MR21B-1776 [WITHDRAWN]
The role of directed van der Waals bonded interactions in the determination of the structures of molecular arsenate solids
On the basis of his famous electrostatic theorem, Feynman (1939) showed that it is not the fluctuating dipole- dipole interactions among neighboring molecules that lead to van der Waals forces, but rather it is the net attraction of the nuclei for the distorted electron density, ED, accumulated between the nuclei of bonded atoms that accounts for the van der Waals R-7 forces. Bader and his coworkers (1990) have since concluded that the distortions in the internuclear region arise by dint of the formation of local maxima and minima of the Laplacian distribution, L(r) = ▽2ρ(r), with the formation of van der Waals bond paths associated with the maxima and minima. The maxima are ascribed to Lewis base domains where the ED is locally charge concentrated, CC, whereas the minima are ascribed to Lewis acid domains where the ED is locally charge depleted, CD. For the As2O3 molecular crystals arsenolite, claudetite I and claudetite II, AsO2 and the As- metalloid, arsenolamprite, the ED of the constituent molecules were found to adopt a configuration where the Lewis acid and base domains of molecules are aligned and connected by As-O, O-O and As-As van der Waals intermolecular bond paths. Despite the relative weakness of the van der Waals bonded interactions relative to the intramolecular As-O bonded interactions, the interactions are concluded to serve as mainstays for the individual molecules in each of the molecular solids. Intermolecular As-O bond paths between the bonded atoms connect Lewis base CC and Lewis acid CD domains whereas the O-O and As-As paths connect Lewis base-pair CC-CC domains and Lewis acid pair CD-CD domains, respectively, give rise to sets of directed van der Waals bond paths. The alignment of the bond paths, like any other bond path, results in the periodic structures adopted by the molecules in the arsenates. The cubic structure adopted by arsenolite polymorph can be understood in terms of sets of As-O and O-O directed bond paths that radiate from the tetrahedral faces of its constituent molecules, serving as face-to-face Fischer key-lock mainstays in forming a periodic tetrahedral array of molecules. The arrangements of the As atoms in the claudetite polymorphs of As2O3 and those in the As metalloid arsenolamprite are similar. The arrangement of the molecules in the thiosulfides realgar, AsS, alancranite, AsS, α-dimorphite, As4S3 and uzonite, As4S5 can also be understood in terms of the Lewis domains and the van der Waals bond paths that radiate from the faces of the constituient molecules. Like the two clausetite polymorphs, arsenolamprite is a molecular solid bound by relatively weak intermolecular van der Waals bonded interactions. The bond critical point and local energy density properties of the intermolecular As-As interactions in arsenolamprite are comparable with those in claudetite I. As such, the structure of claudetite I can be viewed in large part as a stuffed derivative of the arsenolamprite structure with O atoms between pairs of As atoms comprising the layers of the structure.
MR21B-1777
Coupled Finite Volume and Discrete-Finite element Methods for Modeling Hydraulic Fracturing in Geologic Formations
ABSTRACT: High demand for stimulation treatments of fluid-state hydrocarbon reservoirs is driving increased interest in improved understanding of the fundamentals of hydraulic fracturing of geologic formations. In addition, prediction of caprock integrity under the load of geologically sequestered, pressurized CO2 requires better understanding fluid-rock interactions. The approach described here addresses modeling of hydraulic fracturing at the meso-scale, using a discrete-finite element method code (LDEC) coupled to a modified finite volume method to capture compressible flow in a propagating fracture. Leak-off is also addressed through a model parameterized by flow rate and cumulative flow through the fracture face; this approach is used to better approximate the functional form of the dominant underlying chemo-physical phenomena which lead to permeability loss at the fracture face over typical models, which are often parameterized only by time and calibrated, through a set of parameters, to match experimental data. A simulation of a standard fracture injection test is used to compare the results of the proposed leak-off model with the popular Carter leak-off model and shows excellent agreement between the two models. Also, the finite volume approach is verified against analytical solutions for constant aperture parallel plate flow, and results of a validation study comparing simulation results with an experiment on the propagation of a fracture in a brittle, homogeneous polymer are discussed. ACKNOWLEDGEMENTS: This work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.
