 580°C, measured Al solubility agrees well with that predicted from extrapolation of thermodynamic data for Al and Na-Al species; however, measured Al solubility at the solidus is much greater than predicted. Similar observations hold for Na and Si. This implies that there is a strong, pre-melting increase in the extent of polymerization of aqueous species in the narrow 50°C interval immediately below the solidus. Our results illustrate that phase relations in the system Ab-H2O and Ab-Qz- H2O are more complex than previously appreciated, and that polymeric aqueous species involving Na, Al, and Si are predominant in these model near-solidus and melt-saturated crustal fluids.
V12A-02
High Pressure Granitic Melt in Eclogite: Is it Internally Derived?
The Erzgebirge region of Southeast Germany is a dome-shaped structure that contains metamorphic rocks ranging from low-gade units at the periphery to eclogites and diamond-bearing granulites at the center. This study focuses on an eclogite locality on the shore of the Saidenbach reservoir, part of the highest-grade metamorphic core. The eclogites contain abundant K-feldspar-rich quartzofeldspathic material which we interpret to be a former melt. The quartzofeldspathic material is finely dispersed at the grain scale. However, the material commonly forms segregations, lenses, and veins up to several centimeters in diameter with no preferred orientation. The quartzofeldspathic material was in equilibrium with an upper amphibolite assemblage (overprinting the peak eclogite assemblage) that occurred around 800°C and 18 kbar (a pressure minimum for melt formation). The most noteworthy feature of this migmatitic eclogite is the granite composition of the segregations. Despite experiments which show that high pressure eclogite melting forms granites at low melt volumes (e.g., Schmidt et al, 2004), field observations associate eclogite melting with the formation of trondhjemites and tonalites. Thus, by understanding why granite and not trondhjemite formed within these eclogites, we can learn more about what variables control the composition of eclogite melts at high pressures. The origin of the melt can be attributed to one or more of the following processes: 1) it could have formed in the surrounding country rocks and pervasively infiltrated the eclogite; 2) it could have formed within the eclogite as a closed system, requiring a high potassium basaltic protolith; or 3) it could have formed within the eclogite as the result of an infiltrating potassium-rich fluid. The dispersal of small volumes of feldspathic material throughout the host rock strongly supports an internal origin for the granitic melt, and the divergence of trace element trends from expected eclogite-derived melt compositions suggests at least some degree of externally driven metasomatism. Isotope and trace element data will identify common end-members, enabling us to pinpoint which of the abve processes were responsible for the melt generation. By extension we hope to elucidate the nature of fluid and melt transport in high pressure metamorphic systems such as subduction zones.
V12A-03
Fluid evolution of the Varberg-Torpa charnockite-granite intrusion, SW Sweden: Magmatism and metamorphism on a regional scale
The 1.4 Ga Varberg-Torpa charnockite-granite intrusion (Varberg, SW Sweden) consists of the magmatic Varberg charnockite (Opx-Cpx-Bt-Amph-Gt-Plg-Kfs-Qtz; accessory FAp, Zrn, Mt, Ilm, Py, Cp, ±Po, ±Rt), with granitic inlyers, accompanied by in-situ, patchy, fluid-induced, dehydration of local, amphibolite-facies granitic gneisses to charnockite in the vicinity of the intrusion, and the Torpa granite that is both continuous and synmagmatic with the Varberg charnockite, has a similar whole rock chemistry and mineral assemblage (minus the pyroxenes), and contains several charnockite enclaves. P-T estimation, using Gt-Opx Fe-Mg exchange thermometry and Gt-Opx- Plg-Qtz barometry of both the igneous and metasomatically derived charnockite, indicates temperatures of 650 to 700 oC and pressures of 750 MPa during emplacement and crystallization of the charnockite-granite intrusion. The earliest recognized fluid inclusions in both the granite and charnockite consist of H2O- CO2 mixtures (DF = 0.2 to 0.7). Fluid inclusions in the charnockite are characterized by high partial CO2 densities (up to 1.0 g/cm3 for DF=0.7; 40 to 90 % bulk CO2 with minor/no CH4 and/or N2); are of possible magmatic origin; and are best preserved in garnet, plagioclase, and fluorapatite (in order of decreasing CO2 densities). Fluid inclusions with the highest CO2 densities (1.08 to 1.10 g/cm3) are found in quartz (Th -31 to -36 oC) and may have originated under granulite-facies conditions. Magmatic fluids in the granite correspond to aqueous-carbonic inclusions with an estimated bulk composition (mol%) of H2O(73)CO2(25)NaCl(2). The salinity of solutes in the granite is generally higher than for the charnockite (typically 14 to 20 wt% NaCl-eq.). Field, petrographic, mineralogic, geochemical, and fluid inclusion evidence suggests that charnockite vs. granite crystallisation is more a function of relative H2O activity as opposed to relative depth. In those regions of the magma, which crystallized out as charnockite, a preponderance of CO2 lowered the H2O activity thereby allowing for the formation of Opx and Cpx.
