V23C-01 INVITED 13:50h
Timescales of Geological Processes: a Diffuse Spectrum
One of the most significant recent advances in the study of geological processes is the ability to identify and quantify a hierarchy of time scales ranging from days to millions of years. Modeling diffusion related processes provides one of the most versatile tools for studying the lower end of this temporal spectrum. We will demonstrate this versatility with two examples chosen from very different milieus. (1) Volcanic processes in the Chilean Andes: We use somewhat novel modeling techniques to show that magma mixing can cause igneous differentiation in a subduction zone setting over decadal time scales. This provides a link between processes occurring over thousands to millions of years (melt generation and transport) and pre-eruptive processes that occur over days to months. (2) Plutonism and mid-crustal metamorphic processes in the Alps. Here, dike emplacement was triggered over time scales of days, cooling in magma chambers occurred over hundreds of years while metamorphism in the mid crustal level lasted millions of years. The regional distribution of such rates allows us to infer differential uplift across a section of the Alps. Traditionally, diffusion modeling has been plagued by large uncertainties arising from a number of sources. Large extrapolations of diffusion coefficients measured at high temperatures, simplified numerical models and inadequate understanding of diffusion mechanisms are some of these. Technical and theoretical developments now allow us to circumvent many of these problems. For example, the ability to manipulate complex silicate compositions on the nanometer scale through the use of thin film technology (e.g. pulsed laser ablation) allows us to measure diffusion coefficients at conditions that were inaccessible only a few years ago. In addition, these experiments are faster and more accurate. Consequently, it is now possible to determine parameters such as enhancement of diffusion rates in defective crystals or due to reactive processes. Contributions of such diffusion studies to the understanding of deformation rates will also be discussed. Coupling of processes occurring on different time scales is crucial for the stability of geochemical and geological cycles. Knowing the rates of individual processes, analyzing the mechanism of coupling across temporal hierarchies should be a central goal of future research.
V23C-02 INVITED 14:10h
Processes and Rates of Mass Transfer in Ultramafic-Hosted Hydrothermal Systems: An Experimental Study with Implications for Dissolved Inorganic and Organic Components in High-Temperature Vent Fluids
Recently discovered ultramafic-hosted hydrothermal systems at mid-ocean ridges reveal high temperature vent fluids with relatively high SiO2, Ca, H2, and methane and other hydrocarbons. Dissolved Fe concentrations are the highest of any vent systems yet discovered and require a relatively low pH and reducing conditions in subseafloor reaction zones from which the fluids are derived. This, together with the SiO2 concentrations of the vent fluids, strongly indicates fluid buffering by silica-rich phases possibly produced during pyroxene dissolution, the likely abundant presence of olivine notwithstanding. Theoretical predictions of olivine dissolution kinetics at elevated temperatures and pressures, however, suggest relatively rapid conversion of olivine to talc and serpentine with corresponding lowering of dissolved SiO2 and Fe, and increase in pH. Thus, to test this, we performed a series of experiments at 400°C, 500 bars involving olivine (Fo89) alteration in SiO2-bearing systems containing dissolved Na and chloride concentrations approximately equivalent to the Rainbow hydrothermal system. Time series fluid samples indicate steady state conditions. Results confirm unusually slow olivine reaction kinetics, even when coexisting with moderate to high dissolved SiO2. XPS and SEM analysis indicate Fe-enrichment on olivine surfaces, and formation approximately 14 percent talc. The olivine to talc conversion rate suggests a log rate of olivine hydrolysis of -11.97 (moles/cm2/sec), well below that predicted from available rate data extrapolated from lower temperatures and pressures. The relative enrichment of Fe on olivine surfaces may decrease the thermodynamic and kinetic drive for olivine dissolution; effectively precluding pH increases predicted assuming full equilibrium. Rates and processes of mass transfer involving Fe-bearing minerals may also help to catalyze Fisher-Tropsch synthesis of complex hydrocarbons reported for vent fluids issuing from ultramafic-hosted hydrothermal systems, as suggested by results of a companion series of hydrothermal experiments.
