Volume 91, Issue 2-3

Elastic anomalies that accompany cation ordering and displacive phase transitions can be understood in terms of coupling between strain components and the driving order parameter in Landau free energy expansions. Non-convergent cation ordering in spinel, MgAl 2 O 4 , is accompanied by changes in individual elastic constants, shear modulus, and bulk modulus that vary linearly with the order parameter. Convergent cation ordering, such as Al/Si ordering in anorthite, is expected to give changes in elastic properties that scale with the square of the order parameter. The elastic anomalies that develop in association with displacive phase transitions show greater diversity, due to the additional influence of the order parameter susceptibility. These are illustrated for the cases of the proper ferroelastic transition at high pressure in stishovite and the improper ferroelastic transition in SrTiO 3 perovskite. Low temperature transitions in lawsonite show a more complex pattern of softening and stiffening that depends on coupling with both cation ordering and displacive processes. Variations of the spontaneous strain and elastic constants are indicative of the underlying thermodynamic mechanism for a phase transition. If any such transitions occur in minerals of the Earth's crust or mantle they should be identifiable from their distinctive influence on seismic velocities.

The tetranatrolite-paranatrolite transformation has remained a key problem in understanding the paragenesis of zeolites in the natrolite family. It is accepted that when paranatrolite, approximate formula Na 16-x Ca x Al 16+x Si 24-x O 80 · 24 H 2 O, is removed from an aqueous environment and exposed to the atmosphere, it loses water and transforms to tetranatrolite, Na 16-x Ca x Al 16+x Si 24-x O 80 ·nH 2 O (n ≤ 24). Here we show that this transformation is not only reversible, but that tetranatrolite exhibits two sequential pressure-induced hydrations leading first to paranatrolite and then to a superhydrated tetranatrolite above 0.2 and 3.0 Gpa, respectively. We have previously reported similar behavior for the corresponding system with an ordered Si/Al distribution, i.e., natrolite itself, however the ordered version of paranatrolite exists over a much smaller pressure range. The pressure-induced transformations of natrolite and tetranatrolite thus further supports the supposition that paranatrolite is a distinct mineral species, with a pressure-stability field dependent upon composition.

To improve kinetic models for apatite fission-track annealing, we present new experimental annealing data that complement previously published data. To determine the degree of annealing of induced tracks, surface density (ρ), and mean horizontal-confined track lengths (l), were measured, both for basal and randomly oriented faces. Our annealing data were obtained by submitting an apatite sample collected in Itambé, Bahia, Brazil, to 46 different isothermal treatments where temperature ranged from 150 to 600 °C (duration of 1, 10, 100, and 1000 h). To compare the behavior of Itambé to Durango apatite, the latter was also annealed for 1 h in 9 isothermal experiments at temperatures between 240 and 380 °C. Our results show that the l/l 0 values in Durango are systematically smaller than those in Itambé sample, both in basal and random faces. The curves depicting relative track density reduction, ρ/ρ 0 , and relative mean confined track length reduction, l/l 0 , as a function of time and temperature, are similar for ρ/ρ 0 > ~0.5, but different for ρ/ρ 0 < ~0.5. In this interval, ρ/ρ 0 can be measured but the measurement of l/l 0 is very difficult because the confined tracks become undetectable. Measurements of ρ/ρ 0 and l/l 0 for tracks revealed in basal surfaces are systematically lower (but this difference is <3%) than those in randomly oriented ones.

Alpersite, Mg 0.58 Cu 0.37 Zn 0.02 Mn 0.02 Fe 0.01 SO 4 ·7H 2 O, a new mineral species with direct relevance to reactions in mine waste, occurs in a mineralogically zoned assemblage in sheltered areas at the abandoned Big Mike mine in central Nevada at a relative humidity of 65% and T = 4 °C. Blue alpersite, which is isostructural with melanterite (FeSO 4 ·7H 2 O), is overlain by a light blue to white layer dominated by pickeringite, alunogen, and epsomite. X-ray diffraction data (MoKα radiation) from a single crystal of alpersite were refined in P2 1 /c, resulting in wR = 0.05 and cell dimensions α = 14.166(4), b = 6.534(2), c = 10.838(3) Å, β = 105.922(6)°, Z = 4. Site-occupancy refinement, constrained to be consistent with the compositional data, showed Mg to occupy the M1 site and Cu the M2 site. The octahedral distortion of M2 is consistent with 72% Cu occupancy when compared with the site-distortion data of substituted melanterite. Cuprian pentahydrite, with the formula (Mg 0.49 Cu 0.41 Mn 0.08 Zn 0.02 )SO 4 ·5H 2 O, was collected from an efflorescent rim on a depression that had held water in a large waste-rock area near Miami, Arizona. After dissolution of the efflorescence in de-ionized water, and evaporation of the supernatant liquid, alpersite precipitated and quickly dehydrated to cuprian pentahydrite. These observations are consistent with previous experimental studies of the system MgSO 4 -CuSO 4 -H 2 O. It is suspected that alpersite and cuprian pentahydrite are widespread in mine wastes that contain Cu-bearing sulfides, but in which solubilized Fe 2+ is not available for melanterite crystallization because of oxidation to Fe 3+ in surface waters of near-neutral pH. Alpersite has likely been overlooked in the past because of the close similarity of its physical properties to those of melanterite and chalcanthite. Alpersite is named after Charles N. Alpers, geochemist with the United States Geological Survey, who has made significant contributions to our understanding of the mineralogical controls of mine-water geochemistry.

