V34A-01 16:00h
Volatile Contents in Mafic Magmas from two Aleutian volcanoes: Augustine and Makushin
There are several competing theories for the origin of tholeiitic (TH) vs. calc-alkaline (CA) fractionation trends in arc magmas. One relates to water (TH-dry magma, CA-wet magma), another to pressure (TH-low pressure crystallization, CA-high pressure), and a third to primary magma composition (TH-low Si/Fe\#, CA-hi Si/Fe\#) These theories have been difficult to test without quantitative measures of the water contents and pressures of crystallization of arc magmas. We are in the process of studying several Aleutian arc tephra suites (phenocrysts and melt inclusions) with the aim of obtaining volatile element concentrations (by SIMS), major and trace element concentrations and thermobarometric data (by EMP and laser-ICPMS). We report preliminary results on olivine-hosted melt inclusions from Augustine and Makushin volcanoes that support the role of water in calc-alkaline fractionation. Basaltic melt inclusions from Augustine, a low-K$_{2}$O, calc-alkaline volcano, are hosted in Fo$_{80-82}$ olivine. The inclusions yield high water contents, up to 5 wt%, and contain 60-90 ppm CO$_{2}$, 3000-4500 ppm S, and 3000-6000 ppm Cl. Inclusions record vapor-saturation pressures near 2 kbar. Cl/K$_{2}$O ratios in Augustine inclusions (ave. 1.9) are among the highest documented in an arc setting, and likely record a Cl- and H$_{2}$O- rich fluid from the subducting plate. High water contents in Augustine primary melts may have contributed to the strong calc-alkaline trend observed at this volcano. Basaltic melt inclusions from Pakushin, a medium-K$_{2}$O, tholeiitic cone on the flanks of Makushin volcano, are hosted in Fo$_{80-86}$ olivine. These inclusions have low water contents ($<$0.15 wt%) and low CO$_{2}$ contents ($<$125 ppm), and record shallow vapor saturation pressures ($<$300 bars). The high sulfur (2000-4000 ppm) and Cl ($>$2000 ppm) in Pakushin melt inclusions, however, indicate that degassing was minimal. The low water contents and low vapor saturation pressures recorded in Pakushin melt inclusions are consistent with development of its tholeiitic trend, but we cannot distinguish whether the low water contents at Pakushin reflect a difference in the primary magma, or in crustal stresses that favor low pressure fractionation and degassing. These questions will be addressed with further work on the trace element contents of the melt inclusions and on phenocryst-melt geobarometers.
V34A-02 16:15h
The effect of iron on the chlorine concentration in felsic melts
The concentration of Cl in hydrous magmas plays a role in a variety of processes including the timing of volatile exsolution and the mass transfer of metals from melt into an exsolved Cl-bearing volatile phase. Literature data indicate that the concentration of Cl in hydrous felsic melts increases with the activities of network modifying cations (Na+K+Fe+Ca) at the expense of Si (Metrich and Rutherford,GCA,1992). Cl in melt is correlated positively with Na, K, and Fe, negatively with Si and is independent of Al. Synthetic Fe-free peralkaline melts (Al/(Na+K)<1) can contain higher Cl concentrations than peraluminous melts (Al/(Na+K>1); Cl can increase four-fold with increasing melt peralkalinity (i.e., from 2000 to 8000 ppm as ASI decreases from 1 to 0.4). In natural Fe-bearing felsic melts, Cl in melt can be >2000 ppm higher than in Fe-free melts at an identical ASI suggesting the potential for Fe-Cl complexation in the melt. Additionally, one study has postulated a relationship between increasing Cl concentration and decreasing Fe3+/Fe2+ ratio in felsic melts (Webster and DeVivo, AmMin,2002). We have performed a series of experiments (n=20) in Mg- and Ca-free, Fe-bearing system (magnetite (mt)+haplogranite melt+vapor+brine) that definitively link increasing concentrations of Fe and Cl in felsic melts. Natural mt was reacted with synthetic haplogranite melt and an HCl-, NaCl- and KCl-bearing aqueous vapor -O brine (initial molar K:Na=1 and K:H varied from 1 to 25) in Au and Pt capsules held inside traditional and rapid-quench cold-seal vessels at 800,aC, pressures of 100 to 145 MPa and bulk salinities ranging from 1.8 to 20 wt% NaCl eq. The oxygen fugacity of the experiments was fixed at either