MR21B-1778
Acoustical properties of dry and saturated porous media
Our objective is to determine the macroscopic acoustical properties of porous media (either dry or saturated by an interstitial fluid) and to relate them to the mechanical and hydromechanical characteristics of the medium and its components. Wave propagation in a dry elastic material is governed by the elastodynamic equation. For a dry medium, the stress is zero on the pore surface. The medium is supposed to be spatially periodic and composed of identical cells. When the wave length λ is very large when compared to the scale l of the heterogeneities, the medium behaves in a first approximation as an equivalent homogeneous material. All the fields can expanded as series of the small parameter η= l/2πλ, in terms of two space variables x and y associated to the scales λ et l, respectively. This expansion is introduced into the elastodynamic equation with appropriate boundary conditions. A series of non homogeneous partial differential equations are found for the successive orders in η. The predominant order corresponds to the equivalent homogeneous material. The first order equation provides the polarization correction, the second one the celerity dispersion and the third one the attenuation. These equations are discretized by a finite volume formulation in a tetrahedral mesh which is either structured or not. The resulting linear system is solved by a conjugate gradient method. Each elementary volume may have specific properties. Wave propagation in a saturated medium is more complex since it is influenced by the solid and liquid phases. When a periodic oscillation is imposed, the solid displacements are governed by the elastodynamic and the Stokes equations coupled by boundary conditions at the interface. The solutions to these equations yield the macroscopic characteristics of the medium. The first equation yields two independent problems in the solid, one identical to dry media and one corresponding to a medium submitted to an interstitial macroscopic pressure. The second equation yields the dynamic permeability which is a complex function of frequency. Various media have been addressed. Model media may be random packings of spheres which can be partially consolidated by calcite, unimodal or bimodal reconstructed media. Real three dimensional samples made of several minerals were also obtained by computed microtomography. All the results relative to these media will be presented and discussed.
MR21B-1779
Ab-initio structural and electronic properties of Fe, Si, GaAs and NiO from a novel DFT+U+V approach
The electronic and structural properties of materials as diverse as Fe, Si, GaAs and NiO are investigated using a novel DFT+U energy functional named DFT+U+V. In this approach the corrective ("+U") Hamiltonian is constructed on an extended Hubbard model that includes both on-site (U) and inter-site (V) electron-electron interactions. The competition between these interaction parameters avoids the over-stabilization of filled atomic orbitals that is usually observed with the "standard" DFT+U approach and allows to describe systems (like covalent semiconductors) whose valence electrons are "localized" on hybridized orbitals (bonds) rather than on pure atomic states. Thus the novel DFT+U+V functional not only improves significantly the energetics (crucial for performing structural optimizations and studying phase transitions) but also dramatically extends the applicability of the "standard" DFT+U to a much broader class of systems. Also, the additional inter-site interaction parameter V can be straightforwadly obtained (at no additional cost) from the same linear-response approach used to calculate U [1]. The effectiveness and reliability of the novel functional are demonstrated by the good results obtained with semiconductor (Si and GaAs)and charge-transfer insulator (NiO) materials. Some still remaining difficulties with metals are also discussed. [1] M. Cococcioni and S. de Gironcoli, Phys. Rev. B 71, 035105 (2005).
MR21B-1780
Visualization and Analysis of Structural and Dynamical Properties of Hydrous Silicate Melt
We have carried out a detailed visualization-based analysis of position-time data produced by first principles molecular dynamics simulations of hydrous MgSiO3 liquid to gain insight into its structural and dynamical behavior. A wide range of pressure (0 to ~100 GPa) and temperature (2000 to 6000 K) is covered and the water content is also varied (5 and 10 wt percent water in the melt). By exploring a number of structural parameters associated with short- and mid-range orders, we have shown that the melt structure changes substantially on compression. The speciation of the water component at low pressures is dominated by the isolated structures (with over 90 percent hydrogen participated) consisting of hydroxyls, water molecules, O- H-O bridging, and four-atom (O-H-O-H and H-O-H-O) groups, where every oxygen atom may be a part of polyhedron or free (i.e., bound to only magnesium atom). Hydroxyls slightly favor polyhedral sites over magnesium sites whereas molecular water is almost entirely bound to magnesium sites, and also interpolyhedral bridging (Si-O-H-O-Si) dominates other bridging. As compression increases, these isolated structures increasingly combine with each other to form extended structures involving a total of five or more O and H atoms and/or containing three-fold coordination species, which together consume over 60 percent hydrogen at the highest compression studied. Protons in the melt can be considered on equal footing with other cations (or more precisely as network modifier cations), and they are shown to increase and decrease, respectively, the contents of non-bridging and bridging oxygen. Relatively long runs are used to calculate the self-diffusion coefficients of all atomic species, which are enhanced in the presence of water compared to those of anhydrous melt. This is consistent with the prediction that water depolymerizes the melt structure at all pressures. Our analysis suggests that proton diffusion involves two processes the transfer of H atoms (requiring the rupture and formation of O-H bonds) and the motion of hydroxyls as hydrogen carriers (requiring the rupture and formation of Si-O and/or Mg-O bonds). Both the processes are operative at low compression whereas only the first process is operative at high compression.