The redox states of volcanic and impact melts from the Earth, Moon, and Mars have been estimated from the valence state of V in basaltic glasses (Sutton et al. 2005). The V valence has been determined using synchrotron micro X-ray absorption near-edge structure spectroscopy (XANES) (Sutton et al. 2005), which allows for in situ measurements on samples with a micrometer spatial resolution and ~100 ppm elemental sensitivity. Here, we interpret those results for the natural samples and compare them to the literature. The results show that terrestrial melts are dominated by V 4+ , lunar samples by V 3+ , with Martian melts a mixture of both V 3+ and V 4+ . The f O ₂ estimates derived from the V valence are consistent with those determined by other proven methods, whereby terrestrial basalts experience f O₂ conditions within 1 or 2 log units of the QFM buffer, lunar basalts equilibrate at 1 to 2 log units below the IW buffer, and Martian basalts fall somewhere between the QFM and IW buffer. The results illustrate the usefulness of this technique; i.e., a robust oxybarometer covering over six orders of magnitude, applicable to samples that record f O₂ conditions from reduced extraterrestrial bodies to the oxidized Earth.

Use of infrared spectroscopy as an accurate, quantitative method to measure concentrations of hydrous species in minerals requires consideration of the interactions of anisotropic crystals with infrared light. Ensuring that contributions are identified from species at all orientations in the crystal requires combining three measurements, taken with the electric field polarized along three mutually perpendicular directions. This is typically accomplished by determining the orientation of a crystal in advance, and then sectioning it perpendicular to its principal axes. In many instances, however, natural or experimental samples are not suitable for such handling. Here we demonstrate a method that instead uses at least three randomly sectioned grains, considered to be multiple samples of a homogeneous population. We explain the theory whereby: (1) the orientations of the polarization vectors of measurements taken on these grains are determined by comparison to oriented standards of the same mineral, and (2) the principal-axis spectra of the sample are synthesized from the randomly oriented spectra. By comparison to complementary electron backscatter diffraction (EBSD) data, we demonstrate that determination of orientations using the silicate overtone bands in Fourier-Transform infrared (FTIR) spectra is accurate and precise, with typical angular errors of 6°. We show that this precision is sufficient for the synthetic principal-axis spectra to be essentially indistinguishable from X-ray oriented standard spectra. We demonstrate the application of this technique to determining the OH concentrations in a population of hydrated olivine grains recovered from a high-pressure, high-temperature multi-anvil experiment.

We performed new experiments on incorporation of hydrogen in olivine at high pressures (2-12 GPa) and temperatures (1000-1300 °C). OH concentrations were calculated using the Bell et al. (2003) calibration applied to principal-axis infrared absorption spectra synthesized from polarized measurements on randomly oriented grains. Starting materials for the experiments included both fine-grained powders and larger single crystals. Hydrogen was incorporated during grain growth in the former case and by volume diffusion in the latter. The spectra of Fe-bearing olivines exhibit similar structure regardless of the starting material, and are dominated by bands in the wavenumber range from about 3500 to 3650 cm -1 . We do not observe bands at 3525 and 3573 cm -1 , which are predominant in many natural olivines as well as olivines annealed in experiments at lower pressures, and are attributed to humite-related defects. Furthermore, bands between 3300 and 3400 cm -1 , attributed to high silica activity or high oxygen fugacity, are weak or non-existent. Our measurements indicate that OH solubility in Fe-bearing olivine is 2.5-4 times higher than that measured by Kohlstedt et al. (1996). Although this is largely due to the use of a new calibration in our study, correction of previous values is not straightforward. In the pure Mg-system, in contrast to Fe-bearing olivine, order-of-magnitude apparent differences in OH solubility can be obtained using different experimental procedures. This raises questions about attainment of equilibrium in experimental studies of hydrogen incorporation in nominally anhydrous minerals, particularly when crystals are grown from a hydrous melt.

We present a numerical model for calculating the temperature distribution inside resistance-heated high-pressure solid-medium axi-symmetric cell assemblies that incorporates both composition- and temperature-dependent thermal conductivity. The code was validated using both analytic solutions of simplified thermal diffusion problems and comparisons to actual laboratory experiments and was found to be reliable in matching the temperature characteristics of multi-anvil experiments. Calculations for various cell assembly designs resulted in temperature fields that are consistent with experimental measurements of thermal gradients. These calculations also illustrated the influence of temperature-dependence of thermal conductivity, an important and often-overlooked property, on the thermal profiles. This model may be used to fine-tune the design of cell assemblies, either to minimize thermal gradients or to produce a desired temperature distribution. The four "typical" multi-anvil cells that we used to demonstrate this technique have temperature profiles across the sample that range from 25 to 75 °C when the thermocouple temperature is 1200 °C. The thermocouple in all four is in a region where the temperature gradient is on the order of 100 °C per millimeter, which could lead to experimental temperature uncertainties that are correlated with the thermocouple location.

One natural and two synthetic spinels with compositions Mg(Al 1-z Fe z 3+ ) 2 O 4 (with 2z = 0.078, 0.138, and 0.200, respectively) were studied by in situ, high-temperature, single-crystal X-ray diffraction. All samples were heated from room temperature to 1050 °C. Thermal expansion was monitored by measuring the cell edge variation. Cation disorder was monitored by measuring the variation of the oxygen positional parameter u, which is closely correlated with the inversion parameter i. All samples showed extensive Al reordering at the M site between 550 and 650 °C, followed by an increase of disorder at T > 650 °C due to both Mg-Al and Mg-Fe 3+ intersite exchanges. The measured cation distributions were compared with those calculated using the general thermodynamic model for spinel binary solid-solutions of O'Neill and Navrotsky (1984). Measured and calculated inversion parameters compare satisfactorily at T > 650 °C, i.e., at conditions under which equilibrium was achieved at any temperature. In fact, at T > 650 °C, both IV Fe 3+ and IV Al increase with increasing T, following the equilibrium path. The reproducibility of IV Al occupancies was very high, whereas IV Fe 3+ occupancies were not satisfactorily matched. The cation distribution relaxation observed between 550 and 650 °C was interpreted on the basis of kinetic considerations. In this temperature range, inversion decreases to a minimum because the amounts of Al that reorder are far more abundant than those of Fe 3+ that disorder. The Mg-Fe 3+ exchange was confirmed to proceed at a faster rate than the Mg-Al exchange. Moreover, the Mg-Fe 3+ exchange was found to be active at laboratory times at about 550 °C, whereas the Mg-Al exchange was hard to monitor below 600 °C.