V23C-03 14:30h
Fluid Flow in Subduction Zones and Mountain Belts: The Importance of Permeability Heterogeneity and Anisotropy
Fluids are generally expected to be driven upward in the deep parts of orogens, but permeability heterogeneity and anisotropy must also be considered to properly interpret fluid infiltration and kinetic reaction histories preserved in the rock record. This paper focuses on new 2-D models of Darcian fluid flow incorporating permeability contrasts between rock units, the permeability tensor, and reactive fluid sources (e.g., dehydration). Factor of ten contrasts between the minimum and maximum permeability values in anisotropic rocks can strongly divert flow, but contrasts of as little as a factor of two still influence flow behavior. The first example considers fluid flow in subduction zone m\'{e}lange, Syros, Greece. Geochemical evidence suggests that the interiors of meta-mafic blocks of oceanic crust in the m\'{e}lange underwent limited fluid-rock reaction, despite extensive dehydration and decarbonation of the subduction complex. Modeling shows that if the blocks have lower permeability than the surrounding serpentine-rich matrix, then flow is diverted around the blocks resulting in little infiltration except at block margins, consistent with field relations. In this way, the subducted oceanic crust could preserve little evidence of fluid infiltration, even though considerable flow occurred through the m\'{e}lange. The largest fluid fluxes are concentrated in matrix where blocks are in close proximity, and this effect increases as the anisotropy of the matrix increases. The lack of fluid infiltration into blocks could account for the observed limited metamorphism and strong kinetic overstepping of reactions that in some cases allowed preservation of ocean-floor mineral assemblages even at blueschist-eclogite facies conditions. The second example examines fluid flow through a folded sequence in which the direction of maximum permeability is parallel to the folded layering, and is based on field relations of Barrovian metamorphic sequences in CT, USA, and Scotland. As the degree of anisotropy increases, flow is directed along fold limbs and away from fold hinges, becoming more and more concentrated at the tips of antiforms and synforms where cross-layer flow components dominate. Surprisingly, the largest model fluxes are limited to relatively small regions at the tips of fold hinges. The model results predict that permeability heterogeneity and anisotropy have the potential to strongly divert metamorphic fluid flow directions and, thus, influence the spatial distribution of fluid fluxes and the sites of fluid-rock reaction. Consequently, models of deep orogenic flow that consider only homogeneous rock should be re-examined, and new measurements of permeability contrasts and anisotropy relevant for deep crust/upper mantle conditions are needed to better constrain models and, ultimately, field observations.
V23C-04 14:45h
Effect of Mg on the Grain Growth and Dislocation Creep of Calcite
We tested the effect of variations in the amount of the solute impurity (Mg) on grain growth and strength of calcite aggregate. Synthetic marbles were produced by hot isostatic pressing mixtures of powders of calcite and dolomite at $850\deg$C and 300 MPa confining pressure for different intervals (2 to 30 hrs). The HIP treatment resulted in homogeneous aggregates of calcite with Mg content from 0.5 to 17 mol$%$. Stress stepping tests and constant strain rate tests were used to examine the effect of Mg content on the dislocation creep of calcite. The grain growth rate under static conditions was decreased with Mg content from 7 to 17 mol$%$, indicating perhaps that grain boundary mobility is suppressed by the solute drag effect. In the diffusion creep at stresses below 40 Mpa, the strength of calcite decreases with increasing Mg content owing to the difference in grain size at $800\deg$C and 300 MPa confining pressure. The contribution of dislocation creep increases with increasing stress, and the transition between diffusion and dislocation creep occurs at higher stresses for the samples with higher magnesium content and smaller grain size. The creep data were fit assuming a composite flow law consisting of a linear combination of diffusion and dislocation creep and a single-valued grain size. The best agreement was obtained by using a dislocation creep law with exponential dependence of strain rate on stress (e.g. Peierls law). More evidence from microstructure is needed to identify the dominant deformation mechanism conclusively. Most of the samples were compressed up to strains of 0.25; small recrystallized grains are formed resulting in a bimodal grain size distribution in some of the deformed samples. Preliminary data shows that the recrystallized grain sizes are smaller for Mg-calcite compared with that of pure calcite. This study will help to understand the effect of impurities on grain-growth kinetics and strain weakening in localized shear zones.
