Volume 96, Issue 10

An experimental apparatus was built to measure P-wave velocity (v P ) of sandy sediment during hydrate formation from brine and free methane gas. The influences of hydrate saturation, initial brine saturation, conversion ratio of water to hydrate, and grain size of sand upon v P were investigated. The experimental results demonstrate that v P strongly depends on both hydrate saturation and initial brine saturation, whereas the influence of the grain size of sand is unremarkable. During the formation of hydrate for different experimental runs at identical initial brine saturation, v P increases with the increase of hydrate saturation; while it decreases dramatically with the increase of initial brine saturation for a given hydrate saturation. The correlation between v P and conversion ratio of water into hydrate was studied, and it was found that the relation between v P and the conversion ratio of water to hydrate is approximately the same for different experimental runs, regardless of how big the differences in initial saturation of brine and the size of sand grain are. v P was calculated for four different hydrate distribution models. The results suggest that low conversion ratio of water to hydrate, hydrate mainly acts as a load component of a dry frame matrix, whereas at higher conversion ratio of water to hydrate it partly acts as a cement that coats grains at grain contacts, and thereby leads to a dramatic increase in v P . Compared with the hydrate saturation, the conversion ratio of water to hydrate has a large role in distributing hydrate in the pores of sediments.

Eight turquoise samples covering a wide range of compositions in the turquoise-chalcosiderite solid-solution series were analyzed by Mössbauer spectroscopy, X‑ray diffraction (XRD), electron microprobe analysis (EMPA), and Fourier transform infrared (FTIR) spectroscopy. Two of the turquoise samples display evidence of alteration from weathering processes. The unit formulas were calculated on the basis of 24 (O,OH) anions and 11 cations using the results of EMPA, assuming all Fe as Fe 3+ , as confirmed by Mössbauer spectroscopy. The altered turquoise samples show deficiencies in both cations and anion groups, indicated by EMPA, but they preserve the crystal structure of turquoise, as verified by XRD. They also show large amounts of Si and Ca in their microprobe data, due to the presence of kaolinite and Ca carbonate, respectively, which are identified by FTIR spectroscopy. The isomorphous substitution of Fe 3+ for Al in the turquoise structure broadens and shifts the IR bands to lower frequencies, in particular the OH-stretching bands. The Mössbauer spectra, collected at room temperature, are fitted with two generalized Fe 3+ sites, using a Voigt-based quadrupole-splitting distribution method, which are assigned to the M3 (smaller quadrupole splitting) on the one hand and M1 and M2 octahedral sites on the other hand. The Fe 3+ distribution over the M3 and M1,2 sites, calculated from the Mössbauer relative areas and EMPA, indicates that Fe 3+ prefers the larger M3 octahedron in the turquoise-chalcosiderite solid-solution series.

Twenty sepiolites of known composition from different origins were analyzed to quantify the variability in structural parameters and its possible relationships with composition and morphology. Morphology SEM analyses led to classify the sepiolites into several groups, beyond the two simple macroscopic or clay groups. X‑ray powder diffraction with synchrotron light was used to discuss the variability of the a and b cell parameters with the nature of the cations and occupancy of the octahedral shell. Rietveld refinement using the ideal sepiolite model is performed on sepiolites at two temperatures: 225 °C (for zeolitically dehydrated sepiolite) and 25 °C (for hydrated ambient sepiolite). The latter permitted to locate ca. six molecules of the zeolitic H 2 O within the tunnels. A few samples were selected to evaluate the feasibility and potential of single-crystal diffraction methods: X‑ray microdiffraction and electron diffraction. The macroscopic sepiolites gave wellstructured and rich X‑ray fiber diffraction patterns, in excellent agreement with ab initio simulations. High-quality single-crystal electron diffraction patterns for three axis zones are indexed and compared with simulations. The experimental and modeling results for X‑ray microdiffraction and electron diffraction open a new path for quantitative crystallography on sepiolite and other fibrous clays from the sepiolite-palygorskite group.