The kinetics of cation ordering (quench method) in two natural Mg(Al 2-y Cr y )O 4 spinels (y~ 0.03-0.06 and 0.24), highly ordered in terms of Mg-Al, were studied by means of X-ray single-crystal diffraction. The equilibrium distribution of Mg and Al between tetrahedral and octahedral sites was investigated at 650 °C (in disordering and ordering) and at 850 °C (in ordering), through several time-steps to monitor the rate of cation distribution before equilibrium was achieved. The cation distributions for both disordering and ordering experiments were obtained by measuring the oxygen positional parameter u, which is correlated to the inversion parameter x (Al in T site), and then to the composition of the samples. The Mueller kinetic model, satisfactorily applied to the experimental data, allowed the calculation of the kinetic ordering constants K, linearly related to temperature by means of Arrhenius equations. The kinetics of ordering processes are influenced by Cr content. The equilibrium for both the isotherms at 650 and 850 °C was reached at different elapsed times by the low- and high-Cr spinels: the time for the low-Cr sample was, in both the ordering experiments, about double that of the high- Cr sample. Consequently, the activation energy (186 and 175 kJ/mol for low- and high-Cr samples, respectively) for the intracrystalline Mg-Al ordering decreases with Cr increase.

The equations of state and structures of andalusite and sillimanite have been determined using high-pressure single-crystal X-ray diffraction. A third-order Birch-Murnaghan equation-of-state fit to 14 P-V data points measured between 1 bar and 9.8 GPa for andalusite yields values of K T0 = 144.2(7) GPa and K’ = 6.8(2). A similar analysis for sillimanite involving a fit to 13 P-V data points between 1 bar and 8.5 GPa results in K T0 = 164(1) GPa and K’ = 5.0(3). The axial compression of both structures is nonlinear and highly anisotropic (~60%) with the c-axis being the least compressible axis in both structures. The axial moduli determined with a parameterized form of the third-order Birch-Murnaghan equation of state are: K a ₀ = 163(1) GPa, K b ₀ = 113.1(7) GPa, and K c ₀ = 297(1) GPa with K’ a ₀ = 2.1(3), K’ b ₀ = 5.08(19), and K’ c ₀ = 11.1(4) for sillimanite, and K a ₀ = 99.6(7) GPa, K b ₀ = 152.2(9) GPa, and K c₀ = 236(3) GPa with K’ a₀ = 5.83(19), K’ b₀ = 7.6(3), and K’ c₀ = 5.5(9) for andalusite. The major compression mechanism in both structures involves shortening of bond lengths within the AlO 6 octahedra with volume reductions of 7.4% and 5.1% in sillimanite and andalusite, respectively, over the pressure ranges studied. In andalusite there is also significant compression of the AlO 5 polyhedra and, to a lesser degree, the SiO 4 tetrahedra that display reductions of 5.0% and 3.1% in volume, respectively. In sillimanite there is no significant compression of either the AlO 4 or SiO 4 tetrahedra which behave as rigid, incompressible units.

The tetragonal hollandite structure (KAlSi 3 O 8 hollandite) has been studied up to 32 GPa at room temperature using high-pressure in-situ X-ray diffraction techniques. A phase transformation from tetragonal I4/m phase to a new phase was found to occur at about 20 GPa. This transition is reversible on release of pressure without noticeable hysteresis and hence this new high-pressure phase is unquenchable to ambient conditions. The volume change associated with the transition is found to be small (not measurable), suggesting a second order transition. The diffraction pattern of the highpressure phase can be indexed in a monoclinic unit cell (space group I2/m), which is isostructual with BaMn 8 O 16 hollandite. The γ angle of the monoclinic unit cell increases continuously above the transition. A Birch-Murnaghan equation of state fit to pressure-volume data obtained for KAlSi 3 O 8 hollandite yields a bulk modulus K 0 = 201.4 (7) GPa with K’ 0 = 4.0.

Oursinite is a rare Co-bearing uranyl silicate of the uranophane group. The structure of oursinite, (Co 0.8 Mg 0.2 )[(UO 2 )(SiO 3 OH)] 2 (H 2 O) 6 , is orthorhombic, space group Cmca, a = 7.0494(5), b = 17.550(1), c = 12.734(1) Å, V = 1575.4(2) Å 3 , Z = 4. It was solved by direct methods and refined on the basis of F 2 for all unique reflections using least-squares techniques to an agreement index (R1) of 2.66%. The structure contains an approximately linear (UO 2 ) 2+ uranyl ion that is present as a uranyl pentagonal bipyramid, one symmetrically distinct SiO 3 OH acid silicate group, and one M 2+ (OH,H 2 O) 6 octahedron (M is dominated by Co). The uranyl pentagonal bipyramids and silicate tetrahedra are linked by the sharing of edges and vertices, giving a sheet based upon the uranophane anion topology. Adjacent sheets are linked by M 2+ (OH,H 2 O) 6 octahedra located in the interlayer, and by hydrogen bonds. Each M 2+ (OH,H 2 O) 6 octahedron contains two OH groups that are apical ligands of silicate tetrahedra in adjacent uranyl silicate sheets. Although several uranophane-group minerals contain sheets that are based upon the uranophane anion topology, the oursinite sheet involves novel orientations of silicate tetrahedra.

