The significance of ferroelectric phase transitions for the Earth and planetary interiors

With the compelling evidence for orthorhombic perovskite ABO₃ structures in a major part of the earth's mantle, the question of whether any of these are ferroelectric (FE) or antiferroelectric should be of supreme importance. To answer this question the authors have conducted dielectric property studies at pressures up to 5.5 GPa on single crystals and polycrystalline aggregates of BaTiO₃ as well as other FE materials representing FE polarization mechanisms. The results indicate that: (1) h.p.-induced FE phases are indeed likely to play an important part in the earth's mantle; and (2) existing FE state, occuring as elastic and dielectric inhomogeneities, can persist to much higher pressures than expected on the basis of the pressure dependence of isolated and unclamped FE crystals. It is suggested that the presence of FE states in the earth's mantle may be responsible for some of its anomalous elastic and dielectric features, especially its attenuating properties. It is conjectured that effects of FE states on properties of the Jovian planets might even be more prominant, especially those involving dielectric constant and polarization.     

Tetrahedral occupancy of ferric iron in (Mg,Fe)O: Implications for point defects in the Earth's lower mantle

We investigated the concentration and site occupation of ferric iron (Fe³⁺) in (Mg,Fe)O to understand the influence of point defects on transport properties such as atomic diffusion, electrical conductivity and viscosity. We conducted Mössbauer spectroscopy of (Mg_0.8Fe_0.2)O single crystals synthesized at temperatures from 1673 to 2273 K and pressures from 5 to 15 GPa with Re–ReO₂ and Mo–MoO₂ oxygen fugacity buffers. The isomer shift of the Mössbauer spectra suggests that Fe³⁺ occupies mostly the tetrahedral site at reduced conditions and both the octahedral and tetrahedral sites at oxidized conditions. We formulate a thermodynamic model of point defect dissolution in (Mg,Fe)O which suggests that unassociated tetrahedral Fe³⁺ is more stable than unassociated octahedral Fe³⁺ at high-pressure and low oxygen fugacity due to the effect of configurational entropy. The pressure dependence of Fe³⁺ concentration indicates a change in the dominant site occupancy of Fe³⁺: (1) Fe³⁺ in the tetrahedral site, (2) Fe³⁺ in the octahedral site, and (3) defect clusters of Fe³⁺ and cation vacancy, in the order of increasing oxygen fugacity and decreasing pressure. This is in reasonable agreement with previously reported experiments on Fe³⁺ concentration, Mg–Fe interdiffusivity and electrical conductivity. We consider it plausible that (Mg,Fe)O accommodates Fe³⁺ in the tetrahedral site down to the lower mantle. Based on our results and available experimental data, we discuss the solubility competition between Fe³⁺ and protons (H⁺), and its implications for transport properties in the lower mantle.     

Iron oxidation state in lower mantle mineral assemblages: II. Inclusions in diamonds from Kankan, Guinea