MR21B-1781
Electronic Structure of Solid FeO at High Pressures by Quantum Monte Carlo Methods
We have carried out calculations of electronic structure and energy-volume of solid FeO for antiferromagnetic B1 and iB8 structure at a range of volumes. The calculations are done in a genuine many-body framework using the fixed-node diffusion Monte Carlo (FNDMC) method. Several important methodological developments were employed: twist-averaged sampling of the Brilloiun zone, Ne-core Dirac-Fock pseudopotentials with 3s and 3p states in the valence space and optimized one-particle orbitals from hybrid DFT functionals for building high-accuracy Slater-Jastrow explicitly correlated wavefunctions. This enabled us to minimize the systematic errors such as the finite size supercell and fixed-node biases. The calcualted equilibrium parameters such as cohesion, equilibrium lattice constant, bulk modulus and electronic gap show an excellent agreement with experiment. This is a remarkable progress for strongly correlated systems such as transition metal oxides, considering that the FNDMC method does not contain any free nonvariational parameter(s). We also estimate the transition pressure B1 to iB8 and compare the pressure-volume data with available experiments. The estimated transition pressure of 65(5) GPa lies at the lower end of the pressure range reported by several experiments over the past two decades. The implications of these results and comparisons with other approaches will be discussed as well.
MR21B-1782
An Effective Semi-empirical Ansatz for Computing Anharmonic Free Energies
An ansatz to include anharmonic effects neglected in quasiharmonic free energy calculations is developed. A parameterized temperature dependent modification of the vibrational density of states (VDOS) designed to be used directly on the quasiharmonic free energy expression is introduced. The relationship between this modified VDOS and the renormalized VDOS is established. This parameterization is shown to produce the correct low and high temperature behavior for free energy and other thermodynamics properties. The thermodynamics properties of MgO (periclase) and alpha- and beta-Mg2SiO4 (forsterite and wadsleyite) improve considerably after inclusion of anharmonic effects. Anharmonicity is shown to have a dramatic effect on the phase transformation between forsterite and wadsleyite. It can reconcile the discrepancies between the experimental Clapeyron slopes and the slopes predicted by quasiharmonic-type calculations. Research supported by NSF/EAR 0635990, and NSF/ITR 0428774 (VLab). Computations were performed at the Minnesota Supercomputing Institute.
MR21B-1783
Resolution and Smoothing Effect of Tomogram and Their Impact on Computational Velocity Estimation
Pore microstructure and pore-scale simulations have been used to understand physical properties of rocks and their interrelations. Pore microstructures are typically obtained from the X-ray tomographic technique, and we can obtain up to one micron spatial resolution from high-resolution scanning facilities. Though the accuracy of pore-scale simulations depends on grain size distribution, transport properties (permeability and electrical conductivity) can be accurately estimated with current spatial resolution and are recently used widely in many applications. However, the elastic properties can be problematic, because they are sensitive to grain contact areas, which are very difficult to resolve accurately by tomography. In this paper, we are presenting results on the effect of resolution and smoothing of tomogram on pore-scale velocity calculation. We used three different rocks: 17%-porosity sandstone from offshore Korea; 23%-porosity Aztec sandstone; and 42%-porosity beach sand. Three dimensional microstructures were obtained from different high-resolution scan facilities with different resolutions. We found that the resolution of tomogram highly affect velocity estimation from FEM elastic simulation. Two-micron spatial resolution is sometimes not enough to resolve the contact areas between 200-micron grains. Especially for high porosity sandstone, the calculated velocity was overestimated by tens of percent. In addition, smoothing effect from image processing on tomogram acquisition can cause non-negligible velocity overestimation; however, this effect is smaller than that from resolution and can be corrected by anti-smoothing filtering techniques. In conclusion, care should be taken to estimate the velocity of unconsolidated or high-porosity sandstone from pore-scale simulation and smoothing effect is also investigated more carefully.
