MR12A-01 INVITED
First Principles Studies Of Hydrous Minerals Under High Pressure
Hydrogen atom is the lightest and the most abundant element in the universe. Existence of hydrogen in the deep earth is paid attention because it is known to affect the physical properties of the earth constituent minerals. Experimental techniques for determining the hydrogen positions under high pressure are still developing and the effects of hydrogen on physical properties have not been quantitatively clarified. We have investigated by first principles, the effects of hydrogen on the structural, vibrational and elastic properties of hydrous and nominally anhydrous minerals (AlOOH, MgSi2O6H2, hydrous wadsleyite etc.) under high pressure condition and found that the strength of the hydrogen bond depends significantly on pressure as well as the local geometry of the hydrogen bonds. The nature of hydrogen bonds correlates with compressional, elastic and optical properties of minerals. Especially, the hydrogen bonds in δ-AlOOH and phase D, which are thought to be carriers of water into deep mantle, are predicted to have symmetric forms at lower mantle pressures. Our findings are consistent with recent high-pressure experimental results, suggesting that the density functional first principles techniques can be efficient tools for investigating the hydrous systems under high pressure. Research supported in part by JSPS.
MR12A-02 INVITED
Quantum Monte Carlo Computations for Equations of State, Phase Transitions, and Elasticity of Silica
We have performed Quantum Monte Carlo (QMC) computations for silica in the quartz, stishovite, and α- PbO2 structures as functions of compression as benchmark computations. QMC uses no approximate density functional, and the many-body, correlated, Schrödinger equation is effectively solved stochastically. In spite of the great success of DFT there are still some fundamental problems that need improvement. First is the need for increased accuracy for some rather ordinary materials such as silica. Although the local density approximation (LDA) gives excellent results for individual silica phases, such as the CaCl2 transition [1,2], it is not so good for comparing energetics of very different structures, such as quartz versus stishovite. LDA predicts stishovite to be the stable ground state structure rather than quartz. One of the first great successes of the Generalized Gradient Approximation (GGA) was to give the correct energy difference between quartz and stishovite [3]. Less well appreciated is the fact that almost all other properties, such as structure, equations of state, elastic constants, etc., are worse with the GGA than the LDA. Our QMC results will be used to improve density functionals, and show the way towards more accurate computations for Earth materials. Thermal contributions are included using density functional perturbation theory with the code ABINIT. We have also computed the shear elastic constant c11-c12 in stishovite, which is associated with the phase transition to the CaCl2 structure [1], with QMC. We find excellent agreement with experiments. We find that the main differences between QMC and DFT are crystalline phase dependent energy and pressure shifts. This work is supported by NSF grants EAR-0530282, EAR-0310139, and by DOE contract DE-FG02-99ER45795 to John Wilkins. Computations were performed on Blueice at NCAR under a BTS grant, Tungsten and Abe at NCSA, Franklin at NERSC within the friendly user program, and at the Carnegie Institution of Washington. [1] Cohen, R.E., First-principles predictions of elasticity and phase transitions in high pressure SiO2 and geophysical implications, in High Pressure Research in Mineral Physics: Application to Earth and Planetary Science, M.H. Manghnani and Y. Syono, Editors. 1992, AGU: Washington, D.C. p. 425-432. [2] Kingma, K.J., R.E. Cohen, R.J. Hemley, and H.-K. Mao, Transformation of silica to a denser phase at lower- mantle pressures. Nature, 1995. 374: p. 243-245. [3] Hamann, D.R., Generalized gradient theory for silica phase transitions. Phys. Rev. Lett., 1996. 76(4): p. 660- 663.