Inclusions of ferropericlase and former (Mg,Fe)(Si,Al)O₃ perovskite in diamonds from Kankan, Guinea believed to originate in the lower mantle were studied using Mössbauer spectroscopy to determine Fe³⁺/ΣFe. Fe³⁺ concentration in the (Mg,Fe)(Si,Al)O₃ inclusion is consistent with empirical relations relating Fe³⁺/ΣFe to Al concentration, supporting the inference that it crystallised in the perovskite structure at lower mantle conditions. In ferropericlase there is a nearly linear variation of trivalent cation abundance with monovalent cation abundance, suggesting a substitution of the form Na_0.5M_0.5³⁺O (M=Fe³⁺, Cr³⁺, Al³⁺). Excess positive charge is likely balanced by cation vacancies, where their abundance is observed to increase with increasing iron concentration, consistent with high-pressure experiments. The abundance of cation vacancies is related to oxygen fugacity, where ferropericlase inclusions from Kankan and São Luiz (Brazil) are inferred to have formed at conditions more oxidising than Fe-(Mg,Fe)O equilibrium, but more reducing than Re-ReO₂ equilibrium. Fe²⁺/Mg partition coefficients between (Mg,Fe)(Si,Al)O₃ and ferropericlase were calculated for inclusions co-existing in the same diamond using Mössbauer data and empirical relations based on high-pressure experimental work. Most values are consistent with high-pressure experiments, suggesting that these inclusions equilibrated at lower mantle conditions. The measured ferropericlase Fe³⁺ concentrations are consistent with diamond formation in a region of redox gradients, possibly arising from the subduction of oxidised material into reduced lower mantle. Reduction of carbonate to form ferropericlase and diamond is consistent with a slight shift of Kankan δ¹³C values to isotopically heavy compositions compared to the worldwide dataset, and could supply the oxygen necessary to satisfy the high Fe³⁺ concentration in (Mg,Fe)(Si,Al)O₃ perovskite, as well as account for the high proportion of ferropericlase in the lower mantle paragenesis. The heterogeneity of lower mantle diamond sources indicates that the composition of lower mantle diamonds do not necessarily reflect those of the bulk mantle.     

The boundary conditions for simulations of a shake-table experiment on the seismic response of 3D slope

Boundary conditions can significantly affect a slope’s behavior under strong earthquakes. To evaluate the importance of boundary conditions for finite element (FE) simulations of a shake-table experiment on the slope response, a validated three-dimensional (3D) nonlinear FE model is presented, and the numerical and experimental results are compared. For that purpose, the robust graphical user-interface “SlopeSAR”, based on the open-source computational platform OpenSees, is employed, which simplifies the effort-intensive pre- and post-processing phases. The mesh resolution effect is also addressed. A parametric study is performed to evaluate the influence of boundary conditions on the FE model involving the boundary extent and three types of boundary conditions at the end faces. Generally, variations in the boundary extent produce inconsistent slope deformations. For the two end faces, fixing the y-direction displacement is not appropriate to simulate the shake-table experiment, in which the end walls are rigid and rough. In addition, the influence of the length of the 3D slope’s top face and the width of the slope play an important role in the difference between two types of boundary conditions at the end faces (fixing the y-direction displacement and fixing the (y, z) direction displacement). Overall, this study highlights that the assessment of a comparison between a simulation and an experimental result should be performed with due consideration to the effect of the boundary conditions.     

Determination of phase transition pressures of ZnTe under quasihydrostatic conditions

Pressure behavior of ZnTe at room temperature was studied using an X-ray energy dispersive method on a DIA type cubic anvil apparatus (SAM-85) at NSLS-X17B1. By using powdered polyethylene, the sample and NaCl for a pressure scale were held under quasihydrostatic conditions, which were confirmed by X-ray diffraction method. Two high-pressure phase transitions were confirmed using X-ray powder diffraction simultaneously with electrical resistance measurements. The phase transition pressures under quasihydrostatic conditions were determined to be 9.6 GPa, at which the resistance increased, and 12.0 GPa, which was the midpoint of a large resistance decrease. Errors in the pressure determinations were estimated to be less than 0.2 GPa. These pressure values may depend on grain size and anisotropic stress effects on the calibrant. From X-ray observation of ZnTe, the bulk modulus of the zinc blende structure was calculated to beK ₀=51(3) GPa andK ₀ =3.6(0.8), and the first transition at 9.6 GPa was found to have about 9% volume change. It was consistent with an anomaly in the pressure generating curves.     

The system enstatite-wollastonite at high pressures and temperatures,with emphasis on diopside

High-pressure phase transformations for three intermediate compositions (including diopside) in the system enstatite (MgSiO₃)-wollastonite (CaSiO₃) were investigated in the pressure range 100–300 kbar at about 1000°C in a diamond-anvil press coupled with laser heating. The phase behaviour of the two end components (enstatite and wollastonite) at high pressure has been reported earlier. The results of this study reveal that there is very limited solid solution of diopside (CaMgSi₂O₆) in the various high-pressure phase assemblages of enstatite. At pressures greater than about 200 kbar, diopside and a composition between diopside and wollastonite were found to transform into non-quenchable phases, as does wollastonite. It is thought probable that diopside and wollastonite form solid solutions with the perovskite structure at high pressure, but that on release of pressure it is not possible to preserve the high-pressure modification.     