nickel-nickel oxide (NNO) or quartz-fayalite-magnetite (QFM). Varying the HCl concentration, imposed by varying the initial K:H of the fluid, allows us to control the final melt aluminosity, with ASI's ranging from 0.5 to 1.1. When melt concentration data for all cations are regressed against Cl, the positive relationship between Fe and Cl is significantly greater than that between Cl and all other cations. In slightly peralkaline melts (ASI=0.94),a negative correlation obtains between Cl and both Na and K. Runs of short duration, yielding heterogeneous melt compositions, display the most dramatic absence of a systematic relationship between ASI and Cl; in some runs there is a negative correlation between Cl and increasing melt peralkalinity. These data imply that Fe, and not ASI, is the fundamental control on Cl content in the melt. The concentration of Cl in melt is significantly higher in melts produced at QFM (up to 0.6 wt% Cl) relative to NNO (up to 0.3 wt% Cl) and this is directly correlated to the increase of Fe in melt at QFM (up to 4 wt% Fe) relative to NNO (up to 3 wt% Fe). The strong positive correlation between Fe and Cl concentrations at oxygen fugacities ranging from NNO to QFM indicates that the Fe content of water-saturated, peralkaline to peraluminous aluminosilicate melts plays a determinant role on Cl concent in the melt. These experimental results are broadly consistent with the work of Webster and De Vivo (2002) who suggest the timing of iron-bearing mineral crystallization in the melt plays a critical role in controlling the absolute Cl concentration in the melt. Our findings have important implications for magmatic degassing, hydrothermal alteration and the mass transfer of ore metals from melt to the volatile phase; the latter being critically dependent on the partitioning of Cl between melt and volatile phase.
V34A-03 INVITED 16:30h
Controls on Transition Metal Concentrations in Crustal Brines
Experimental studies of mineral solubilities have systematically explored the effects of pH and other parameters on metal concentrations over relatively narrow temperature ranges. This study has compiled a data base of brine analyses, ranging from low temperature shield and formation brines to magmatic brines, including geothermal and metamorphic brine analyses. The data includes both analyses of samples from drilling, and fluid inclusion analyses, and there is a span of over an order of magnitude in chloride concentration. Concentrations of Fe, Mn, Zn and Pb vary systematically across the entire data set, and the principal controls on their concentrations are salinity and temperature. In each suite of analyses in the data set, metal concentrations increase linearly with Cl over the entire salinity range, with a slope of between 1 and 1.5 in log mol units. For Fe and Mn in all the data sets, Me/Cl remains nearly constant over a wide range of salinities at constant temperature, but there is almost 6 orders of magnitude variation in Me/Cl between low-T formation brines and magmatic brines. Larger scatter in the Fe data may be attributed to variations in redox, and correlates with Mn/Fe. The slope of the data array on a Zn-Cl plot may be somewhat higher for formation waters than for magmatic fluids, indicating a possible change in complexing with temperature, but at no temperature is there evidence for a change in complexing with Cl concentration. Pb data is sparse but shows similar trends, though with less dependence on temperature. The continuity in crustal brine chemistry from sedimentary to metamorphic and magmatic fluids demonstrates the importance of wall rock buffering for the control of crustal fluid composition, and shows that the variation in pH, {\it f}S$_{2}$ and redox environment between different lithologies is not sufficiently large for variation in these parameters to dominate the variation in metal contents of fluids. In contrast, temperature and salinity emerge as the major controls on transition metal contents of crustal fluids, with little evidence for significant changes in chloride complexing over the range of compositions and conditions encountered in the crust. Preliminary results on low salinity, gas-rich fluid inclusions suggest that their transition metal contents fit on the same trends as those of brines.