MR12A-03 INVITED
Ab-initio study of transition-metal compounds through a consistent, linear-response LDA+U approach
Hubbard U-corrected LDA or GGA have proven very effective in describing several systems characterized by strongly localized electronic states for which these standard approximations to DFT fail. I introduce here our scheme to evaluate the effective electronic interaction of the "+U" functional in a fully consistent way. This approach is based on the linear response of the considered system to a potential shift acting on the localized orbitals of the correlated sites. Using the occupations of these orbitals as the relevant electronic degrees of freedom we compute the needed on-site electronic coupling as the difference between the inverse of the bare and fully interacting response matrices. The computed U thus corresponds to the effective kernel of the electron-electron on-site interaction entering the second quantization expression of the "+U" energy functional. In this way the strength of the "+U" correction is consistently evaluated from the same DFT scheme we aim to correct and the LDA+U is transformed in a completely ab-initio method with no need for any semi-empirical or apriori evaluation of the effective coupling. The results are also largely independent on the choice of the localized orbitals: the same occupation matrix that enters the expression of the "+U" correction is consistently used to compute the effective interaction parameter. A further development in this approach also allowed us to obtain the the effective U "auto-consistently" from a correlated (LDA+U) ground state. With this approach we successfully studied the structural, electronic, chemical and electrochemical properties of several transition-metal compounds. Examples of applications will include minerals in the Earth's interior [1], cathode materials for next-generation lithium batteries [2] and catalytic reactions on molecules [3,4]. [1] M. Cococcioni and S. de Gironcoli, PRB (2005). [2] F. Zhou, M. Cococcioni, A. C. Marianetti, D. Morgan and G. Ceder, PRB (2004). [3] H. J. Kulik, M. Cococcioni, D. Scherlis and N. Marzari, PRL (2006). [4] D. A. Scherlis, M. Cococcioni, P. Sit, and N. Marzari, JPC (2007).
MR12A-04
Quinone-Hydroquinone complexes as components of humic acids: Theoretical studies of structure, stability and visible-UV spectra
Quinones and hydroquinones form complexes called quinhydrones which have optical absorption energies lying below those of either of the separate components. It has been suggested that these and similar complexes are components of humic acids which strongly influence their visible-UV spectra and are responsible for the smooth, featureless exponentially rising absorption of humics in the visible and near UV. We have used the methods of molecular quantum mechanics to calculate the structure, energetics and spectra of such complexes. Since the chemical forces holding the quinone and hydroquinone together are of van der Waals and H-bonding type, it is necessarily to use a high level quantum mechanical method which accurately treat such interactions. We rely mainly upon 2nd order Moller-Plesset theory with doubly polarized triple zeta basis sets for the structural and energetic studies and time dependent density functional theory with the B3LYP potential for the spectral studies. Our calculated C-C intermolecular distance for quinhydrone is in excellent agreement with the experimental value. Our energetic calculations indicate that the free energy of formation of quinhydrone is negative in the gas-phase but positive in aqueous solution. We have established that our TD DFT methods give accurate spectra for the gas-phase molecule quinone. The complexes give lower energy absorptions, but these are quite sensitive to geometry and to the charge state of the quinhydrone complex. The transitions within the quinhydrone complex are mainly from the highest energy occupied MO to the lowest energy empty MO, as expected from the donor acceptor model. We are now carrying out studies of the effects of dynamics and structural perturbation on the spectra.
MR12A-05 INVITED
Applications of Efficient First-Principles Methods in Mineral Sciences: the SIESTA Program
The possibility of treating large systems with first-principles electronic-structure methods has opened up new opportunities in many disciplines. In particular, the SIESTA program (www.uam.es/siesta) has become quite popular and is increasingly being used by researchers in geosciences and biophysics. The code's efficiency for large problems stems from the use of strictly localized basis sets and from the implementation of linear-scaling algorithms which can be applied to suitable systems. A very important feature of the program is that its accuracy and cost can be tuned in a wide range, from very fast exploratory calculations to highly accurate simulations matching the quality of other approaches, such as plane-wave and all-electron methods. Even at the lower end of the range it provides a fully self-consistent solution that is more accurate than empirical approaches in problems involving charge transfer or coordination changes. Apart from computing the electronic structure (including the effects of spin), the program can perform a full range of molecular dynamics simulations (including constant-stress) and structural optimizations, making it suitable for a wide spectrum of research topics in mineral sciences, some of which will be highlighted in the talk. This presentation is supported by the European Science Foundation (ESF) under the EUROCORES Programme EuroMinScI (www.esf.org/eurominsci), through contract No. ERAS-CT-2003-980409 of the European Commission, DG Research, FP6. http://www.uam.es/siesta