Simplified approximate method for analysis of rocking systems accounting for soil inelasticity and foundation uplifting

A simplified approximate method to analyze the rocking response of SDOF systems lying on compliant soil is introduced, accounting for soil inelasticity and foundation uplifting. The soil–foundation system is replaced by a nonlinear rotational spring, accompanied by a linear rotational dashpot, and linear horizontal and vertical springs and dashpots. Considering a square footing on clay under undrained conditions, the necessary moment–rotation (M–θ) relations are computed through monotonic pushover finite element (FE) analyses, employing a thoroughly-validated constitutive model. Cyclic pushover analyses are performed to compute the damping–rotation (C_R–θ) relations, necessary to calibrate the rotational dashpot, and the settlement–rotation (Δw–θ) relations, required to estimate the dynamic settlement. The effectiveness of the simplified method is verified through dynamic time history analyses, comparing its predictions with the results of 3D FE analyses. The simplified method is shown to capture the entire rotation time history θ(t) with adequate accuracy. The latter is used to compute the time history of dynamic settlement w(t), employing a simplified approximate procedure. The proposed simplified method should, by no means, be considered a substitute for more sophisticated analysis methods. However, despite its limitations, it may be utilized for (at least preliminary) design purposes.     

High-pressure reconnaissance investigation in the system Mg3Al2Si3O12-Fe3Al2Si3O12

Two synthetic end-members and two natural solid solutions of the system Mg₃Al₂Si₃O₁₂-Fe₃Al₂Si₃O₁₂ have been found to display successive phase transformations at increasingly high pressures when they were compressed in a diamond-anvil cell and heated with a YAG laser to temperatures of approximately 1400–1800°C. X-ray diffraction studies of the quenched samples show that the iron-rich garnets apparently first transform to a garnet-related high-pressure phase, then disproportionate into a mixture of magnesiowüstite plus an unknown phase(s). The latter phase(s) may further transform to a still denser unknown phase(s). The ultimate high-pressure phase may be a perovskite-like structure as was previously found for the magnesium-rich garnets. One of the unknown phases may be the high-pressure phase of Al₂O₃ · nSiO₂ compounds. Magnesium-rich garnets display similar phase transformations as do the iron-rich garnets with the exception of the garnet-related high-pressure phase. These results disagree with a previous interpretation for the high-pressure phase of iron-silicate garnets recovered in shock-wave experiments reported by Ahrens and Graham (1972).     

Geobarometry of ultramafic xenoliths from Loihi Seamount,Hawaii, on the basis of CO2 inclusions in olivine