A new occurrence of rankamaite is here described at the Urubu pegmatite, Itinga municipality, Minas Gerais, Brazil. The mineral forms cream-white botryoidal aggregates of acicular to fibrous crystals, intimately associated with simpsonite, thoreaulite, cassiterite, quartz, elbaite, albite, and muscovite. The average of six chemical analyses obtained by electron microprobe is (range in parentheses, wt%): Na 2 O 2.08 (1.95-2.13), K 2 O 2.61 (2.52-2.74), Al 2 O 3 1.96 (1.89-2.00), Fe 2 O 3 0.01 (0.00-0.03), TiO 2 0.02 (0.00-0.06), Ta 2 O 5 81.04 (79.12-85.18), Nb 2 O 5 9.49 (8.58-9.86), total 97.21 (95.95-101.50). The chemical formula derived from this analysis is (Na 1.55 K 1.28 ) Σ2.83 (Ta 8.45 Nb 1.64 Al 0.89 Fe 3+ 0.01 Ti 0.01 ) Σ11.00 [O 25.02 (OH) 5.98 ] Σ31.00 . Rankamaite is an orthorhombic “tungsten bronze” (OTB), crystallizing in the space group Cmmm. Its unit-cell parameters refined from X‑ray diffraction powder data are: a = 17.224(3), b = 17.687(3), c = 3.9361(7) Å, V = 1199.1(3) Å 3 , Z = 2. Rietveld refinement of the powder data was undertaken using the structure of LaTa 5 O 14 as a starting model for the rankamaite structure. The structural formula obtained with the Rietveld analyses is: (Na 2.21 K 1.26 ) Σ3.37 (Ta 9.12 Nb 1.30 Al 0.59 ) Σ11.00 [O 26.29 (OH) 4.71 ] Σ31.00 . The tantalum atoms are coordinated by six and seven oxygen atoms in the form of distorted TaO 6 octahedra and TaO 7 pentagonal bipyramids, respectively. Every pentagonal bipyramid shares edges with four octahedra, thus forming Ta 5 O 14 units. The potassium atom is in an 11-fold coordination, whereas one sodium atom is in a 10-fold and the other is in a 12-fold coordination. Raman and infrared spectroscopy were used to investigate the room-temperature spectra of rankamaite.

Although Brønsted-Lowry (B-L) sites in solids, such as zeolites, have been studied extensively, all previous investigations were conducted on a bulk (average) basis. In contrast, the imaging and distribution of B-L sites on atomic flat Al-rich chlorite are presented by using Scanning probe and Kelvin probe force microscopy. These techniques are used to correlate, at the nanoscale, the contrast due to the surface potential (related to the B-L proton) with the surface morphology and crystal chemistry. Quantum mechanical modeling (DFT) is consistent with the experimental results. Imaging of the distribution of B-L sites in solids and the existence of two-dimensional (2D) arrays of zeolitic-type B-L sites in chlorites is shown. The study demonstrates the general validity of the Brønsted-Lowry acid-base theory extended to pure solids without any solution medium. The experimental approach developed here can facilitate the search of B-L site architectures in minerals.

Iron oxide-alloy equilibration experiments were conducted in H 2 -CO 2 gas mixtures at 1 atm and 1125-1240 °C using strips of Au 74 Pd 26 (wt%) and produced Au-Pd-Fe alloys with 0.03-13 wt% iron. A thermodynamic calibration for the mixing of Au 74 Pd 26 with iron using an asymmetric regular solution leads to W G-Fe = -45.0 ± 1.8 kJ/mol and W G-AuPd = +19.5 ± 7.7 kJ/mol (1σ). Internal oxidation of iron was observed in a reversal experiment, suggesting that oxygen can be transferred across capsule boundaries during high-temperature experiments. This thermodynamic calibration is applicable to a wide range of oxygen fugacities and iron activities relevant to petrological and metallurgical applications at 1 atm and, as previous studies suggest excess volumes in this system are small, it can also be used to predict Fe activities in experiments at elevated pressure (up to 3 GPa). By pre-doping Au-Pd capsules to match Fe activities expected for the sample during an experiment, it is possible to maintain samples with little to no loss of iron. Pre-saturation of the capsule also provides a method for controlling the oxygen fugacity of samples if no formal oxygen buffer is available.

We investigate charge-balanced hydrous magnesium and silicon defects [(2H) X Mg , (4H) X Si ] by first principles. Two new lowest-energy hydrogen configurations are proposed for (4H) X Si . With these new configurations, the distribution of O-H stretching phonon frequencies in Group I (>3450 cm -1 ) are better reproduced. Substitution of silicon with four hydrogen atoms gives rise to significant elongation of distances between O atoms at the tetrahedron of the silicon vacancy. Our calculations indicate that the correlation between O-O distances and O-H stretching phonon frequencies, which has been well established for hydrous minerals, does not apply directly to nominally anhydrous minerals and should not be used to determine the identity of hydrous defects responsible for infrared absorption peaks.