Aluminum is an important minor constituent of a number of high-pressure mantle silicates in which it substitutes for octahedrally coordinated silicon. In several cases, its solid solution may be linked to the presence of oxygen vacancies; in others, to charge balance with H + . Here we present new data from high-resolution, high-field (18.8 Tesla) 27 Al NMR of aluminous stishovite and of a non-stoichiometric perovskite with nominal composition MgSi 0.95 Al 0.05 O 2.975 . For the stishovite, we characterize the local structure of the symmetrical, octahedral site for Al. These results, combined with 27 Al{ 1 H} REDOR NMR, are consistent with hypothesized H + charge balance, although the presence of a significant fraction of randomly distributed oxygen vacancies could remain undetected. As in a recent previous study of a related perovskite composition, the observed ratio of Al at symmetrical octahedral B sites to that of Al at large, central A sites is about 2:1, indicating the presence of oxygen vacancies to account for charge neutrality in this phase. Such vacancies are not preferentially associated with the Al octahedra, however, suggesting a random distribution in the structure.

The shear viscosities of Cl-bearing melts in the system Na 2 O-Fe 2 O 3 -Al 2 O 3 -SiO 2 were determined at temperatures of 550-950 °C in the range of 10 8.5 to 10 12.0 Pa s using the micropenetration technique. The compositions are based on addition of Fe 2 O 3 or FeCl 3 to aluminosilicate glasses with a fixed amount of SiO 2 (67 mol%). Although there was loss of Cl - during the glass syntheses, no loss occurred during the viscometry experiments. It is to be expected that Cl - takes the structural position of O 2- , and thus reduces the polymerization of the melt structure, and therefore the viscosity of the melt; as F - does. Our measurements show that the presence of Cl - increases or decreases the viscosity of the melts as a function of melt composition. In the present melts, at least 10% of the Fe exists as network-modifying or charge-balancing Fe 2+ ; whereas the rest exists as network-forming Fe 3+ . It is proposed here that the different effects of Cl on viscosity are due to the preferred Cl - -Fe 2+ NBO bonding together with the different structure of peralkaline and peraluminous melts. In peralkaline aluminosilicate melts, the addition of Cl 2 O -1 will destroy 2 NBOs and create one BO if Cl - bonds primarily to the Fe 2+ creating non-bridging O atoms. This would result in an increase in viscosity. In peraluminous melts, the addition of Cl 2 O -1 may result in Cl - bonds to the charge-balancing Fe 2+ , creating 2 new tri-clusters [assuming (Al 3+ ,Fe 3+ )Si 2 O 5 tri-clusters exist]. The preference of Cl - to form bonds to the NBO-forming Fe 2+ is indicated by the small amount of Cl - soluble in the peraluminous melt structure in comparison to that soluble in the peralkaline structure.

Fluid-bearing microinclusions in diamonds (<1 μm) provide a unique source of information on the diamond-forming medium. Transmission electron microscopy (TEM) investigation of such microinclusions enables the detailed study of their size, external habit, internal morphology, and mineralogy, and yields information on the chemical composition and crystallography of the included phases. Here we present a detailed TEM examination of microinclusions in four fibrous diamonds from Canada and Siberia, each with a distinctive inferred original fluid composition. Most microinclusions contain multi-phase assemblages that include carbonate, halide, apatite, possible pyroxene, and high-silica mica (6.8-7.7 Si atoms per formula unit) whose composition lies along the phlogopite-Al-celadonite join. The TEM results, together with the tight range of composition detected by electron probe microanalysis (EPMA) and the volatiles detected by infrared (IR) spectroscopy, suggest that the microinclusions trapped a uniform, dense, supercritical fluid and that the crystallized minerals grew as secondary phases during cooling. Carbonates appear in all assemblages, together with either halides or silicates, indicative of the importance of carbonatitic high-density fluid during diamond growth and fluid evolution. The presence of halide-carbonate or silicate-carbonate assemblages is in agreement with the bulk composition of the microinclusions as detected by EPMA. The high K content of some microinclusions detected by EPMA cannot be accounted for by the solid phases analyzed by TEM. This discrepancy suggests that K is concentrated in the residual fluid that is lost during TEM sample preparation. In addition to microinclusions, large cavities containing amorphous phases were found in the inner parts of one Siberian and one Canadian diamond. An Al-rich phase is the most abundant, and it is accompanied by Ca-rich and Si-rich phases. These phases may be explained by amorphization of crystalline phases. A breakdown of a single melt into three immiscible components is less likely.

The densities of 8 Zn-bearing silicate melts have been determined, in equilibrium with air, in the temperature range of 1363 to 1850 K. The compositional joins investigated [sodium disilicate (NS2)-ZnO; anorthite-diopside 1 bar eutectic (AnDi)-ZnO; and diopside-petedunnite] were chosen based on the pre-existing experimental density data set, on their petrological relevance, and in order to provide a test for significant compositionally induced variations in the structural role of ZnO. The ZnO concentrations investigated range up to 25 mol% for sodium disilicate, 20 mol% for the anorthitediopside 1 atm eutectic, and 25 mol% for petedunnite. Molar volumes and expansivities have been derived for all melts. The molar volumes of the present liquids decrease with increasing ZnO content. The partial molar volume of ZnO derived here from the volumetric measurements for each binary system is the same within error. A multicomponent fit to the volumetric data for all compositions yields a value of 13.59(0.55) cm 3 /mol at 1500 K. We find, herewith, no volumetric evidence for compositionally induced coordination number variations for ZnO in alkali-bearing vs. alkali-free silicate melts nor for Al-free vs. Al-bearing silicate melts.