V34A-04 16:45h
Experimental determination of the critical PVTX properties of H$_{2}$O-NaCl-KCl-CaCl$_{2}$ brines
Near-critical and supercritical fluids are of great scientific interest because of their unique thermodynamic and physical properties, such as anomalously high heat capacity, large compressibility, low viscosity versus density ratio, high diffusivity and light scattering. Geochemists - among other scientists - have long studied these characteristics to understand the role of critical and supercritical fluids in various geological environments. Such fluids play an important role in numerous geologic processes. Supercritical fluids are important mass transfer media in hydrothermal systems (e.g. ore deposit formation), or are present as metamorphic fluids in most of the middle to upper grade metamorphic facies. They are also responsible for metasomatic processes: for example, those related to subducting slabs at converging plate margins. Economic and industrial interests have also turned to critical and supercritical fluid systems, since high P-T fluids have successfully been applied as high efficiency solvents, often used for contaminant removal. The objective of our research is to examine the critical behavior and phase equilibrium properties of complex, four-component (H$_{2}$O-NaCl-KCl-CaCl$_{2}$) aqueous solutions using synthetic fluid inclusions. Critical PVTX properties in some of the binary subsystems in the H$_{2}$O-NaCl-KCl-CaCl$_{2}$ system have already been studied by previous authors. Experimental data have been published on critical phase relations in the system H$_{2}$O-NaCl, in dilute H$_{2}$O-KCl solutions, and in H$_{2}$O-CaCl$_{2}$ solutions up to 3 molal. Until recently, no experimental work has been done on the critical properties of more than two component systems, even though these solutions approximate the complexity of natural fluids. In order to examine these properties we used the synthetic fluid inclusion technique introduced by Bodnar and Sterner (1987). Critical points of brines up to 3 molal salinity have been determined. Our results show that 2 molal H$_{2}$O-KCl solutions have a higher critical temperature than H$_{2}$O-NaCl and H$_{2}$O-CaCl$_{2}$ brines of the same molality, whereas the latter one has significantly higher critical pressures. Critical points of three component solutions fall on slightly curved lines in P-T space and fall between the critical points of the corresponding binary end members. The critical point of the quaternary system falls in a triangle defined by the critical points of the three binary end members.
V34A-05 17:00h
Hydrogen Bonding, Hydration of Species, Ion Pairing and Clusterization in H$_2$O-NaCl-CaCl$_2$-CO$_2$-NaHCO$_3$-Na$_2$CO$_3$ Fluids: Molecular Dynamics Simulation of the Effects of Temperature, Pressure and Composition
Molecular computer simulation is an especially valuable tool to study the structural and dynamic properties of carbon-bearing aqueous fluids on the fundamental atomic time- and length- scale because these fluids are not readily studied experimentally using conventional X-ray or EXAFS methods. In this case, experimental methods can produce ambiguous results, because the carbon and oxygen atoms of the solute species are not easily distinguishable from the oxygen atoms of solvent water. Systematic molecular dynamics (MD) computer simulation studies of several fluid compositions $-$ H$_2$O-CO$_2$, H$_2$O-CO$_2$-NaCl, H$_2$O-NaHCO$_3$, and H$_2$O-Na$_2$CO$_3$ $-$ were performed to study the effects of temperature, pressure (fluid density) and concentration on the structural, energetic, spectroscopic and dynamic properties of these solutions characterized on the atomic scale via the statistical parameters of individual hydrogen bonds and H-bonding networks, local hydration structures of dissolved species, and ion pair formation. Similar molecular-level characteristics of pure water and H$_2$O-NaCl and H$_2$O-CaCl$_2$ solutions are calculated and used for comparison with the properties of carbon-bearing fluids. H$_2$O-rich and CO$_2$-rich compositions of the ternary H$_2$O-CO$_2$-NaCl system demonstrate strikingly different structural and dynamic behavior at about the same average density. In dense CO$_2$-rich fluids, dissolved H$_2$O molecules exhibit a high degree of hydrogen bonding and form relatively stable H-bonded clusters structurally similar to those observed in supercritical water at a much lower density. In contrast, CO$_2$ molecules dissolved in water-rich fluids occur in clathrate-like cages formed by surrounding H-bonded water molecules. The hydration shells of carbonate and bicarbonate ions both contain approximately 10 water molecules, but the water structure around the carbonate ion is much more pronounced due to the higher anion charge. This also leads to the formation of