A combined Raman spectroscopy and electron probe microanalysis study of the heavily shocked Grove Mountains (GRV) 052049 meteorite revealed the largest chemical fractionation of natural ringwoodite, and composition-dependent variation of the intensities and/or wavenumbers of Raman bands. With Fa content [atomic ratio of Fe/(Fe+Mg)] of ringwoodite varying from 27.8 to 81.6 mol%, the peak position of the single band around 290 cm -1 (SB1), which relates to the SiO 4 translation mode, shifts from 296.0 to 284.6 cm -1 , and one of the doublets around 790 cm -1 (DB1), which relates to the symmetric stretching of SiO 4 , shifts from 796.3 to 782.7 cm -1 . In addition, the relative intensities of SB1 and the other band of the doublet around 840 cm -1 (DB2), which relates to asymmetric stretching of SiO 4 , increases with Fa content. Based on the paired Raman-EPMA data, single-peak and two-peak calibrations were established, which can be used to derive Fa contents of ringwoodite from the Raman spectra. The accuracy of Raman-derived Fa content of ringwoodite is better than ±5 mol%. The correlation of SB1 intensity with the Fa content of ringwoodite suggests that the vibration of SB1 is enhanced with the substitution of Mg 2+ by Fe 2+ . The correlation between Raman spectra and the chemical composition of ringwoodite have potential applications in on-line measurement of high-pressure experiments and in situ mineralogical determination in future planetary explorations.

An in situ synchrotron powder X‑ray diffraction study on (Fe 0.89 ,Ni 0.11 ) 3 S was conducted up to 141 GPa and 1590 K. (Fe 0.89 ,Ni 0.11 ) 3 S has a tetragonal structure, which is the same structure as Ni-free Fe 3 S. Fitting a third-order Birch-Murnaghan equation of state to data at ambient temperature yielded a bulk modulus of K 0 = 138.1(7.2) GPa and its pressure derivative K 0 ′ = 4.5(3) with a zero pressure volume V 0 = 375.67(4) Å 3 . The density of (Fe 0.89 ,Ni 0.11 ) 3 S under the core-mantle boundary condition is 1.7% greater than that of Fe 3 S. The axial ratio (c/a) of (Fe 0.89 ,Ni 0.11 ) 3 S decreases with increasing pressure. The addition of nickel to Fe 3 S leads to a softening of the c-axis. Assuming that the nickel content of the outer core is about 5 at%, we estimated 12.3-20.8 at% sulfur in the outer core for the given 6-10% density deficit between the outer core and pure iron at 136 GPa.

The structure of quartz over the temperature domain (298 K, 1078 K) and pressure domain (0 GPa, 20.25 GPa) is compared to the following three hypothetical quartz crystals: (1) Ideal α-quartz with perfectly regular tetrahedra and the same volume and Si-O-Si angle as its observed equivalent (ideal β-quartz has Si-O-Si angle fixed at 155.6°). (2) Model α-quartz with the same Si-O-Si angle and cell parameters as its observed equivalent, derived from ideal by altering the axial ratio. (3) BCC quartz with a perfectly body-centered cubic arrangement of oxygen anions and the same volume as its observed equivalent. Comparison of experimental data recorded in the literature for quartz with these hypothetical crystal structures shows that quartz becomes more ideal as temperature increases, more BCC as pressure increases, and that model quartz is a very good representation of observed quartz under all conditions. This is consistent with the hypothesis that quartz compresses through Si-O-Si angle-bending, which is resisted by anion-anion repulsion resulting in increasing distortion of the c/a axial ratio from ideal as temperature decreases and/or pressure increases.

Neutralization of acidic fluids by means of fluid-olivine interactions is important in volcanic environments and has been proposed as a practical scheme for the neutralization of acidic sulfaterich fluids. To understand the interaction of olivine with highly acidic fluids we have reacted whole olivine crystals and a dunite cube with different sulfuric acid solutions at temperatures ranging from 60-120 °C. Reaction of olivine with 2 and 3.6 M acid concentrations produced a layered amorphous silica pseudomorph of the original olivine grain. The mechanism of pseudomorphic replacement was studied by reacting olivine with an 18 O-enriched acid solution and examining the products using Raman spectroscopy. Peak shifts in the Raman spectra show that 18 O was incorporated into the silica rim, including the siloxane ring structures. The formation of a layered silica pseudomorph, the incorporation of 18 O into the silica rim and the dependence of the replacement rim strength on the acid concentration indicate that the pseudomorphic replacement occurred by means of an interface-coupled dissolution-reprecipitation mechanism. When olivine was reacted with 1 M sulfuric acid amorphous silica was produced but no longer formed a pseudomorph of the olivine grain. Reaction with 0.1 M acid, or solutions containing Na, encouraged the formation of hematite as well as amorphous silica. From the known Fe-phase stabilities for our experimental conditions and the dependence of hematite formation on the presence of Na we propose that initially jarosite phases precipitated, which transformed into hematite during the experiment.