MR21B-1784
Effect of pressure on the structure and compressibility of hydrous ringwoodite: a first principle investigation
First principle calculations were performed to investigate the structural characteristics of hydrous ringwoodite and their evolution with pressure up to 36 GPa. We focus on hydrous ringwoodite with two water contents: Mg1.875H0.25SiO4 (1.65 wt % H2O) and Mg1.75H0.5SiO4 (3.3 wt% H2O). Hydrogen defects are introduced by replacing Mg atoms. Protons tends to bond to an oxygen atom to align the O-H dipoles along the edge of the Mg-vacant octahedron. The energy difference between the two energy minima along an O O pair is small enough for the two sites to be exchangeable at Earth conditions. The calculated bulk modulus decreases linearly with water content with dK/d(CH2O) = -9.9 (GPa/ wt%). Ab initio molecular dynamics simulations show that protons are contained inside the Mg-vacant octahedral cage sampling the six O hosts and do not have long-range diffusion at temperature up to 2000K.
MR21B-1785
High-density Cotunnite-TiO2 Phase as Determined by Theory and Experiment
The nature of bonding in titanium dioxide, TiO2, is of interest as a superhard material with many industrial applications. Additionally, TiO2 is an accessory mineral structurally similar to SiO2, a major mineral in the Earth. Using density functional theory based first-principle computations, and complementary high-resolution synchrotron x-ray powder diffraction, we have determined the equations of state (EoS) of the orthorhombic I (OI) and orthorhombic II (cotunnite OII) phases of TiO2. Our static first- principle calculations predict that the volume across the OI → OII phase transition decreases by ~8.3% and ~7.6% for LDA and GGA, respectively. This computationally predicted large volume change is fully supported by our experimentally observed volume decrease during OII synthesis (at 56 GPa and 1800 K) of ~8.3%. The large magnitude of the volume decrease observed in TiO2 is also consistent with the systematics of the volume change across the OI → OII transition in other oxides, such as ZrO2, HfO2, SnO2, and PbO2.
MR21B-1786
First-principle Simulation of Magnesium-aluminum Spinel (MgAl2O4)
11033768 First-principle Simulation of Magnesium-aluminum Spinel (MgAl2O4) Materials with the spinel crystal structure, AB2O4 are believed to be an important component of Earth's mantle and may be related to density and seismic wave velocity discontinuities at the transition zone from 400km to 660km depth. Using Ab-initio calculations, five phases are predicted to have a stability range at zero temperature: magnesium-aluminum spinel (MgAl2O4), two of its polymorphs, which are of Pbnm and Cmcm space groups, periclase (MgO) and corundum (Al2O3). Pbnm-MgAl2O4 has the calcium-ferrite structure and Cmcm-MgAl2O4 takes the calcium-titanate structure. Calculations are preformed using the PWSCF (Plane-Wave Self-Consistent Field) codes. The free energy of the compressed volume was calculated directly for each of the phases above. Based on the energy-volume results from the calculations, dissolution of MgAl2O4 into MgO + Al2O3 occurs at 12GPa and the mixture (MgO + Al2O3) is expected to recombine to form the calcium-ferrite type phase at about 27GPa. The two phase transition pressures are consistent with experimental results. Cell parameters of the five phases simulated and their bulk modulus derived from the energy-volume curve are also in good agreement with experimental work. But unlike the conclusions drawn from some previous experimental work, the calcium-ferrite type structure (Pbnm-MgAl2O4) did not transform to the calcium-titanate type structure (Cmcm-MgAl2O4) at around 40GPa, which provides the possibility that calcium-ferrite type phase may be stable to even higher pressures (up to100GPa). Derived parameters, bulk modulus and density of each phase are in good agreement with experimental results. The differences are within 4%. Compared to seismic velocity profiles of the earth, these phase transitions pressures match the discontinuity pressures at transition zone 400km (Fd3m- MgAl2O4 -> MgO + Al2O3) and 660km (MgO + Al2O3 -> Pbnm-MgAl2O4) respectively, suggesting that the spinel phase transitions may be partially responsible for the seismic discontinuity at the transition zone. Future calculations will add the energy contribution of phonons which will allow calculations at finite temperature in the first-principle simulations. This is expected to provide more accurate predictions for application to realistic mantle temperature-pressure conditions.