MR12A-06
Multi-Scale Modeling of Post-Perovskite Deformation Processes: From Atomic Scale to Polycrystal Plasticity
The D" layer that extends up to several hundred kilometers at the transition between the silicate mantle and the metallic core is an essential feature of the mantle's convecting system. This layer is known to display a strong and heterogeneous seismic anisotropy. At the same time, seismic velocities change laterally, indicating large thermal and chemical variation. It is generally accepted that D" is essentially composed of (Mg,Fe)SiO3, which can be found in the Perovskite (Pv) structure or in the post perovskite (PPv) structure depending on the temperature of the region of interest. The interpretation of the origin of the anisotropy of D" is still in debate. Essentially, two propositions appear: strain-induced Crystal Preferred Orientation (CPO) of mineral (MgSiO3 PPv or (Mg,Fe)O2) and/or shape preferred orientation (SPO) of inclusions. With the last assumption, to produce azimuthal anisotropy, inclusions have to be inclined or strain-induced CPO has to be invoked. To explore the contribution of CPO due to plastic flow of PPv to seismic anisotropy in the D" layer, we use a multi- scale modeling approach that couple atomistic/continuum models of dislocations at D" pressures to polycrystal plasticity simulations (using a viscoplastic self-consistent (VPSC) model). Indeed to model the development of CPO, we need to determine a fundamental parameter that controls the activity of a dislocation glide system: the critical resolved shear stress (CRSS). Nowadays, dislocation core properties can be calculated from first principles calculation through the Peierls- Nabarro model using the generalized stacking faults approach. Dislocation properties such as planar core spreading and Peierls stresses are thus modeled under a pressure of 120 GPa for 10 potential glide systems of the PPv structure. As lattice friction is commonly considered as the major factor in plastic deformation of minerals, we have then chosen to take the CRSS proportional to the Peierls stresses for additional VPSC simulations. VPSC calculations show that CPO development are very sensible to CRSS, here with a strong alignment of [010] in the shear direction and of (001) in the shear plane, reflecting the high activity of [100](010), [110](001) and [110](11̄0) glide systems. Finally, these results also illustrate that, when several slip systems are activated with similar contributions, is not possible to directly determine the dominant glide system from the pole figures.
MR12A-07 INVITED
Phase stability and shear softening in CaSiO3 perovskite at high pressure
We predict the phase diagram of CaSiO3 perovskite, finding the tetragonal I4/mcm structure transforming to cubic Pm\bar 3m with increasing temperature. The transition temperature is 1150 K at 0 GPa, and 2450 K at 140 GPa. The c/a ratio of the tetragonal structure is 1.018 at 100 GPa and increases on compression, as does the static enthalpy difference between tetragonal and cubic structures. The elastic constants of the tetragonal phase at static conditions differ substantially from those of the cubic phase with the Voigt-Reuss-Hill shear modulus 29 % less at 100 GPa. Computations are based on density functional theory in the local density and generalized gradient approximations. The phase diagram and high temperature elastic constants are computed using a mean field theory with parameters of the Landau potential determined via structurally constrained density functional theory calculations. We present a simple scheme for systematically searching for the ground state over all perovskite structures derivable from octahedral rotations within the context of symmetry- preserving relaxation, which confirms tetragonal I4/mcm as the ground state in density functional theory. We argue that the experimental x-ray diffraction pattern can be explained by the I4/mcm phase by considering the development of preferred orientation under uniaxial compression.
MR12A-08
VLab: a service oriented architecture for first principles computations of planetary materials properties
We present an overview of the VLab, a system developed to handle execution of extensive workflows generated by first principles computations of thermoelastic properties of minerals. The multiplicity (102-3) of tasks derives from sampling of parameter space with variables such as pressure, temperature, strain, composition, etc. We review the algorithms of physical importance that define the system's requirements, its underlying service oriented architecture (SOA), and metadata. The system architecture emerges naturally. The SOA is a collection of web-services providing access to distributed computing nodes, controlling workflow execution, monitoring services, and providing data analyses tools, visualization services, data bases, and authentication services. A usage view diagram is described. We also show snapshots taken from the actual operational procedure in VLab. Research supported by NSF/ITR (VLab) http://www.vlab.msi.umn.edu