Under pressure and temperature conditions like those found in deep geological repositories (DGRs), rare-earth cations may react with silicates to form rare-earth disilicates. This study establishes the stability range of yttrium disilicates in response to changes in pH. The α, β, γ, and δ polymorphs of Y 2 Si 2 O 7 were synthesized by the sol-gel process at temperatures between 1100 and 1650 °C and subjected to pH stat leaching tests. By measuring the Y and Si leaching rates and monitoring the transformation of the crystalline and amorphous phases, we showed that yttrium disilicates were stable at pH > 5. At pH < 5, the pH stability sequence was consistent with the temperature-dependent stabilities of Y 2 Si 2 O 7 phases, with the δ polymorph showing the lowest leaching rates. Because rare-earth compounds can be used as a proxy for analogous actinide hosts, the results of this study can be used to predict the performance of engineered barriers in DGR.

Atomic diffusion along grain boundaries in solids is a key process in many geological environments and in ceramics research. It is closely related to the grain boundary width, which is an important parameter in numerous equations describing diffusional or rheological processes, including plastic deformation of polycrystals, intergranular failure, and recrystallization. Here, we studied diffusion along a single well-characterized, near Σ5 grain boundary in yttrium aluminum garnet (YAG) using different transmission electron microscopy methods at atomic resolution. For the diffusion experiment, YAG thin-films containing Yb as diffusant were deposited perpendicular to the grain boundary on the bicrystal. We investigated the grain boundary using a focal series in combination with multislice calculations that yield the electron exit wave. This, coupled with chemically sensitive Z-contrast images, as well as the Yb distribution over the grain boundary measured using electron energy loss spectroscopy, show the zone of enhanced Yb diffusion parallel to the grain boundary. This zone of enhanced diffusion is often considered as the effective grain boundary width. Profiles from the boundary into the crystal volume suggest a highly permeable zone of about 18 nm, which we assume to be the effective grain boundary width for diffusional processes. The effective width strongly differs from the structural grain boundary width of about 2 nm in the present study. Furthermore, it is much shorter compared to the calculated volume diffusion profile. We conclude that the combination of small samples and transmission electron microscopy at atomic resolution are excellent tools to study varying processes, such as diffusion, deformation, or reactions at the atomic scale.

Using static electronic structure calculations, we determine the equation of state, the full elastic constant tensor and the sound wave velocities of cementite (Fe 3 C) at pressures up to 410 GPa. Fe 3 C is ferromagnetic (fm) at ambient pressures. Upon compression, the magnetic moment of the Fe atoms are gradually lost and, at around -62 GPa, Fe 3 C becomes non-magnetic (nm). We find that the pressurevolume results for the Fe 3 C (fm) phase are well represented by a Vinet equation of state with K 0 fm = 183 GPa, K 0 = 5.9, and V 0 fm = 151.6 Å 3 and that of the Fe 3 C (nm) phase are well represented by a Vinet equation of state with K 0 nm = 297 GPa, K 0 ′ = 4.9, and V 0 nm = 143.2 Å 3 . A third-order Birch-Murnaghan equation of state formulation for the Fe 3 C (nm) phase yields similar parameters with K 0 nm = 304 GPa, K0 = 4.5, and V 0 nm = 143.3 Å 3 . At pressures relevant to the Earth’s inner core, the full elastic constant tensor of Fe 3 C (nm) reveals significant P-wave anisotropy (-10%). A crystal preferred orientation with the [110] directions of Fe 3 C aligned along the pole axis would be required to explain the inner core anisotropy. Comparing, pure hcp Fe and iron carbides with varying stoichiometry, we find that the shear wave velocity decreases linearly with the increasing C content.

A systematic study of the Raman spectra of the titanomagnetite solid-solution series (Fe 3-x Ti x O 4 ) for x = ~0.0, 0.2, 0.4, and 0.6 has been conducted. The samples showed combinations of five previously predicted Raman peaks at ~190, 310, 460, 540, and 670 cm -1 that correspond to vibrational modes with T 2g (1), E g , T 2g (3), T 2g (2), and A 1g , respectively. The calibration of Raman spectra for titanomagnetite with known values of Ti concentrations reveals a strong dependence of relative intensity for the T 2g (2) and T 2g (3) modes on Ti concentration. The most prominent feature is the appearance and increase in the relative intensity of a T 2g (3) peak above x = ~0.2. On the other hand, the Raman peak for the T 2g (2) mode gradually diminishes as Ti increases and nearly disappears at x = ~0.6. Combining the two relative intensities potentially provides a sensitive indicator of Ti content. The technique was applied to study titanomagnetite in grains from Hana Volcanics and melatroctolite from Rhode Island.