Abundant fluid inclusions in olivine of dunite xenoliths (～1–3 cm) in basalt dredged from the young Loihi Seamount, 30 km southeast of Hawaii, are evidence for three coexisting immiscible fluid phases—silicate melt (now glass), sulfide melt (now solid), and dense supercritical CO₂ (now liquid + gas)—during growth and later fracturing of some of these olivine crystals. Some olivine xenocrysts, probably from disaggregation of xenoliths, contain similar inclusions.Most of the inclusions (2–10 μm) are on secondary planes, trapped during healing of fractures after the original crystal growth. Some such planes end abruptly within single crystals and are termed pseudosecondary, because they formed during the growth of the host olivine crystals. The “vapor” bubble in a few large (20–60 μm), isolated, and hence primary, silicate melt inclusions is too large to be the result of simple differential shrinkage. Under correct viewing conditions, these bubbles are seen to consist of CO₂ liquid and gas, with an aggregate ? = ～ 0.5–0.75 g cm^?3, and represent trapped globules of dense supercritical CO₂ (i.e., incipient “vesiculation” at depth). Some spinel crystals enclosed within olivine have attached CO₂ blebs. Spherical sulfide blebs having widely variable volume ratios to CO₂ and silicate glass are found in both primary and pseudosecondary inclusions, demonstrating that an immiscible sulfide melt was also present.Assuming olivine growth at ～ 1200°C and hydrostatic pressure from a liquid lava column, extrapolation of CO₂P-V-T data indicates that the primary inclusions were trapped at ～ 220–470 MPa (2200–4700 bars), or ～ 8–17 km depth in basalt magma of ? = 2.7 g cm^?3. Because the temperature cannot change much during the rise to eruption, the range of CO₂ densities reveals the change in pressure from that during original olivine growth to later deformation and rise to eruption on the sea floor. The presence of numerous decrepitated inclusions indicates that the inclusion sample studied is biased by the loss of higher-density inclusions and suggests that some part of these olivine xenoliths formed at greater depths.     

Distribution patterns of gypsum and kalistrontite in a dry lake basin of the southwestern Kalahari (Omongwa pan,Namibia)

The deposits of the Omongwa pan, southwestern Kalahari, Namibia, are partly gypsiferous and locally contain small amounts of kalistrontite. The distribution patterns of gypsum (CaSO₄.2H₂O) and kalistrontite (K₂Sr(SO₄)₂) are mainly determined by groundwater depth and by the lithological composition of the deposits. The latter determines the hydraulic conductivity for saturated and unsaturated flow and therefore controls the depth of the interval within the deposits where evaporation and mineral precipitation can take place. Other factors that affect the distribution of the evaporite minerals are the patterns of groundwater flow within the basin and the occurrence of a redistribution of salts during short flooding stages. Gypsum occurs as crystals of four distinct morphological types, which each developed in different conditions: prismatic crystals formed both as subaqueous precipitates and as crystals that developed within the sediment matrix of a brine-saturated surface layer; sub/euhedral tabular crystals formed within brine-filled macropores; hemi-bipyramidal crystals formed in a nearly permanently brine-saturated part of a subsurface horizon (phreatic zone), as the product of recrystallization of an older gypsum occurrence; and lenticular crystals formed by intrasediment growth in the vadose zone. The development of these morphological types of gypsum crystals is marked by differences in the differential inhibition of growth of the various crystal forms. The further growth of lenticular crystals, which are most strongly affected by differential growth inhibition, is still characterized by a difference in degree of inhibition between the {1 11} and {1 03} forms, expressed as a change in orientation of the plane of flattening. Copyright © 1999 John Wiley & Sons, Ltd.     

Dynamischer Zinkaustausch an umweltrelevanten anorganischen Festphasen

The fixation of slightly soluble substances on inert carriers offers the possibility to examine the exchange of heavy metals at these phases under dynamic conditions. Investigations are presented for the exchange of Zinc under various conditions (pH, different concentrations of other cations such as Na⁺ and Ca²⁺ or of NTA as a chelating agent) on fixed MnO₂-hydrate, Fe(OH₃), CaCO₃ and a natural clay-slate. Batch-experiments showed that all phases are able to sorb Zinc with MnO₂-hydrate as the best sorbent for this metal (highest sorption constant and capacity). Column desorption experiments demonstrate that a decrease of the pH of the aqueous phase or an addition of NTA remobilise Zinc from MnO₂-hydrate, Fe(OH)₃, CaCO₃ and also from the clay-slate. The same effect of releasing sorbed metal (with exception of the CaCO₃-phase) show Na⁺ or Ca²⁺ containing solutions. Therefore Zinc is bonded under natural conditions on the oxihydrates and the clay-slate in a labile, ion-exchangeable form in contrast to CaCO₃, on which Zinc is sorbed probably as the carbonate.     