V12A-04
Reaction-Enhanced Permeability in Gabbroic Crust, IODP Site 1309, mid Atlantic Ridge
Thermal cracking is normally assumed to be the main mechanism of permeability generation in the oceanic crust. Here we present microstructural evidence that metamorphic reactions played a significant role in creating porosity and permeability in gabbroic rocks beneath a detachment fault at 30o N in the Atlantic ocean. At least two mechanisms for permeability generation have been identified: 1) In basaltic intrusions, euhedral zoning in amphibole replacing clinopyroxene suggests a dissolution/precipitation mechanism similar to textures observed in epidosites from the Troodos ophiolite. The basaltic sills lack macroscopic fractures and alteration patterns suggest a fingering instability allowed fluid to enter the rock. 2) Volume increase reactions in olivine gabbros and troctolites. These generated radial crack networks filled with secondary minerals where olivine was surrounded by other phases, and are seen both in serpentinization reactions and in reaction between olivine and plagioclase forming tremolite-chlorite coronas. Discrete-element modelling shows how isolated olivines can be linked by propagating networks of cracks allowing fluid to access initially impermeable crust. Assuming that most fluid flow occurs along major fractures, reaction permeability provides an effective mechanism for fluid to access intervening unaltered rock, and may be important for scavenging ore-forming components from the crust. The most permeable part of the system is likely to be at the reaction front before precipitating minerals occlude the porosity, and this may promote front-parallel fluid flow.
V12A-05
Experimental Evidence for High-Pressure Phase Separation in the H2O-CO2-CaCl2 System: Implications for Rock Rheology
As part of a study to examine the effect of CO2 on deformation mechanisms in quartz, axial compression experiments were carried out at 900°C and 1500 MPa on cores of Black Hills quartzite (BHQ) with a layer of dolomite powder (± 0.05 wt% H2O) in the center of each charge (some runs included buffer assemblages at sample ends). All runs released CO2 via the reaction dol + qtz = diop + CO2 during run-up to experimental conditions. BHQ starting material contains three types of naturally occurring fluid inclusions (FIs): pure H2O, H2O + 6-18 wt% CaCl2, and pure CO2. Deformation experiments on as-is BHQ (no dol powder) result in destruction of most optical FIs. In contrast, experiments with wet dol powder produced visible FIs in nearly all samples, though most were too small to analyze by microthermometry. One hydrostatic experiment with dolomite generated FIs up to 15 microns across near the reaction zone. FIs within this sample fall into two types: (1) superdense CO2 (homogenization to liquid below -50°C), and (2) H2O-CO2-CaCl2 solutions with variable X(CO2) and bulk density and up to 40 wt% CaCl2 (referenced to aqueous phase only). Both inclusion types occur within the same clusters, and likely result from interaction of CO2 released by dol breakdown with H2O and FI fluids released from the starting material. Isochores from the Type 1 CO2 FIs record pressures of 1200- 1400 MPa at 900°C. Estimation of bulk density for Type 2 FIs is hampered by complex microthermometric behavior and incomplete equation of state data for this fluid system, but model isochores overlap with those of Type 1 FIs at 900°C. Entrapment of the two types of FIs and variable phase proportions in Type 2 inclusions are consistent with fluid phase separation at experimental conditions. Deformation experiments run at f(O2)<graph-CO2 show significantly less strain close to the dol reaction zone, where CO2 inclusions are most abundant, relative to sample ends where CO2 was reduced to graphite. CO2 reduction may move rocks out of the two-fluid field and result in an increase in both a(H2O) and f(H2O), which in turn will facilitate strain accommodation by dislocation and/or diffusion creep. Shmulovich & Graham (2004 CMP) documented a large two-fluid field in the H2O-CO2-CaCl2 system at 800°C and 900 MPa. Our study shows that the region of immiscibility extends to higher pressures and temperatures, and can be anticipated in both lower crustal and upper mantle rocks as well as in subducting slabs. Our data also demonstrate that variations in f(O2) may be generated and preserved over short distances, and that strain accommodation mechanisms can co-vary with fluid composition, f(O2) and f(H2O) in COH-salt systems. We thus anticipate that significant localized differences in rock strength will result from metamorphic reactions that move rocks into and out of the two-fluid field at high pressure.