The experimental alteration of monazite to allanite, REE-epidote, fluorapatite, and/or fluorapatitebritholite was investigated at 450 to 610 MPa and 450 to 500 °C. Experiments involved monazite + albite ± K-feldspar + muscovite ± biotite + SiO 2 + CaF 2 and variety of fluids including H 2 O, (KCl + H 2 O), (NaCl + H 2 O), (CaCl 2 + H 2 O), (Na 2 Si 2 O 5 + H 2 O), 1 M HCl, 2 M NaOH, 2 M KOH, 1 M Ca(OH) 2 , 2 M Ca(OH) 2 , and (CaCO 3 + H 2 O). The reaction products, or lack thereof, clearly show that the stability relations between monazite, fluorapatite, and allanite or REE-epidote are more dependent on the fluid composition and the ratio of silicate minerals than on the P-T conditions. A high Ca content in the fluid promotes monazite dissolution and the formation of fluorapatite and allanite or REE-epidote. Lowering the Ca content and raising the Na content in the fluid decreases the solubility of monazite but promotes the formation of allanite. Replacing Na with K in the same fluid causes fluorapatite, with a britholite component, to form from the monazite. However, allanite and REE-epidote are not formed. Monazite is stable in the presence of NaCl brines. In KCl brine, monazite shows a very limited reaction to fluorapatite. When the fluid is (Na 2 Si 2 O 5 + H 2 O), strong dissolution of monazite occurs resulting in the mobilization of REEs, and actinides to form fluorapatite-britholite and turkestanite. These experimental results are consistent with natural observations of the partial to total replacement of monazite by fluorapatite, REE-epidote, and allanite in fluid-aided reactions involving the anorthite component in plagioclase at mid- to high-grade metamorphic conditions. In contrast, an alkali-bearing environment with excess Na prevents the growth of allanite and eventually promotes the precipitation of secondary monazite. The results from this study provide implications for geochronology and for deducing fluid compositions in metamorphic rocks.

Quantum-mechanical calculations with a hybrid HF/DFT Hamiltonian (B3LYP) model yielded the Raman-active vibrational modes of the Mg- and Fe-cordierite structure. Maximum and mean deviation between experimentally derived bands and calculated modes of synthetic Mg- and natural Fe-rich cordierite are ±19 and 7 cm -1 . Most of the observed bands could be related to specific vibrational modes of tetrahedral and octahedral sites of the cordierite structure, although the large number of Ramanactive modes (87) prevents a complete assignment. Atomic motions in cordierite are compared with those of the structurally similar mineral beryl. The calculations enable more accurate interpretation of the Raman spectra with respect to structural changes of cordierite, in particular Al-Si ordering and Mg-Fe exchange.

Alkali feldspar crystals from the Klokken syenite pluton (South Greenland) have a coarse patch perthite exsolution texture, on the scale of hundreds of micrometers, produced by dissolution-reprecipitation reactions at ~500 °C. We have discovered that Ab-rich patches (Ab 92 An 7 Or 1 ) subsequently exsolved by spinodal decomposition to give peristerite, with a periodicity of 17 nm. This is the first time a peristerite has been characterized by TEM within a perthite, and the first peristerite to be found outside regional metamorphic rocks and pegmatites. The coexisting Or-rich patches, Ab 16 Or 84 (An 0.3 ), unmixed over the same temperature interval to yield coherent film microperthite with periodicities ≤1 μm. Because the peristerite and film perthite from a single hand specimen from the Klokken intrusion have shared exactly the same cooling history, they provide a novel opportunity to test the utility of diffusion coefficients obtained in the laboratory for prediction of coarsening rates over 105 yr time scales, in the light of a calculated cooling trajectory for the pluton. We find good agreement between observed and predicted periodicities for the film perthite, which began to exsolve by coherent nucleation at the coherent solvus for ordered feldspars at ~420 °C, ~32 000 yr after crystal growth, and took a further 13 000-39 000 yr to coarsen to a periodicity of 1 μm. Peristerite exsolution began at the conditional spinodal at ~450 °C, but on the basis of the best existing experimental data would have taken ~520 000-4 600 000 yr to coarsen, incompatible with the cooling history. We speculate on changes to these “best” parameter values to describe NaSi-CaAl interdiffusion that would lead to peristerite coarsening rates compatible with the cooling history of the Klokken intrusion. We find that an interdiffusion expression: D NaSi-CaAl = (9 × 10−12)exp(-220 000/RT) (where D and the pre-exponential term, A, are in units of m2/s and the activation energy, Ea, is in J/mol) leads to coarsening rates compatible with the experimental data, the cooling path of the Klokken intrusion, and the coarsening rate of the coexisting film perthite. The new interdiffusion parameters are consistent with fine-scale peristerite intergrowths in low-grade metamorphic rocks and with the existence of the coarsest known peristerite. In the light of our discovery, it seems likely that peristerite is commonplace in albitic plagioclase in granitic rocks.