UHT metamorphosed aluminous granulites from the Eastern Ghats Belt, India, contain contrasting mineral assemblages in closely spaced domains. Reaction textures and mineral chemistry in one domain are consistent with prograde metamorphism. A unique intergrowth texture is found in garnet porphyroblasts from the other domain. Lamellar to vermicular intergrowth of spinel solid-solution occurs in garnet, particularly in the central part of the porphyroblast. Locally, sapphirine, overgrowing spinel, is also intergrown with garnet. Three possible mechanisms are discussed to account for the garnet-spinel intergrowth: (1) from early stabilized orthopyroxene + sillimanite; (2) spinel forming pseudomorphs after early sillimanite inclusions in garnet; and (3) from an early stabilized hypothetical, highly aluminous orthopyroxene involving a complex reaction between the (Fe,Mg)-Tschermak and Fe3 + -Tschermak components. Sapphirine is produced by a reaction involving spinel and aluminous pyroxene. Regardless of the exact mechanism, available thermodynamic data suggest that all such reactions presumably occurred due to loading and/or heating during prograde metamorphism consistent with the conclusion from the other domain. Such an intergrowth texture has not been reported previously in aluminous granulite parageneses, but these could provide crucial information regarding the prograde segment of the metamorphic evolution of deep crustal rocks.

High-temperature (867-1534 ºC) density measurements were performed in air on 10 liquids in the Na 2 O-Fe 2 O 3 -FeO-SiO 2 (NFS) system and 5 liquids in the K 2 O-Fe 2 O 3 -FeO-SiO 2 (KFS) system using Pt double-bob Archimedean method. Replicate measurements indicate an average reproducibility of 0.22%. Compositions (in mol%) range from 4 to 18 Fe 2 O 3 , 0 to 3 FeO, 18 to 39 Na 2 O, 25 to 37 K 2 O, and 43 to 67 SiO 2 . Errors in the gram formula weight are ~0.4%. The molar volumes were fitted to a linear compositional model, which gives a compositionally independent partial molar volume (±2σ) for the Fe 2 O 3 component (V̄ Fe₂O₃ ) of 41.52 ± 0.34 cm 3 /mol and zero thermal expansivity. The average residual to the fit is ±0.36% for our 57 measurements on 15 liquids at various temperatures. The value for V̄ Fe₂O₃ in silicate liquids when Fe 3+ is in fourfold vs. sixfold coordination is estimated to be ~45.5 ± 1 vs. ~34 ± 1 cm 3 /mol, respectively. Thus, the fitted value of 41.5 cm 3 /mol appears to reflect an average Fe 3+ coordination number between 4.5 and 5.0, which is consistent with recently published X-ray absorption fine structure (XAFS) spectroscopy and molecular dynamics (MD) simulations on Fe 3+ - bearing silicate glasses. In the literature, V Fe 3+ is inferred to be present in trigonal bipyramidal sites, in contrast to the square pyramidal sites for V Ti 4+ . The lack of a strong compositional or temperature dependence for V̄ Fe₂O₃ in these alkaline silicate liquids, in contrast to what is observed in the literature for V̄ TiO₂ in similar melts, may reflect the different geometries for V Fe 3+ and V Ti 4+ .

A model for the modulated crystal structure of an antigorite polysome with m = 16 (where m is related to the number of tetrahedra spanning a wavelength along a) was refined by single-crystal synchrotron diffraction data in C2/m, using crystals coexisting with the m = 17 polysome from Val Malenco, Italy, which was previously determined structurally. Lattice parameters [a = 81.664(10), b = 9.255(5), c = 7.261(5) Å, β = 91.409(5)°] were determined using a single-crystal diffractometer equipped with an area detector at the Desy synchrotron (Hamburg). The structure was solved by direct methods, and the model refined using 19 222 symmetry-related reflections. The final R 4σ factor was 0.0951, calculated for 7246 reflections. The structure of the m = 16 antigorite polysome strongly resembles that of the m = 17 polysome. A continuous, wavy octahedral sheet is linked to a tetrahedral sheet, reversing its polarity through sixfold tetrahedral and eightfold tetrahedral rings. The half-wave has a curvature radius of 80.1 Å. Polyhedral geometry, ditrigonalization angles, and interlayer O-O distances are similar in the two polysomes. The only differences concern the number of tetrahedra for the m = 16 polysome (an even number which leads to symmetric half-waves) and the periodic b/2 shift involving the eightfold rings (to produce the doubling of the a parameter and a C-centered cell).

Oxidative drop solution calorimetry is being developed as a general method for sulfide thermochemistry. The samples are dropped from room temperature into molten 3Na 2 O·4MoO 3 solvent at 975 K, with oxygen bubbling through the melt to ensure rapid and complete conversion of sulfide to dissolved sulfate. Complete dissolution of sulfides and retention of sulfur in the solvent is documented by furnace tests and visual observation, consistent determination of enthalpy of drop solution, and comparison with previous data for the heat of formation of ZnS, PbS, and CdS. Enthalpies of formation (kJ/mol) from the elements (ΔH 0 f ) are determined for sphalerite (ZnS) (-206.53 ± 4.03 kJ/mol), galena (PbS) (-98.12 ± 4.37 kJ/mol), greenockite (hexagonal CdS) (-148.79 ± 4.13 kJ/mol), and hawleyite (cubic CdS) (-147.65 ± 4.28 kJ/mol). Thus, hawleyite appears to be energetically very similar to greenockite but possibly slightly metastable by about 1 kJ/mol. The results confirm that oxidative drop solution calorimetry in molten sodium molybdate is a viable method for sulfide thermochemistry. It will be most useful for sulfides with moderate heats of oxidation (e.g., the Fe-S, Co-S, and Ni-S systems), and should be applicable to ternary compounds, e.g., the Fe-Ni-S and Ni-Co-S systems, and systems showing large homogeneity ranges, as well as to other chalcogenides and pnictides.