MR21B-1787
Investigation of high-pressure post-magnesite phases of MgCO3
We investigate several high-pressure phases of magnesite (MgCO3) by first-principles evolutionary simulations and Car-Parrinello molecular dynamics. Three predicted high-pressure post-magnesite structures are studied at pressures from 140 GPa to 850 GPa. The lowest-pressure structure in this pressure range has corner-sharing CO4-tetrahedra, and is stable at p > 140 GPa while the highest-pressure structure, which contains edge-sharing CO6-octahedra, is stable above ~ 750 GPa. In addition, the intermediate structure is found to be stable in the range ~ 400-750 GPa. By analogy to silicates, we also compare the predicted post-magnesite structures with hypothetical perovskite and post-perovskite MgCO3 structures. It is found that perovskite and post-perovskite are energetically unfavoured throughout the investigated pressure range, implying that the chemistry of silicates and carbonates are very different not only at low, but also up to extremely high pressures.
MR21B-1788
Web-based implementation of atomistic visualization
Atomistic (molecular) visualization is one of the most widely studied applications of scientific visualization. It deals with time-varying three dimensional positional data representing snapshots of atomic configurations produced by molecular dynamics simulations of a variety of materials including geomaterials. We have recently developed an efficient scheme, which integrates the analysis and rendering tasks together in order to support interactive visualization at space-time multi-resolution of these data. Our scheme allows us to gain better insight into bonding, radial distribution, atomic coordination, clustering, structural stability and distortion, and diffusion. We are currently extending the support for web-based access to atomistic visualization by developing a three-level distributed application with platform independence and portability. The first layer contains off-screen rendering engine whose functionality is exposed using Web Service. This layer supports batch-style rendering that allows remote analysis of data and provides general way to access service from different types of clients. The second layer is a web application that enables user to interact with data using Web Service as entry point to rendering engine. Finally, the front-end of the system is a web browser (e.g. Firefox, Safari, Internet Explorer). We will also take the advantage of relational database to store simulation results and retrieve them from rendering service. We will present the details of the implementation and applications.
MR21B-1789
Low-pressure clino- to high-pressure clino-enstatite phase transition: a phonon related mechanism
We have investigated by first principles the compressional behavior of low-pressure (LP) and high-pressure (HP) MgSiO3 clinoenstatite. We have carefully examined cell shapes, chain angles, and polyhedral volume responses, such as angle variances and quasi-elongations, under pressure at room temperature. We have observed opposite behavior of the tetrahedra in the S-rotated and O-rotated chains with pressure in the LP phase, with a slight increase (decrease) in angle variance and quasi-elongation in the former (latter). Inspection of zone center modes of both phases under pressure reveals a transition path that converts the S- rotated chain in the LP phase into the O-rotated chain in the HP phase. This conversion is related to a slight softening of an Ag "metastable" Raman mode under pressure. The thermodynamics of the transformation was also investigated. As in other polymorphic transitions in silicates, phase boundaries determined by the GGA and LDA functionals bracket experimentally measured boundaries with the GGA (LDA) overestimating (underestimating) the same trend that the transition pressure. The calculated Clapeyron slopes are similar and in close agreement with the experimentally determined values. Research supported by NSF/EAR 0635990 and NSF/ITR 0428774 (VLab). Computations were performed at the Minnesota Supercomputing Institute.
MR21B-1790
Structure of iron at Earth Core Conditions
Ab-initio molecular-dynamics and electronic-structure calculations using the density-functional approach indicate that the face-centered cubic (fcc) phase of Fe is mechanically stable at outer-core pressures and temperatures of ~150-160 GPa and 6000-6600 K (atomic volume 7.7Å3). Bain-path and stress-anisotropy calculations document the mechanical stability of fcc iron at these conditions. Based on free-energy calculations with temperature-dependent phonon and electron contributions, the hexagonal close-packed (hcp) phase is found to be more stable than fcc iron by 7-9 mRyd at inner-core conditions of ~330-360 GPa and 6000-6600 K. This is consistent with the fact that fcc phase as a competitive candidate to hcp is based on theoretical simulations explaining the temperature quenched metastable laser heated high pressure samples and truly do not demonstrate the stability of fcc phase. Whereas the simulations reveal that fcc iron is mechanically stable but energetically less stable than the hcp phase, they show that bcc Fe is both mechanically and energetically unstable at Earth-core conditions. The small energy differences between the three phases mean that their relative stability and the mechanical instability of bcc iron may be influenced by the incorporation of light elements.