The site-specific behavior of iron in an orthoenstatite-structured 57 Fe-enriched (M1)(M 2 )Si 2 O 6 powdered sample was explored using synchrotron Mössbauer spectroscopy and diamond-anvil cells in two independent experiments at ambient temperature. In one experiment, NaCl was used as the pressure- transmitting medium (ambient pressure to 36 GPa), and in the other experiment, Ne surrounded the sample (4.1 to 26.8 GPa). The hyperfine parameters of the M1 and M2 sites at room pressure are in excellent agreement with previous literature values obtained using conventional Mössbauer spectroscopy and yield (Mg 0.980 Fe 0.020(5) )(Mg 0.760 Fe 0.240 )Si 2 O 6 as the chemical formula. Analyses of both data sets reveal a change in the trend or discontinuity in the hyperfine parameters around 10 GPa, indicative of a transformation in orthopyroxene. However, the detailed behaviors of the iron sites with pressure appear to depend on the local stress conditions provided by the different pressure media. Our observations may help explain the reported variations in structural transition behavior for orthopyroxenes at high pressures.

Low- and high-temperature single-crystal X‑ray diffraction studies of synthetic pyrope garnet have been conducted at 20 temperature-points over a wide temperature range from 96.7 to 972.9 K. From precise structure refinements, the possibility of static disorder and anharmonic thermal vibration of Mg in the dodecahedral site, which has long been under debate, has been assessed together with the thermal expansion behavior. Application of the Debye model to the temperature dependence of the resulting mean square displacements (MSDs) of atoms shows that they clearly include significant static disorder components in all the constituent atoms, of which Mg has the most dominant static disorder component. The residual electron density analyses and the structure refinements based on the split-atom model at a low temperature of 96.7 K provide direct proof of the Mg static disorder. Anharmonic structure refinements applying the Gram-Charlier expansion up to the fourth-rank tensor show that the anharmonic contribution to atomic thermal vibrations begins to appear in all atoms except Si at a high temperature of T > 800 K. However, at lower temperatures, anharmonic refinements show no sign of anharmonic thermal vibrations on any atom and instead provide an indication of their static disorder.

The single-crystal elastic properties of Fe-bearing wadsleyite with 1.93 wt% H 2 O (Mg 1.634 Fe 0.202 H 0.305 SiO 4 ) have been determined by Brillouin scattering. At ambient conditions, the aggregate bulk and shear moduli (K S0 , G 0 ) of this wadsleyite are 156.2(5) and 98.0(3) GPa, respectively. Compared to the corresponding anhydrous wadsleyite, 1.93 wt% H 2 O lowers K S0 and G 0 by 8.1% and 9.3%, respectively. High-pressure measurements up to 12 GPa show that the pressure derivative of the bulk modulus, K′ S0 = 4.8(1), is similar to that of the anhydrous Fe-wadsleyite with reported values of 4.6-4.74, but the addition of H 2 O increases the pressure derivative of the shear modulus, G 0 ′ from 1.5(1) to 1.9(1). This contrasts with the G 0 ′ of Fe-free wadsleyite, which is the same within uncertainty for the hydrous and anhydrous phases. As a result, both the compressional- and shear-wave velocities (v P , v S ) of hydrous Fe-bearing wadsleyite are about 200(±24) m/s slower than anhydrous Fe-bearing wadsleyite at transition zone pressures.