High-pressure phase transformations and isothermal compression in CaTiO3 (perovskite)

A polycrystalline CaTiO₃ (perovskite) was investigated under static pressures up to 38 GPa and temperatures up to 1000°C by using a diamond anvil pressure cell, a YAG laser, and the ruby fluorescence pressure calibration system. In situ x-ray diffraction data reveal that at room temperature, the orthorhombic CaTiO₃(I) transforms into a hexagonal CaTiO₃(II) at ∼ 10 GPa with a volume of change of 1.6%. At 1000°C, the orthorhombic CaTiO₃(I) first transforms into a tetragonal CaTiO₃(III) at 8.5 GPa and then transforms further into a hexagonal CaTiO₃(II′) at ∼ 15 GPa with molar volume changes of 0% and 1.6%, respectively. All three high-pressure polymorphs found in this study are nonquenchable.Isothermal compressibility of the orthorhombic CaTiO₃ was derived from measurements under truly hydrostatic environments (i.e., ⩽ 10.4 GPa). By assuming K^′₀ = 5.6 obtained ultrasonically on SrTiO₃ perovskite, the value of the bulk modulus (K₀) was calculated with the Birch-Murnaghan equation to be 210 ± 7 GPa.     

Bridge–embankment interaction under transverse ground excitation

Seismic performance and dynamic response of bridge–embankments during strong or moderate ground excitations are investigated through finite element (FE) modelling and detailed dynamic analysis. Previous research studies have established that bridge–embankments exhibit increasingly flexible performance under high‐shear deformation levels and that soil displacements at bridge abutment supports may be significant particularly in the transverse direction. The 2D equation of motion is solved for the embankment, in order to evaluate the dynamic characteristics and to describe explicitly the seismic performance and dynamic response under transverse excitations accounting for soil nonlinearities, soil–structure interaction and imposed boundary conditions (BCs). Using the proposed model, equivalent elastic analysis was performed so as to evaluate the dynamic response of approach embankments while accounting for soil–structure interaction. The analytical procedures were applied in the case of a well‐documented bridge with monolithic supports (Painter Street Overcrossing, PSO) which had been instrumented and embankment participation was identified from its response records after the 1971 San Fernando earthquake. The dynamic characteristics and dynamic response of the PSO embankments were evaluated for alternative BCs accounting for soil–structure interaction. Explicit expressions for the evaluation of the critical embankment length L_c are provided in order to quantify soil contribution to the overall bridge system under strong intensity ground excitations. The dynamic response of the entire bridge system (deck–abutments–embankments) was also evaluated through simplified models that considered soil–structure interaction. Results obtained from this analysis are correlated with those of detailed 3D FE models and field data with good agreement. Copyright © 2007 John Wiley & Sons, Ltd.     

Plastic deformation of garnets: systematics and implications for the rheology of the mantle transition zone

Plastic properties of materials with garnet structure have been studied under wide temperature conditions, ranging from room temperature to 95% of the melting temperatures, using uniaxial compression and hot microhardness tests. Garnets studied include single crystals of oxide garnets (Y₃Al₅O₁₂, Gd₃Ga₅O₁₂ and Y₃Fe₅O₁₂) and silicate garnets (various solid solutions, including grossular, almandine, andradite, pyrope, spessartine and uvarovite). Both uniaxial compression and hot hardness tests indicate that there is a general trend in the plasticity of garnets when the data are compared at normalized conditions (T/T_m andσ/μ), and that the resistance to plastic deformation in garnets is significantly higher than most of the other minerals in the Earth's mantle. Based on both stress-dip tests and microstructural observations, it is proposed that the creep strength of garnet is largely controlled by the resistance to dislocation glide rather than by recovery processes. This conclusion is consistent with the high Peierls stress inferred from the hot hardness tests. The high Peierls stress in garnets is, presumably, due to the large unit cell (i.e., long Burgers vectors) and/or the bcc packing, which are common to all garnets. We postulate, therefore, that the present results can be applied to the strength of high-pressure garnet (majorite) and suggest that garnet-rich layers in the Earth, such as subducted oceanic crust in the transition zone or a possibly garnet-rich (bottom part of the) transition zone, will be considerably stronger than surrounding regions.     