The effect of pressure on the Fe 3+ /ΣFe ratio of an anhydrous andesitic melt was determined from 0.4 to 3.0 GPa at 1400 °C with oxygen fugacity controlled internally by the Ru + RuO 2 buffer. Values of Fe 3+ /ΣFe were determined by Mössbauer spectroscopy on quenched glasses with a precision of ±0.01, one standard deviation. This precision was verified independently by XANES spectroscopy of the same samples. The XANES spectra show a systematic increase in energy and decrease in intensity of the 1s → 3d transition with increasing pressure. The results to 2.0 GPa are in good agreement with predictions from density and compressibility measurements fitted to a Murnaghan equation of state, but the datum at 3.0 GPa has higher Fe 3+ /ΣFe than predicted from the trend established by the lowerpressure data. This might be due to a coordination change in Fe 3+ at high pressure; although there is no evidence for this in the Mössbauer spectra, such a change could account for the change in intensity of the 1s → 3d transition in the XANES spectra with pressure.

We investigated the thermal decomposition of muscovite in natural granite powders heated to 1175 °C for durations from 5 min to 68 h, at 1 bar, paying special attention to the early stages of decomposition. This study shows that muscovite is completely transformed after 5 min. Muscovite pseudomorphs consist of glass, mullite, and Al-rich oxides. For short durations (5 and 40 min), the Al-rich phase was identified by XRD, electron diffraction, and TEM microanalysis as γ-Al 2 O 3 containing 4-8 wt% FeO (total Fe), probably a few weight percents of MgO, and possibly up to 10 wt% SiO 2 . Faint superstructure spots and diffuse streaks observed in electron-diffraction patterns suggest vacancy or trace elements ordering in the γ-Al 2 O 3 defect spinel structure. γ-Al 2 O 3 displays an unexpected acicular morphology, elongated along three directions at 120° in the basal (001) musc planes and parallel to lateral faces of the former muscovite. Mullite forms rods elongated in the basal (001) musc planes along a direction at 90° from one set of γ-Al 2 O 3 needles. The γ-Al 2 O 3 structure appears to be a metastable phase that is replaced by corundum for longer durations.

We describe here the synthesis of amphiboles along the nominal Na(NaMg)Mg 5 Si 8 O 22 (OH) 2 - Na(LiMg)Mg 5 Si 8 O 22 (OH) 2 join, at 800 °C, 0.4 GPa. High amphibole yields (>90%) plus minor quartz and enstatite have been obtained at all compositions; amphibole crystals are acicular and their size rarely exceeds 20-30 × 0.5-3 μm. TEM analysis shows the presence of h+k odd reflections in all samples, indicative of a P-lattice. By similarity with closely related amphiboles from the literature (e.g., Oberti et al. 2000; Cámara et al. 2003) a P2 1 /m space group was assigned to the amphiboles synthesized here. Refined cell-parameters from X-ray powder-patterns show a linear decrease as a function of increasing Li at M4, a and β being the most affected parameters. The four infrared OH-stretching spectra all show two main bands at 3741-3748 and 3712-3716 cm -1 , respectively. They are assigned to two independent O-H groups in the P21/m structure, interacting with a strongly delocalized A Na. The spectra show in addition two minor absorptions at about 3688 and 3667 cm -1 , respectively; these bands are assigned to vacant A-sites in the structure and indicate slight departure of the nominal composition toward cummingtonite. The present work shows that one apfu of B Na can also be completely replaced by one apfu of B Li (M + ), in synthetic Na(M + Mg)Mg 5 Si 8 O 22 (OH) 2 , and that all compositions have P2 1 /m symmetry at ambient conditions.

In this study, we investigated the dissolution kinetics of wollastonite, a pyroxenoid with composition CaSiO 3 , at earth-surface temperatures. We used vertical scanning interferometry (VSI) to quantify the surface normal retreat (or advance) of a crystallographically defined cleavage face on a single wollastonite crystal as a function of time. Nanometer-scale measurements of (100) faces show that the wollastonite surface undergoes a net expansion normal to the surface and relative to the pristine mineral surface. These observations were made during the dissolution process at acidic conditions. The swelling is the result of the formation of a cation-depleted layer at the wollastonite-water interface. We combined VSI measurements with electron microprobe measurements to track the chemical evolution of the leached layer during dissolution, and then used previously reported Ca release rates to calculate the density of the cation-depleted layer. Using an average Ca release of 9.09 × 10 -9 mol/m 2 ·s at pH 3 and 25 °C, the density of the leached layer is 1.2 g/cm3, and grows to a thickness of ~50 nm after 20 hours of reaction. Although our observations do show direct evidence of the formation of a cation-depleted layer during dissolution, they do not favor a particular mechanism of formation of this layer. At this time, our results cannot rule out either cation migration through the depleted crystal structure or re-precipitation at the mineral surface.