We have examined a paleosol in Pronto, Canada, which formed at ~2.45 Ga and under very low O 2 conditions, to reconstruct ancient weathering processes. Chlorite, which replaced biotite in parent granite, consists of almost complete chlorite layers as revealed by high-resolution transmission microcopy. Chlorite grains were fragmented gradually with increased intensity of weathering. The fragmented chlorite grains were partly or largely replaced by sericite formed from K-metasomatism, which strongly suggests that almost complete chloritization of biotite preceded sericitization. In the upper part of the paleosol, chlorite-biotite mixed layers formed but the formation of these layers cannot be explained by mere chloritization and sericitization, which are usually considered to occur concurrently during post-weathering events. Our observations and analyses suggest that the chlorite-biotite mixed layers were formed as follows: chlorite, with no or little biotite, was formed from biotite in parent granite prior to weathering. Chlorite was then weathered by dissolution and fragmentation at ~2.45 Ga, and chlorite-vermiculite (or smectite) mixed layers were formed by layer-by-layer transformation in the weathering profile (note that vermiculite is not actually observed in the Pronto paleosol). Following burial of the weathering profile, the vermiculite layers in the chlorite-vermiculite mixed layers were converted to biotite layers, by means of K-metasomatism, to form the chlorite-biotite mixed layers. The chlorite-biotite mixed layers in the Pronto paleosol were therefore chlorite-vermiculite mixed layers at the time of weathering.

Eliseevite, Na 1.5 Li[Ti 2 Si 4 O 12.5 (OH) 1.5 ]∙2H 2 O, is a new microporous titanosilicate of the lintisitekukisvumite family [monoclinic, C2/c, a = 27.48(1), b = 8.669(4), c = 5.246(2) Å, β = 90.782(8)°, V = 1249.7(9) Å 3 , Z = 4]. The mineral is found in two different peralkaline veins in an ijolite-foyaite- malignite differentiated complex of the Lovozero alkaline massif, Kola Peninsula, Russia. At Mt. Alluaiv, eliseevite occurs in an aegirine-eudialyte-sodalite-microcline vein as long-prismatic to fibrous crystals (up to 2 mm long) growing in voids of natrolitized sodalite in close association with albite, analcime, catapleiite, chabazite-Ca, gmelinite-K, manganoneptunite, microcline, murmanite, and an ussingite. At Mt. Punkaruaiv, it is found within a ussingite-aegirine-microcline vein as longprismatic crystals (up to 0.8 mm long) in close association with chabazite-Ca, chkalovite, eudialyte, manganoneptunite, punkaruaivite, rhabdophane-(Ce), sodalite, sphalerite, and steenstrupine-(Ce). It is a late-stage, hydrothermal mineral formed as a result of alteration of murmanite. The mineral is transparent, pale creamy to colorless, with a vitreous luster and a white streak. Cleavage is perfect along {100}, fracture is splintery. Mohs hardness is about 5. In transmitted light, the mineral is colorless, biaxial (-): α = 1.665(2), β = 1.712(2), γ = 1.762(5) (for λ = 589 nm); Y = b, Z^a = 8-12°. Dispersion is medium, r < v. D calc = 2.706 g/cm 3 , D meas = 2.68(4) g/cm 3 . The mean chemical composition (n = 7) determined by the Penfield method (water), ICP-MS (Li), and electron microprobe (other elements) is (wt%): H 2 O 10.50, Li 2 O 2.85, Na 2 O 9.15, K 2 O 0.08, CaO 0.05, Fe 2 O 3 0.21, Al 2 O 3 0.08, SiO 2 46.87, TiO 2 29.40, Nb 2 O 5 0.72, total 99.91. The empirical formula calculated on the basis of Si = 4 apfu is: (Na 1.51 K 0.01 Ca 0.01 ) Σ1.53 Li 0.98 [(Ti 1.89 Nb 0.03 Fe 3+ 0.01 Al 0.01 ) Σ1.94 Si 4 O 12.26 (OH) 1.74 ]∙2.12H 2 O. The simplified formula taking into account the results of a single-crystal study is Na 1.5 Li{Ti 2 O 2 [Si 4 O 10.5 (OH) 1.5] }∙2H 2 O. The six strongest reflections in the X‑ray powder-diffraction pattern [d in Å, (I), (hkl)] are: 13.76(100) (200), 6.296(60)(310), 3.577(80)(710), 3.005(70)(421), 2.881(70)(910), 2.710(50)(621). The mineral is named in honor of Nikolai Aleksandrovich Eliseev (1897-1966), a remarkable Russian geologist and petrologist, Professor at Leningrad State University, for his contributions to the geology and petrology of metamorphic and alkaline complexes.