V12A-06
CO2 Strengthening of Quartz in the Dislocation Creep Regime?
Trace amounts of water have long been known to significantly reduce the strength of quartz deforming by dislocation creep, but the effect of carbonic fluids has not been documented experimentally. In naturally deformed rocks, however, there are indications that carbonic fluids act to strengthen quartz-rich rocks. In schists from the eastern Alps, Selverstone (2005, JMG) observed that quartz in graphitic layers with carbonic fluid inclusions (FIs) deformed in a brittle manner whereas quartz in closely adjacent non-graphitic layers with aqueous FIs deformed by dislocation creep. We have conducted an experimental study to investigate the effect of carbonic fluids on quartz deforming by dislocation creep. Constant displacement rate, axial compression experiments were conducted at 900°C, 1.5 GPa, and 10-5/s in a modified Griggs apparatus using cylinders of as-is Black Hills quartzite (BHQ) encapsulated in Pt (inner) and Ni (outer) jackets; all samples were shortened 50%. BHQ is a pure (<1% feldspars, clays, and Fe oxides), quartzite with equant grains (average diameter ~100 μm); it contains three types of FIs (carbonic, H2O + NaCl, and H2O + CaCl2) and has an average bulk water content of 0.09 wt % as measured by FTIR. During deformation and recrystallization, most of the original optical-scale FIs appear to be expelled to grain boundaries. In order to test the effect of carbonic fluids, CO2 was generated in some experiments at P and T by the reaction of quartz + dolomite goes to diopside + CO2. In these experiments a ~140 μm thick layer of dried, mixed, fine-grained (~5-20 μm diameter) quartz and dolomite powder was placed between two short cores of BHQ. Some samples were held at P and T for 15 hours and some for 85 hours, before starting the deformation. The amount of CO2 produced by complete reaction is ~0.005 g, or 0.5 wt %. The formation of the diopside layer did not appear to affect the sample deformation. The bulk strength of samples with added CO2 was similar to or greater than that of as-is samples of BHQ deformed at the same conditions. However, the distribution of deformation was strikingly different. As-is samples are very homogeneously strained, consistent with our very low T gradients. In contrast, the samples with CO2 produced from the central reaction layer have an hourglass shape; the region close to the reaction layer has abundant CO2 FIS and is almost unstrained, whereas the regions further away, toward the ends of the sample, have very few CO2 FIs and are highly strained and recrystallized. Thus it appears that the areas with higher CO2 were stronger. It is not yet clear whether the CO2 FIs prevent grain boundary migration recrystallization and/or whether the CO2 decreases the water fugacity.