very stable ion pairs and larger ionic clusters of Na$^+$ and CO$_{3}^{2-}$ even under ambient conditions. Due to the ion cluster formation, the diffusion rates of both Na$^+$ and CO$_{3}^{2-}$ are $\sim$ 3-6 times lower than in similar NaCl solutions. The effect of dissolved Na$_2$CO$_3$ on the water structure is comparable to that of CaCl$_2$. In contrast, the structure and dynamics of NaHCO$_3$ solutions is very similar to that of H$_2$O-NaCl, where ion pairing is non-existent at lower temperatures, and the diffusion rates of HCO$_{3}^{-}$ are about the same as those of Cl$^-$. The molecular simulations also predict that a high-density fluid ($\rho$ $\sim$ 1.1 g/cm$^3$) of the composition H$_2$O/CO$_2$/NaCl=60/28/12 mol%, homogeneous at 1000$\deg$C, should experience a phase separation upon cooling at approximately 700$\deg$C by forming a low-salinity CO$_2$-rich phase (H$_2$O/CO$_2$/NaCl $\sim$ 50/48/2 mol%) and a concentrated brine (H$_2$O/CO$_2$/NaCl $\sim$ 66/20/14 mol%) in good agreement with available experimental data and thermodynamic calculations. The structural and dynamic properties of both fluid phases are rationalized on the molecular level in terms of the electrostatic and H-bonding interactions between the fluid species in order to understand physical mechanisms driving this phase separation process.
V34A-06 17:15h
The Role of Brines in low Temperature, Fault-related Deformation of Quartz Arenites
Fluids play an integral role in deformation within the Earth\'{}s crust over a wide range of physical conditions. At low temperatures ($<$$300\deg$C) the effect is dominantly mechanical, largely through the effects of pore fluid pressure. At higher temperatures ($>$$300\deg$C), chemical processes, such as diffusive mass transfer, advective mass transfer, and hydrolytic weakening dominate. Brines, because of their greater reactivity, enhance certain chemical processes during deformation. In the transition between high and low temperature regimes, both mechanical and chemical processes operate and interact in complex ways. This study investigates the role of brines in the deformation of quartz arenite in a map scale fault zone deformed under conditions transitional between low and high temperature regimes. The fault zone is also known to have been a conduit for fluids thought to be largely basinal brines. The Cove Mountain fault zone in south central Pennsylvania contains several map-scale blocks of quartz arenite which display a wide range of brittle and ductile microstructures. Abundant evidence of fluids is present in the form of quartz veins, microveins, fluid inclusion planes, cataclastic bands, and stylolites. Three different fluids are recognized based on cathodoluminescence color of quartz, and homogenization (T$_{h}$) and melting temperatures (T$_{m}$) from fluid inclusions. Blue-green luminescing quartz has a T$_{h}$ of 185 to $215\deg$C and a T$_{m}$ of -15 to -$17\deg$C; red luminescing quartz has a T$_{h}$ of 165 to $200\deg$C and a T$_{m}$ of -16 to -$20\deg$C; and zoned quartz with both red and blue-green luminescence has a T$_{h}$ of 180 to $220\deg$C and a T$_{m}$ of -18 to -$23\deg$C. The eutectic temperature of all three fluids is approximately -$50\deg$C suggesting that CaCl$_{2}$ is the dominant salt species. Grains adjacent to fluid conduits (microfractures, cataclastic bands, and stylolites) display more crystal plastic microstructures than those farther away. Compared to quartz in veins and in undeformed portions of the rock, which have water contents of ~1,000 and ~12,000 H/10$^{6}$ Si, respectively (determined by FTIR), grains adjacent to fluid conduits have water contents as high as 26,000 H/10$^{6}$ Si indicating that water was able to penetrate these grains and thus promote ductile deformation in a dominantly brittle regime. Access of water to grain centers was provided by both microfracturing and diffusion. Evidence for diffusion is shown by the presence of quartz grains adjacent to fluid conduits which have blue luminescing centers and red rims. Water contents in these grains ranges from 2,000-22,000 H/10$^{6}$ Si with some correlation between water content and luminescence color. The large amount of pressure solution relative to similar rocks in the region is believed to be related to the salinity of the brines. Also, the thick accumulation of goethite in stylolites and fractures precipitated by these fluids is unusual in similar rocks in the region suggesting that the fluid chemistry was atypical. The microstructures in this fault zone indicate that water is important in controlling the operative deformation mechanisms in the transition from low to high temperature deformation. Furthermore, the chemistry of these brines may account for greater solubility of quartz and faster diffusion rates than would be the case with low salinity fluids.