Multi-anvil experiments have been performed in the system CaCO 3 -MgCO 3 at pressures of 5.0-7.0 GPa and temperatures of 800-1600 °C. The reaction dolomite = aragonite + magnesite has been reversed and located near 750 °C at 5.0 GPa, 900 °C at 6.0 GPa, and 1000 °C at 7.0 GPa. Between 5 and 6 GPa, the reaction boundary is strongly curved and its dP/dT slope increases from 2 to 12 MPa/°C, expanding the stability field of dolomite. This increase is attributed to increasing Ca-Mg disorder in the dolomite. In addition to the reaction boundary, an isobaric section of CaCO 3 -MgCO 3 has been determined at 6.0 GPa. There is a two-phase field of aragonite + dolomite on the Ca-rich side, which closes before minimum melting temperatures of 1350 °C are reached. The two-phase field on the Mg-rich side, where dolomite + magnesite coexist, intersects with the solidus. Inferred minimum melts are close to dolomite composition suggesting congruent melting of dolomite at 6 GPa. The melt compositions and temperatures in the pure carbonate system are surprisingly similar to solidus phase relations in the CMAS-CO 2 system, implying that minimum melting conditions in carbonated peridotite at high pressures are dominated and controlled by the carbonate component.

Low- and high-temperature heat capacities of natural ferrocarpholite and magnesiocarpholite samples from the island of Crete (Greece) were measured using the heat-capacity option of the Physical Properties Measurement System (PPMS, Quantum Design), which is based on the principles of heat-pulse calorimetry (HPC), and by differential scanning calorimetry (DSC). Comprehensive chemical analyses also have been performed on these samples by electron-microprobe analysis (EMPA), inductively coupled plasma mass spectrometry (ICP-MS), and Karl-Fischer titration (KFT) for H 1 O. In addition, the ferrocarpholite sample has been investigated by Mössbauer spectroscopy (MS). The measured heat capacities were corrected to end-member compositions of ferrocarpholite and magnesiocarpholite. The standard entropy of end-member ferrocarpholite Fe 2+ Al 2 [Si 2 O 6 ](OH) 4 and end-member magnesiocarpholite MgAl 2 [Si 2 O 6 ](OH) 4 were determined as 248.1 ± 0.3 J/(mol·K) and 221.5 ± 0.2 J/(mol·K), respectively. The C P -polynomial for end-member ferrocarpholite is C P = 652.05 - 6.0995 ×10 3 T -0.5 - 2.0980 × 10 6 T -2 + 2.9396 × 10 8 T -3 [J/(mol·K)], valid in the temperature range of 298.15-625 K, and that for end-member magnesiocarpholite is C P = 644.08 - 6.0921 × 10 3 T -0.5 - 2.2825 × 10 6 T -2 + 3.3182 × 10 8 T -3 [J/(mol·K)], valid in the temperature range of 298.15-725 K.

Chemical and crystallographic properties of clinopyroxene exsolution in wollastonite are described from metamorphosed calc-silicate granulite, Namaqualand, South Africa. The wollastonite is Ca 1.96 Fe 0.01 Al 0.01 Si 2.01 -O 6 belonging to space group P2 1 /a (2M polytype) and the clinopyroxene is Ca 0.99 Mg 0.75-0.80 Fe 0.17-0.21 Na 0.02 Al 0.03 Si 1.99-2.00 O 6 belonging to C2/c. An electron backscattered diffraction investigation suggests that the clinopyroxene lamellae elongated along [11̅1] lie on (120) and (100) of the wollastonite-2M, and [11̅0] of both lamellae are parallel to [001] of the wollastonite-2M. The formation of the exsolution probably results from the relatively high peak metamorphic temperature (800-860ºC) of the Namaqualand granulite and its slow cooling rate.

Melliniite, ideally (Ni,Fe) 4 P, is a new mineral from the Northwest Africa 1054 acapulcoite. It occurs as anhedral to subhedral grains up to 100 µm, associated with kamacite and nickelphosphide. Melliniite is opaque with a metallic luster and possesses a gray streak. It is brittle with an uneven fracture; the Vickers microhardness (VHN 500 ) is 447 kg/mm 2 (range 440-454) (calculated Mohs hardness of 8-8½). The calculated density is 7.88 g/cm 3 (on the basis of the empirical formula). In plane-polarized reflected light, melliniite is cream-yellowish. Between crossed polars it is isotropic, with no internal reflections. Reflectance percentages (R) for the four standard Com wavelengths are 60.5 (471.1 nm), 50.4 (548.3 nm), 52.5 (586.6 nm), and 55.9 (652.3 nm), respectively. Electron-microprobe analyses give the chemical formula (Ni 2.30 Fe 1.64 Co 0.01 ) Σ=3.95 P 1.05 on the basis of total atoms = 5. Melliniite is cubic, space group P2 1 3, with a = 6.025(1) Å, V = 218.71(6) Å 3 , and Z = 4. The crystal structure has been solved and refined to R = 2.72% using single-crystal X-ray diffraction data. It shows the AlAu 4 -type structure, which is an ordered form of the β-Mn structure. The metal atoms occupying the 12b site (M1) are effectively 14 coordinated whereas the metals at the 4a site (M2) are 12 coordinated by three P atoms and nine metals. The phosphorus atoms (4a site) coordinate 12 metal atoms in a somewhat distorted icosahedral arrangement. This new phosphide is the first phase with such a high coordination-number for phosphorus. The new mineral is named after Marcello Mellini, Professor of Mineralogy, who strongly developed the study of meteorites in Italy. The new mineral and mineral name have been approved by the Ima Commission on New Minerals and Mineral Names (2005-027).

The complexity of silicate minerals makes prediction of their dissolution rates a challenging problem. A combination of large cluster ab-initio quantum mechanical models and chemical probe dissolution experiments are used to understand the dissolution process for olivine-group minerals. Rapid release of M 2+ cations by precursor reactions involving H + attack at µ 3 -O surface sites produces a silica-enriched surface. Slower rates of silica release via a ligand exchange reaction involving a proton in the activated complex controls the overall rate of olivine dissolution. Our results provide a physical explanation for the correlation among olivine dissolution rates and water exchange rates for the corresponding aqueous cation.