http://www.agu.org/lilixu2004
V23C-05 INVITED 15:00h
Reaction Rates in Deformation and Hydrostatic Experiments in the Anhydrous System Anorthite - Forsterite
The reaction anorthite $+$ forsterite --$>$ cpx $+$ opx $+$ spinel $\pm$ gnt proceeds at high temperatures and elevated pressures in the lower crust and upper mantle. This solid-solid reaction was studied experimentally at 900$\deg$C in the pressure range of 1000 to 1600 MPa in both shearing deformation and hydrostatic experiments. Powder mixtures (1:1 by vol) of anorthite (An92) and forsterite (Fo93) are hot pressed at 970$\deg$C, 750 MPa for 48 hrs in a Griggs apparatus and deformed ($\dot\gamma$ = 5 $\times$ 10$^{5}$ sec$^{-1}$) after adjustment of P and T to run conditions. H$_{2}$O content of the samples has been measured by FTIR and is $<$ 30 ppm. At small pressure overstepping (ca. 200 to 300 MPa) undeformed samples show only 10 % reaction progress after 168 hrs, whereas reaction progress in deformed samples after 72 hrs is 60 %. At greater pressure overstepping (700 to 800 MPa) the difference between deformed and undeformed samples is less pronounced (95 % after 60 hrs deformed, 75 % after 168 hrs undeformed) but still present. At greater pressure overstepping, undeformed samples show an exponential reaction rate, whereas that of deformed samples is always linear. Samples initially deformed and then kept hydrostatically show a fast initial reaction rate (85 % of total reaction progress after 0.25 of total run time), followed by a slower reaction progress (15 % reaction after 0.75 of total time) under hydrostatic conditions. The difference in reaction progress is mainly attributed to different nucleation rates. In all experiments, enstatite rims form around olivine grains separating those from other reaction products. Such coronas are indicative of diffusion-controlled reactions. Plots of rim thickness vs time indicate a relative increase of the bulk diffusion coefficient by a factor 5 in the deformed samples compared to undeformed. However, as the grain size of reaction products of deformed samples is 10 times smaller than in undeformed ones, the nucleation rate in deformed samples is $\sim$ 5000 times higher. The increased nucleation rate is interpreted to result from higher defect densities and greater driving potential for the reaction in deformed samples. The increased nucleation rate enhances reaction rate and thus rim growth rate, so that a straightforward interpretation of rim growth in terms of diffusion parameters is impossible. Thus, nucleation and length scales of reaction processes in deforming rocks may be very different from those in hydrostatic cases, and for the purpose of extrapolation to nature rates of reaction from hydrostatic experiments provide only minimum values.
V23C-06 15:20h
Synchrotron Radiation Study of the Kinetics of Dehydration of Chrysotile Fiber
Chrysotile fiber from the Thetford region of Quebec in Canada was heated in hydrothermal diamond anvil cells (HDACs) under both hydrothermal and atmospheric pressure conditions. Temperatures between 25\deg\ C and $>$800\deg\ C were attained at varying rates using electrical resistance heating elements surrounding the sample space. The HDACs were fixed in the path of a synchrotron-radiation beam at the Advanced Photon Source (APS) at the Argonne National Laboratory. X-ray diffraction patterns were recorded with high temporal resolution in order to develop time-temperature transformation curves for the decomposition of chrysotile and subsequent formation of crystalline phases such as forsterite, tridymite, talc, and amphibole. During atmospheric-pressure experimentation chrysotile was confined in sample space drilled in rhenium-foil gasket material ranging in thickness between 150-300 microns. Diamond anvils were brought into light contact with the gasket material and no further pressure was applied. For hydrothermal experimentation, rhenium-foil gaskets were loaded with chrysotile and pure water, and pressure measurement utilized pressure-temperature-density diagrams calculated from an equation of state for water. During 72 hours of experimentation at APS approximately 150 diffraction patterns were generated under a variety of temperature/time/pressure conditions. Following are some of the significant results obtained: On rapid heating at atmospheric pressure from 120\deg\ C to 800\deg\ C ($<$3 minutes) chrysotile is more than 90% decomposed, and the growth of forsterite, tridymite, and talc are noted. At 32 minutes with the temperature held at 800\deg C chrysotile is absent and forsterite, tridymite and talc phases are growing. At 103 minutes talc is absent and an 8.3\AA\ peak consistent with amphibole growth appears, while tridymite abundance is decreasing. At 164 minutes forsterite and amphibole are increasing in abundance and tridymite continues to decrease. At 280 minutes tridymite is absent while forsterite and amphibole remain. Notable is the transient state of coexistence of forsterite and tridymite for more than 164 minutes but less than 280 minutes. Investigation of lower temperature reactions exhibited a destabilization of the chrysotile structure at 450\deg\ C after 240 minutes, seen as loss of peak area for major chrysotile d-spacings, accompanied by formation of a number of peaks in the 1.3-3\AA\ range of d-spacing. Chrysotile that was held off-line at 450\deg\ C and at atmospheric pressure for 48 hours revealed a well-developed diffraction pattern consistent with growth of forsterite. Ongoing analysis of the data collected at APS will help define a kinetic framework for the dehydration and decomposition of chrysotile, and help to refine the experimental design for future studies. Additionally, integration of these data with data from our limited hydrothermal experimentation will expand our understanding of the role hydrothermal fluids play in rates of decomposition.