MR21B-1791
First-principles study of low-spin LaCoO3 with structurally consistent Hubbard U
We use the local density approximation + Hubbard U (LDA+U) method to calculate the structural and electronic properties of low-spin LaCoO3. The Hubbard U is obtained by first principles and consistent with each fully-optimized atomic structure at different pressures. With structurally consistent U, the fully-optimized atomic structure agrees with experimental data better than the calculations with fixed or vanishing U. A discussion on how the Hubbard U affects the electronic and atomic structure of LaCoO3 is also given.
MR21B-1792
VLab: Collaborative Grid Services and Portals to Support Computational Mineral Physics
We will review the architecture and current stage of implementation of VLab, a Grid and Web-Service-based system for enabling distributed and collaborative computational chemistry and material science applications to planetary materials. The latter uses the Quantum ESPRESSO as back-end computational package and has been developed to enable distributed and automatic execution of the extensive workflows spanned by realistic and demanding computations of minerals' properties through a consolidated user-friendly interface. The requirements of VLab include job preparation and submission, job monitoring, data storage and analysis, and distributed collaboration. These components are divided into client entry (input file creation, visualization of data, task requests) and back-end services (storage, analysis, computation). In particular, we will describe three aspects of VLab: (1) managing user interfaces and input data with JavaBeans and Java Server Faces; (2) integrating Java Server Faces with the Java CoGKit; and (3) designing a middleware framework that supports collaboration. We will demonstrate its performance on realistic examples onsite.
MR21B-1793
Enhanced Finite Elements With Linear Displacement Jumps for Modeling Propagating Fault and Fracture Systems
Finite elements with embedded discontinuities have been developed to study localized deformation patterns such as shear bands, fractures, and faults. Upon detection of a failure condition, a failure surface can be added in these elements at an orientation consistent with the critical angle of the failure condition, and at a location consistent with the propagating fracture. Until recently, the embedded displacement jump in the element was limited in that it had to be constant in that element, disallowing a jump in the strain and stress fields on opposite sides of the discontinuity. In many fault systems, however, there is relative tension on one side of the fault in the direction parallel to the fault. We enhance the kinematics of the element to account for linear displacement jumps, thus allowing for stress and strain jump across the element, while maintaining traction continuity. This enhancement allows for more realistic capturing of the stresses along a fault, and hence more realistic propagation of the fault or fracture surfaces. The method is applied to problems of fracture and fault extension in multiple rock layers.
MR21B-1794
Modelling dislocation cores in Mg2SiO4 wadsleyite at 15 GPa: Comparison with TEM observations
Wadsleyite (Mg,Fe)2SiO4 is considered as a major constituent of the transition zone of the mantle (between ca. 410 km and 520 km depth). The rheology of this mineral is thus of primary importance for the dynamics of this important layer of the mantle and for matter exchanges between the upper and the lower mantle. A series of deformation experiments followed by detailed dislocation characterizations at the transmission electron microscope (TEM) have led to a first description of the possible dislocation microstructures in this mineral (see Thurel et al's publications below). The aim of this work is to study numerically the dislocation core structures of Wadsleyite at 15GPa. We use the Peierls-Nabarro (PN) model combined with finite elements methods. Generalized Stacking Faults (GSF) calculated at the atomic scale with the GULP code are used as an input for the PN model. We have considered five slip planes in Wadsleyite: (100), (010), (001), (011) and (101). The PN model is used to determine the core structures of dislocations with [100], [010], [001] and [101] Burgers vectors in the appropriate glide planes. Our numerical results are in good agreement with previous TEM study of Thurel et al.. In particular, we show that [100] glide is the easiest. We also reproduce the dissociation of the [010] dislocation into four partial dislocations and the dissociation of the [101](010) dislocations. E. Thurel and P. Cordier (2003) Plastic deformation of wadsleyite: I High-pressure deformation in compression. Physics and Chemistry of Minerals, 30(5), 256-266. E. Thurel, P. Cordier, D. Frost and S.-I; Karato (2003) Plastic deformation of wadsleyite: II High-pressure deformation in shear. Physics and Chemistry of Minerals, 30(5), 267-270. E. Thurel, J. Douin and P. Cordier (2003) Plastic deformation of wadsleyite: III. Interpretation of dislocations and slip systems. Physics and Chemistry of Minerals, 30(5), 271-279.