The pure synthetic end-member of antigorite was studied by in-situ Raman spectroscopy in a diamond anvil cell. It can be metastably compressed up to 10 Gpa at room temperature without occurrence of phase transition or amorphization. The spectrum in the OH region is simpler than in natural antigorite, allowing identification and assignment of the two observed bands at 3672 and 3698 cm -1 to the in-phase stretching modes of the outer O3-H3 bonds (brucite-like) and to the stretching mode of the inner O4-H4 bonds (talc-like), respectively. A broad weak shoulder on the low frequency side of the OH bands near 3650 cm -1 is better resolved above 7 Gpa and assigned to the out-of-phase stretching mode of the outer OH. The two strong OH peaks have small positive pressure dependences (average of 2.3 cm -1 /Gpa), indicating no enhancement of hydrogen bonding at high pressure. One broad OH band at about 3670 ± 20 cm -1 has a large pressure-induced shift (10 ± 2 cm -1 ). It could be related to structural defects. The specific behavior of OH bands in serpentines can influence the high pressure D/H partitioning with respect to other hydrated minerals.

Shock recovery experiments of sinoite (Si 2 N 2 O) were carried out in a pressure range of 28-64 Gpa. Post-shock samples were investigated by optical microscopy, X-ray diffraction, micro Raman spectroscopy, and transmission electron microscopy coupled with energy dispersive X-ray analysis. Sinoite was stable up to ~28 Gpa, being partially amorphous at 34 Gpa, and almost completely amorphized above 41 Gpa. These results are consistent with the observation that sinoite is found in enstatite chondrites classified into shock stage S2, and imply that sinoite may not be a crystallization product from impact melts and that it may be a metamorphic product of high temperatures and relatively low pressure.

A method is proposed for determining the water concentration in silicate melt inclusions (Mi) by confocal micro-Raman spectroscopy, without exposing the inclusions for measurement (a prerequisite of all previous methods). The latter is important for extremely water-rich Mi (e.g., those in evolved granites and pegmatites), which would loose H 2 O on exposure. Furthermore, this technique permits determination of the water concentration in a single Mi. We use a comparative technique, determining the total water content of a sample against a reference glass of known water content. Because this process is non-destructive it does not preclude the subsequent use of other analytical techniques.

The apparent dichotomy of Mg/Si and Al/Si ratios between terrestrial rocks and Martian meteorites has been interpreted as indicative of major differences between the terrestrial and Martian magma source regions. We suggest that this apparent dichotomy is not robust when compared with partly cumulate and non-cumulate terrestrial igneous material. Terrestrial cumulate intra-plate nodules with similar mineralogy to the SNC meteorites plot in the SNC field in this compositional space and far removed from the "terrestrial geochemical fractionation" line (Earth's crust line) of Jagoutz et al. (1979). As is the case for terrestrial partly cumulate igneous rocks, the bulk compositions of cumulate SNC meteorites such Chassigny and Nakhla are dominated by the chemical characteristics of the accumulated minerals, minerals compositionally similar to those found in terrestrial intra-plate magmas. Therefore, the minor amounts of liquid involved play an insignificant role in the Mg/Si and Al/Si signature and no special Martian chemical characteristics can be identified. SNC meteorites considered as possibly representative of liquid compositions are similar in composition and Mg/Si and Al/Si to terrestrial ferropicrites, suggesting that even "liquid" compositions within the SNC space are not uniquely Martian. Therefore, the Mg/Si and Al/Si ratios cannot clearly distinguish Martian from terrestrial rocks.

The structure and equation of state for CaIrO 3 -type MgSiO 3 were determined using high-pressure powder X-ray diffraction from 116 to 144 Gpa. The CaIrO 3 -type phase remained stable over this entire pressure range. At each pressure increment, the sample was heated with a laser to relax the deviatoric stress in the sample. Pressure-volume data could be fitted to the Birch-Murnaghan equation of state with K 0 = 237(1) Gpa using Anderson's gold pressure standard, when K 0 ’ and V 0 were set to 4 and 162.86 Å 3 , respectively. This indicates that the bulk modulus of MgSiO 3 decreases when the phase transition occurs, because the bulk modulus of perovskite-type MgSiO 3 is 250-260 Gpa. The b axis of the unit-cell parameter was more compressive than the a and c axes. The unit-cell volumes at high pressures, observed in high-pressure experiments, were slightly smaller than those predicted by theoretical studies.

This paper describes a high-pressure synchrotron infrared (Ir) spectroscopy study of the synthetic amphibole Na(NaMg)Mg 5 Si 8 O 22 (Oh) 2 . This compound has P 2 1/m symmetry at room conditions; its IR OH-stretching spectrum consists of two main bands at 3743 and 3715 cm -1 , which are assigned to the two symmetrically independent O-H groups in the P structure (sample 403, Iezzi et al. 2004a). For increasing pressure, both bands shift toward higher frequency, suggesting a shortening of the O-H bond. In addition, the two bands progressively merge to give a single, symmetric and broad absorption band at 20-22 Gpa. This behavior suggests that at 20-22 Gpa there is a unique O-H group in the structure, indicative of a C-lattice type. The Ir data thus show that the examined sample undergoes a P2 1 /m ↔ C2/m phase-transition at that pressure. Upon release of pressure, the initial two-band pattern is immediately recovered indicating that the pressure-induced phase-transition is reversible, as already observed for the same kind of transition induced by temperature. By analogy with structurally related pyroxenes, and taking into account the possible crystal structural topologies of amphiboles, we suggest that the C2/m polymorph stable at high pressure is characterized by fully kinked double-chains.