Dielectric properties of the first 100 meters of the Moon

Reviewing 92 measurements of lunar sample dielectric constant versus density at frequencies above 100 kHz, gives the relationK′ = (1.93 ± 0.17)^p by regression analysis, where K′ is the dielectric constant of a soil or solid at a density ofpg/cm³. This formula is the geometric mean between the dielectric constant of vacuum (1) and the zero porosity dielectric constant of lunar material. Similarly, the loss tangent (D) can be described byD = [(0.00053 ± 0.00056) + (0.00025 ± 0.00009)C]p whereD is the loss tangent at densitypg/cm³ withC percent of total FeO + TiO₂ (approximately proportional to ilmenite content). Using the density versus depth relations derived from lunar surface core tubes, and from laboratory studies of lunar soil compression gives a model of the dielectric properties as a function of depth in the lunar regolith. The dielectric constant increases smoothly with depth, as a function of the soil compaction only. The loss tangent, however, is more sensitive to the ilmenite content than it is to density. Neither dielectric constant nor loss tangent varies significantly with the temperature observed in a lunar day.     

Phase relations in the system LiFMgF2 at elevated pressures: Implications for the proposed mixed-oxide zone of the earth's mantle

Solvi and liquidi for various LiFMgF₂ mixtures have been determined at pressures up to 40 kbar by differential-thermal-analysis in a piston-cylinder high-pressure device. The melting curves of pure LiF and MgF₂ were also studied and the initial slopes (dT_m/dP)_{P = 0} were found to be 11.2 and 8.3°C/kbar, respectively. The eutectic composition (LiF)_0.64(MgF₂)_0.36 is independent of pressure to 35 kbar and the eutectic temperature rises approximately 6.3°C per kbar. Initial slopes of 11°C/kbar and 35°C/kbar are inferred for the melting curves of MgO and SiO₂ (stishovite) respectively, on the basis of data for their structural analogue compounds. The observed solid solution of LiF in MgF₂ and other evidence suggest the possibility of solid solution in the system (Mg,Fe)OSiO₂ (stishovite) under mantle conditions which may have important consequences for the elastic properties of a “mixed-oxide” zone of the earth's mantle.     

High-pressure phases of CaCO3: Crystal structure prediction and experiment

Post-aragonite phase of CaCO₃, experimentally known to be stable above 40 GPa [S. Ono, T. Kikegawa, Y. Ohishi, J. Tsuchiya, Post-aragonite phase transformation in CaCO₃ at 40 GPa, Am. Mineral. 90 (2005) 667–671], is believed to be a major carbon-containing mineral in the Earth's mantle. Crystal structure of this mineral phase could not be solved using experimental data or traditional theoretical simulation methods and remained a controversial issue. Using a combination of advanced ab initio simulation techniques and high-pressure experiment, we have been able to determine the crystal structure of CaCO₃ post-aragonite. Here, we performed simulations with the USPEX code [C.W. Glass, A.R. Oganov, and N. Hansen, (in preparation). USPEX: a universal structure prediction program], which is based on an evolutionary algorithm using ab initio free energy as the fitness function. This novel methodology for crystal structure prediction, which uses only the chemical composition as input, is described in detail. For CaCO₃, we identify a number of energetically competitive structures, the most stable of which closely matches the experimental powder diffraction pattern and, in agreement with experiment, becomes more stable than aragonite above 42 GPa. This structure belongs to a new structure type, which is also adopted by the high-pressure post-aragonite phases of SrCO₃ and BaCO₃. It has 2 formula units in the orthorhombic unit cell (space group Pmmn) and contains triangular CO₃²⁻ ions and Ca²⁺ ions in the 12-fold coordination. Above 137 GPa, a pyroxene-type structure (space group C222₁) with chains of CO₄⁴⁻ tetrahedra becomes more stable than post-aragonite. For MgCO₃, this structure becomes more stable than magnesite above 106 GPa and is a good candidate structure for MgCO₃ post-magnesite.     