MR21B-1795
NA Nonlinear Equation-of-state Inversion
A fully non-linear inversion scheme is introduced for the determination of the parameters controlling the equation-of-state and elasticity of mineral phases using the thermodynamically consistent finite-strain formulation introduced by Stixrude & Lithgow-Bertelloni (2005). This inversion exploits a directed search in an eight-dimensional parameter space using the Neighbourhood Algorithm (NA) of Sambridge (1999) to search for the minimum of an objective function representing the misfit to multiple data sets that constrain different aspects of the mineral behaviour. No derivatives are employed and the progress towards the minimum builds on the accumulated information on the character of the parameter space acquired as the inversion progresses. When only a limited range of experimental information is available there is a strong possibility of multiple minima in the objective function, which can pose problems for conventional iterative least-squares or other gradient methods. The addition of many different styles of data tends to produce a better defined minimum. The influence of different data types can be readily assessed by allowing differential weighting. The new procedure is illustrated by application to MgO, for which extensive experimental data are available. These include the variation of relative volume V with temperature T and pressure P from both static and shock-compression experiments, acoustic measurements of compressional and shear (and hence bulk) moduli, and calorimetric determinations of entropy as a function of temperature at atmospheric pressure. Preliminary NA modeling highlighted tensions between marginally incompatible subsets of data. We therefore excluded one-atmosphere V(T) data for T ≥ 1800 K for which the quasi-harmonic approximation is inadequate (Wu et al., 2008) along with elastic moduli derived from Brillouin spectroscopy under conditions (P ≥ 14 GPa) where significant departures from hydrostatic conditions are expected. With these limited exclusions based on sound physical principles, the NA search identified a compact family of models that provide an excellent fit to the diverse experimental data and a measure of the covariance between key model parameters.
MR21B-1796
Lattice Boltzmann Simulation of Calcite Precipitation and Isotope Fractionation
The Lattice-Boltzmann (LB) method can be used to simulate aspects of chemical reactions such as vapor condensation or mineral precipitation from a fluid phase. It has the advantage of allowing the solute phase to evolve depending on local conditions rather than being specified, so that the controls on precipitated phase grain shape can be studied simultaneously with the controls on chemical compositions and growth rate. We have used the LB approach to simulate the precipitation of calcite from oversaturated water as an effort in developing methods for treating complex chemical reaction problems, including isotopic effects. The formation of calcite is a classic problem in diffusion and reaction-limited mineral growth. There are many complexities, but the process is attractive for modeling purposes because there is abundant information on reaction kinetics and the relationships between crystal morphology and growth conditions. In this paper we describe the LB approach used, address strategies for properly conserving mass at a surface of a growing ?06₯730300crystal?06₯730308, the scaling of the calculations to the actual physical problem, and the conditions for growth of dendritic versus compact crystals.
MR21B-1797
Insights into noble gas incorporation in silicates from atomistic and ab-initio simulations
We discuss recent progress in simulating trace- element incorporation in minerals and melts and the implication of these results for analysing and understanding recent partitioning experiments. We use classical forcefield calculations and density functional theory to study the incorporation of noble gases in solid silicates. It is crucial to take into account the local structural environment of each ion in the solid and the change in this environment as a surrounding atom move following the introduction of a foreign atom or atoms. Several different energy terms contribute to the overall energetics of the partitioning process with the strain energy being only one of these. We contrast the behaviour of light atoms such as He with that of heavier noble gases. Applications include incorporation of trace elements in olivines and solid solutions and trace-partitioning behaviour of the noble gases, uranium and thorium. We discuss the importance of allowing explicitly for segregation of trace elements to grain boundaries and other interfaces, particularly when there is a size mismatch between the trace element and the ions in the host solid. In addition, the influence of pressure and temperature on solution energies is studied by direct free energy minimisation methods which indicates a large dependence of energy with pressure but little variation with temperature. We also estimate contributions to the overall solubilities from intrinsic and extrinsic vacancies. Finally, we present calculated solubilities of noble gases in bulk minerals and interfaces in a wide range of minerals as a function of temperature and pressure.