V34A-07 INVITED 17:30h
Geochemical Modeling of Evaporation Processes on Mars: Insight From the Sedimentary Record at Meridiani Planum
The Opportunity rover's analysis of an impure evaporite component present in the Martian sedimentary record reveals a unique geochemical system. The evaporation of basaltic weathering fluids is a process which is rare on Earth, but is likely to have played a major role in the formation of sedimentary rocks at Meridiani Planum. Adequately modeling the evaporation processes in this system must involve adding additional components to current thermodynamic models, namely Fe(II) and Fe(III). The goals of this study are to: (1) develop a thermodynamic database suitable for modeling evaporation of basaltic weathering fluids in the Meridiani system and (2) to apply the model to experimental fluid data obtained in our laboratory from weathering synthetic Martian basalt, which will allow for the testing of hypotheses related to the geochemical evolution of the Meridiani site. The evaporation of these fluids is simulated using an expanded version of the Harvie-Moller-Weare model which employs Pitzer's ion interaction approach in calculating activity coefficients in high ionic strength solutions. This model has been expanded using recent data to include Fe(II) and Fe(III). Although a full set of experimentally-derived data allowing the inclusion of Fe(III) into such models is not yet available, an adequate set of interaction parameters was built, based on viable assumptions and substitutions using analog data (e.g., Al$^{3+}$, Ga$^{3+}$, Cr$^{3+}$). The accuracy of the thermodynamic model in predicting Fe(II) and Fe(III) activities in a multi-component system can be assessed. This is accomplished by comparing calculated Eh values (proportional to a$_{Fe2+}$/a$_{Fe3+}$) to those measured in the field from high ionic strength acid mine waters containing all of the relevant components of the model. The agreement between calculated and observed values suggests that the model calculations are adequate for reaction path calculations. New thermodynamic data for several Fe(II) and/or Fe(III) containing minerals, including a variety of sulfates have also been incorporated. The resulting model is not only relevant to Mars, but acid mine drainage environments as well. The results of the calculations place constraints on the chemical controls of the evaporation system. For example, using fluids derived from a synthetic olivine-bearing Martian basalt, we predict gypsum (or anhydrite), jarosite, melanterite and hydrated Mg-sulfate as major phases produced upon evaporation. Jarosite has been identified by Moessbauer spectroscopy at Meridiani and Mg-sulfate is a likely outcrop component based on geochemical systematics. The redox conditions are unconstrained in this system and the formation pathways of Fe-containing minerals such as jarosite and hematite remain an open question. However, the inclusion of Fe(II) and Fe(III) in the model allows redox conditions to be systematically varied for any calculation. The stability of evaporite assemblages in contact with later fluids can also be modeled, testing hypotheses related to diagenesis. This may shed light on the origin of possible diagenetic features within the outcrop such as hematitic concretions and vugs that have been interpreted to be crystal moulds. Possible diagenetic reactions may have occurred as a result of groundwater recharge into previously deposited sedimentary layers.