Seismic ground response at Lotung: Hysteretic elasto-plastic-based 3D analyses

This paper presents a non-linear finite element study to back-interpret the free field seismic response recorded at the Lotung Large-Scale Seismic Test site. The study is carried out in the time domain by the Finite Element (FE) code PLAXIS 3D, considering the vertical wave propagation of both the horizontal components of motion. The non-linear soil behaviour is simulated through a constitutive model, the Hardening Soil model with Small-Strain Stiffness (HSsmall), capable of describing the cyclic response of the material at different strain levels. In the paper, the constitutive response of the HSsmall model is firstly investigated through numerical simulations of strain-controlled cyclic shear tests under single and multi-directional conditions at low strain levels. Then, it is adopted to back-analyse the recorded free field seismic response, comparing the FE numerical results to the in-situ down-hole and surface signals recorded during two earthquakes occurred on May 20th and July 17th 1986, characterized by different peak ground accelerations.     

Effect of oxygen partial pressure on the dislocation recovery in olivine: a new constraint on creep mechanisms

The dislocation annihilation rate in experimentally deformed olivine single crystals was measured as a function of oxygen partial pressure (P_O2). It was shown that the dislocation annihilation rate decreased with increasing P_O2. This result is inconsistent with the reported P_O2 dependence of creep rate () in single olivine crystals, thus indicating that the creep in single olivine crystals is not rate-controlled by recovery, under the experimentally investigated conditions.     

Molecular modeling of the 10-Å phase at subduction zone conditions

Molecular dynamics (MD) modeling of the 10-Å phase, Mg₃Si₄O₁₀(OH)₂·xH₂O, with x=2/3, 1.0 and 2.0 shows complex structural changes with pressure, temperature and water content and provides new insight into the structures and stabilization of these phases under subduction zone conditions. The structure(s) of this phase and its role as a reservoir of water in the mantle have been controversial, and these calculations provide specific predictions that can be tested by in situ diffraction studies. At ambient conditions, the computed structures of talc (x=0) and the 10-Å phases with x=2/3 and 1.0 are stable over the 350-ps period of the MD simulations. Under these conditions, the 10-Å phases show phlogopite-like layer stacking in good agreement with previously published structures based on powder X-ray diffraction data for samples quenched from high-pressure and high-temperature experiments. The calculations show that the 10-Å phase with x=2.0 is unstable at ambient conditions. The computed structures at P=5.5 GPa and T=750 K, well within the known stability field of the 10-Å phase, change significantly with water content, reflecting changing H-bonding configurations. For x=2/3, the layer stacking is talc-like, and for x=1.0, it is phlogopite-like. The calculations show that transformation between these two stackings occurs readily, and that the talc-like stacking for the x=2/3 composition is unlikely to be quenchable to ambient conditions. For x=2.0, the layer stacking at P=5.5 GPa and T=750 K is different than any previously proposed structure for a 10-Å phase. In this structure, the neighboring basal oxygens of adjacent magnesium silicate layers are displaced by b/3 (about 3 Å) resulting in the Si atoms of one siloxane sheet being located above the center of the six-member ring across the interlayer. The water molecules are located 1.2 Å above the center of all six-member rings and accept H-bonds from the OH groups located below the rings. The b/3-displaced structure does not readily transform to either the talc-like or phlogopite-like structure, because neither of these stackings can accommodate two water molecules per formula unit. There is likely to be a compositional discontinuity and phase transition between the b/3-displaced phase and the phase with phlogopite-like stacking. The simulations reported here are the first to use the recently developed CLAYFF force field to calculate mineral structures at elevated pressures and temperatures.