V12A-07
Growth of Fibrous Talc and Anthophyllite in the Hydrothermal Diamond Anvil Cell (HDAC)
Talc deposits are formed by metasomatism of dolostone or ultramafic rocks during metamorphism or hydrothermal alteration (Van Gosen et al., 2004: Env. Geo., v. 45, p. 920). Although talc is common, fibrous (asbestiform) talc is rare and its origin has been explained by the replacement of asbestiform amphiboles during retrograde metamorphism. The effects of fibrous talc on the human respiratory system remain unresolved. In order to elucidate the formation of asbestiform talc in natural systems, kinetics experiments in the MgO-FeO- SiO2-H2O-HCl system were performed by using an hydrothermal diamond anvil cell (HDAC) apparatus. Two different chamber sizes have been used, with either 650 μm or 300 μm diameter holes in 250 μm thick rhenium foil. We have observed that at elevated pressures and temperatures, deformation of the gasket occurs and the chamber volume is reduced by up to 25%. Forsterite (Fo95) + quartz + 0.1 mM HCl solution (pH = 4 @ 25oC and 0.1 MPa) were added to the chamber and run at temperatures of 560 - 660oC and at pressures up to 1 GPa for a duration of 1 - 8 hours. The experiments were run isothermally (+/- 3oC) at the target temperature. The experiments were recorded by a video camera mounted on a polarized light microscope. Experiments have been performed at pressures and temperatures near the talc + enstatite = anthophyllite equilibrium. Run products consist of talc or anthophyllite with a fibrous habit. Vapor-liquid homogenization temperatures before and after the experiment were used to determine the prograde and retrograde isochores, respectively. Quartz was dissolved upon heating by as much as 5%; dissolution occurred on a time scale of ~ 30 seconds. Whiskers of fibrous anthophyllite crystallized on the high temperature side of the talc + enstatite = anthophyllite equilibrium whereas fibrous talc crystallized on the low temperature side. Composition of the whiskers was determined by using a spindle stage and immersion oils. The talc and anthophyllite whiskers share similar morphological characteristics (aspect ratios ~25:1, average lengths ~40 um) and grew on the forsterite surface and/or on the gasket walls. Growth rates for both minerals are on the order of ~ 0.03 μm/s in the long dimension. Growth of fibrous talc and fibrous anthophyllite may be due to: (1) strong spatial gradients in the aqueous concentration of SiO2 during growth; (2) fluctuations in pressure due to changes in cell volume brought on by gasket deformation. The production of fibrous talc in these experiments demonstrates that growth of talc after amphibole is not necessary for the production of a fibrous habit. The results of this study may be used to determine the conditions under which fibrous talc may form in some deposits.
V12A-08
Growth of Diamond from a Carbonaceous Hydrous Silicate Melt: An Experimental Study
Diamond was grown in a hydrous halide-bearing silicate system at pressures and temperatures near those at which natural peridotitic diamonds form (1000-1300° C, 5-7 GPa) in the Earth' s upper mantle. The mechanisms by which diamonds are made within the earth is still unresolved, and many authors have suggested possible media from which diamonds precipitate; examples include mantle carbonates, sulphides and silicates (1-3). To date, little work has been conducted on silicate melts and the effect of mantle catalysts on diamond formation. This study used a hydrous silicate melt (HSM) to attempt to precipitate diamond. The primary experimental system was MgO-SiO2-C-H2O, with subsystems to document the addition of alkali halides (KCl and NaCl). Previous studies have concluded that alkali halides have a catalytic effect on diamond formation reactions and observed halides in inclusions in natural diamonds. Diamond was successfully grown on seed crystals at temperatures of 1400-1500° C and pressures of 6-7 GPa, in 3-4 hours. No spontaneous nucleation of diamond was observed during these experiments. No diamond growth was observed in experiments at < 1400° C and 6 GPa to date. The addition of KCl to the HSM system allowed diamond to form 200° C lower than the previously published minimum temperature of over 1600° C (3). The effect of NaCl and other catalysts are still under investigation. The starting compositions contain ~ 20.6wt% structural H2O. At run conditions, a hydrous melt coexisted with olivine and orthropyroxene, with the halides either dissolved in the melt or forming a separate brine. This study demonstrates that hydrous silicate melts, especially containing alkali halides, are a viable medium for diamond growth in the Earth' s upper mantle. 1) Arima et al, 2002; 2) Gunn & Luth, 2006; 3) Pal' yanov et al, 2007