Knowledge of the deformation mechanism of post-perovskite is important for interpreting observed seismic anisotropy in terms of mantle flow. Experiments on post-perovskite MgSiO 3 and the low-pressure analog material CaIrO 3 yield different textures, leaving the interpretation of the observed seismic signatures unclear. Here we present results of deformation experiments on CaPtO 3 post-perovskite that may be a better analog to MgSiO 3 . Post-perovskite CaPtO 3 deforms by glide of [100] dislocations on the (010) plane, consistent with previous experimental results on CaIrO 3 . In addition, samples containing a weak minority phase also display shape-preferred orientation with grains elongated in the crystallographic a-direction forming a planar fabric perpendicular to the compression direction. This shape-preferred orientation strengthens the observed crystallographic-preferred orientation and results in a rapid development of texture during deformation. This observation supports the recent suggestion that the D″ reflector might be due to a rapid generation of texture in post-perovskite. Furthermore, the role of shape-preferred orientation in generating seismic anisotropy in multi-phase assemblages should be considered for the D″ assemblage.

The structure of phosphorus-bearing, H 2 O-saturated silicate melts, silicate-saturated aqueous fluids, and silicate-rich single phase (supercritical) liquids has been characterized in situ to 800 °C and 1486 MPa in an Ir-gasketed hydrothermal diamond-anvil cell (HDAC) with the aid of both confocal microRaman and FTIR spectroscopy. Temperature and pressure in the HDAC were recorded with thermocouples (±1 °C uncertainty) and pressure- and temperature-dependent Raman shift of 13 C diamonds (±40 MPa uncertainty). Starting materials were aluminum-free Na 2 O·4SiO 2 (NS4) and with 10 mol% Al 2 O 3 (NA10) substituting for SiO 2 , both with 5 mol% P 2 O 5 . Aluminosilicate species of Q 0 , Q 1 , Q 2 , and Q 3 type exist in coexisting fluid and melt and in single phase liquid together with phosphate species, PO 4 , P 2 O 7 , and Q n P. In the Q n P species, n O atoms bridge between Si 4+ and P 5+ , whereas there is no Si-to-P linkage in the PO 4 and P 2 O 7 species. In melts, the abundance of the most depolymerized silicate species, Q 0 , is positively correlated with temperature and pressure, whereas that of the most polymerized species, Q 3 , decreases with temperature and pressure. In the silicate solute of aqueous fluids, the opposite relationship exists with Q 0 abundance decreasing and Q 3 (and Q 1 and Q 2 ) abundance increasing with increasing temperature and pressure. The silicate melts, therefore, become increasingly depolymerized and the silicate solute in aqueous fluids decreasingly depolymerized because increasing H 2 O solubility in melts causes melt depolymerization, whereas increasing silicate solute abundance in fluids results in silicate polymerization. The P 2 O 7 and Q n P are the dominant phosphate species in fluid, melt, and single phase liquid with orthophosphate, PO 4 , playing a subordinate role. The fluid/melt partition coefficients for P 2 O 7 and Q n P species are in the 0.15-0.7 range with that of Q n P being greater than that of P 2 O 7 . The PO 4 fluid/melt partition coefficients are <0.2. In all cases, the partition coefficients increase with increasing temperature and pressure. There is no clear influence of Al 2 O 3 on partition coefficients. Hence, it appears that P-bearing complexes in fluids and melts are associated with Na + that becomes available as silicate species polymerize. The mobility of phosphorus during metamorphic processes is principally governed by the availability of alkali metals (and perhaps alkaline earths).

Recent SIMS analysis of water, F, and Cl in lunar apatite suggests significantly higher volatile abundances in lunar magmas than previously considered. However, apatite is commonly a latecrystallizing mineral and its volatile content may reflect late-stage open-system processes that have perturbed the magmatic volatile content and obscured direct information regarding the volatile contents of the parental magmas and magmatic source region. Degassing during magma ascent has the potential to perturb not only the absolute but also the relative magmatic volatile abundances. A set of evacuated silica tube degassing experiments were conducted that simulate ascent of high-Al basalt 14053 with 0.5 wt% Cl, 0.5 wt% F, 0.3 wt% S, and 2.2 wt% (and 2.5 wt%) water (in addition to dissolved C-O-H species) from 100 km to within ~20 m of the surface followed by degassing (at an f O₂ of ~QIF). Extensive degassing occurred within 6 h during which 99-100% of the initial water, 89-84% of the initial Cl, 60-61% of the initial F, and 94-92% of the initial S was lost. During degassing, the relative volatile contents showed a strong decrease in water content and an increase in F:Cl ratio. In reflection of the changes in melt volatile contents, apatites crystallizing from the degassed melt would have much lower OH contents and higher F:Cl ratio than apatites crystallized from the non-degassed melt. These results confirm the possibility of significant underestimation of primary magma volatile contents, especially water and Cl, through use of apatite volatile contents and the assumption of simple increases in volatile abundance during magma differentiation.