Volume 94, Issue 1

We used configurational entropy theory to model the viscosity (η) of hydrous melts of NaAlSi 3 O 8 , haplogranite (SiO 2 -KAlSi 3 O 8 -NaAlSi 3 O 8 ), and complex (natural) granite composition from available measurements and recently published configurational heat-capacity data. The equation log η = A e + B e /TS conf (T), where S conf is configurational entropy, reproduces viscosity data for individual samples as well as or better than the empirical three-parameter TVF equation (defined below), and has the advantage of being based on thermodynamic theory. The variables A e , B e , and Sconf(T g ), where T g is glass transition temperature, were parameterized as a function of water content for compilations of viscosity data for hydrous NaAlSi 3 O 8 , haplogranite, and peraluminous granite melts. With the simplest assumption of ideal mixing between silicate and water components, configurational entropy models with between 4 and 10 fitting parameters reproduce experimentally determined η-T-XH 2 O relationships significantly better than previous literature models based on empirical equations. Our preferred configurational entropy models have root-mean-square deviations of 0.26 log units for NaAlSi 3 O 8 (n = 77), 0.16 log units for haplogranite (n = 55), and 0.28 log units for peraluminous granites (n = 79). The best statistical fits to the data sometimes require thermodynamically unlikely variations in A e , B e , and Sconf(Tg) as a function of water content, however, such that further calorimetry data are needed to extract accurate thermodynamic information from viscosity data sets for hydrous melts.

The effect of alteration phases on the Pb-isotope systematics of weakly altered basalts from two major ocean islands (Hawaii and Kerguelen) was investigated using scanning electron microscopy (SEM), X-ray diffraction (XRD), and acid-leaching experiments. Alteration within vesicles and as replacement products of olivine and glass consists mainly of phyllosilicate mixtures (most commonly smectite ± talc, serpentine, chlorite, celadonite), goethite, and zeolite with minor pyrolusite, barite, apophyllite, dolomite, pyrite, and chalcopyrite. The presence of distinct alteration assemblages within the Hawaiian and Kerguelen basalts can be related to differences in their eruption environment, age, and sampling method. In particular, the Hawaiian basalts have been contaminated by highly radiogenic components (seawater and drilling mud), demonstrating the importance of acid-leaching prior to Pb-isotopic analysis even for young basalts that appear to be relatively unaltered. However, for some basalts, leaching may not remove the alteration completely or in a reproducible way, which may reflect variable extents of dissolution of secondary minerals that are not readily soluble in HCl (e.g., celadonite) or inconsistencies in the leaching procedure. The incomplete removal of foreign Pb components, which typically have distinct Pb-isotopic signatures, is a major source of uncertainty and may represent the ultimate limitation on high-precision Pb-isotopic compositions of ocean-island basalts. To achieve the highest precision and accuracy, we recommend that Pb-isotope studies of basalts include an investigation of the alteration mineralogy and an evaluation of the effectiveness of the leaching procedure as part of standard quality control protocols.

Abstracts Results of the investigation of paleontological blue and gray bone fragments of small vertebrates coming from stratigraphic layer 770 at San Josecito Cave (Nuevo Leon, Mexico, dating between 28 000 and 19 000 years BP), are presented. Structural and elemental analyses combining X-ray diffraction (XRD), Fourier-transform infrared spectroscopy (FT-IR), transmission electron microscopy (TEM), and particle-induced X-ray and γ-ray emission (micro-PIXE/PIGE), as well as spectroscopic investigations [i.e., UV/visible/near-IR reflectance spectroscopy and X-ray absorption near-edge structure (XANES) spectroscopy] were performed to identify precisely the origin of the blue stain. Prior research has shown that Mn 5+ in tetrahedral coordination could be responsible for the turquoise blue color in mastodon ivory some tens of million years old that was modified by a heat process. Manganese is present in the anionic form of (MnO 4 ) 3- and partially substitute for (PO 4 ) 3- in the hydroxyapatite matrix. The spectroscopic data of the present study have revealed a heat-induced modification, revealed traces of Mn among the typical bone constituents (Ca, P, Sr, Zn), and provided insights into the color origin of the blue paleontological bones from San Josecito Cave. Cations of Mn 5+ in a tetrahedral environment of four O 2- ions in the apatite structure are found in these bones, the same color origin as in the blue mastodon ivory. As indicated by XANES, Mn 4+ ions in octahedral coordination as in pyrolusite are found in gray bones. The presence of submicroscopic Mn oxide inclusions might explain the color of the San Josecito gray bones. The formation of Mn 5+ very likely is induced by heat treatment of the bones under oxidizing conditions. The heat-induced modification of both types of paleontological bones also is indicated by the direct observation of apatite crystals using TEM. The question remains, however, how the heat originated inside the cave.

An albite-omphacite rock, containing the three sodic pyroxenes of the jadeite-omphacite-diopside/ hedenbergite ternary system, has been investigated in detail to better characterize the phase relationships in this system. The rock is from a serpentinite mélange in the Kurosegawa Zone of Central Kyushu, Japan, and exhibits an omphacite overgrowth over diopside/hedenbergite and the further overgrowth of jadeite over omphacite, indicative of the successive replacement of pyroxenes during metamorphism. Partial replacement of omphacite by diopside/hedenbergite at the margins of K-feldspar veins suggests that all three pyroxenes were stable locally at the same stage of metamorphism. The peak temperature and pressure conditions are roughly estimated to be 350 °C and 5-10.8 kbar. Partitioning of Fe 2+ -Mg between pyroxene pairs is consistent with a recent thermodynamic model, indicating the attainment of local equilibrium for pyroxene pairs. The observed compositional gap is also consistent with the miscibility gap reported in many previous studies. The compositional field of omphacite in the ternary jadeite-aegirine-diopside/hedenbergite diagram extends from the jadeite-diopside/hedenbergite line toward the aegirine-rich field, maintaining approximately constant the jadeite content, resulting in an asymmetric phase diagram.

H 2 O can affect the thermophysical properties of the mantle, and nominally anhydrous mantle minerals, such as olivine, pyroxenes, and garnet, may be an important reservoir of mantle H 2 O. However, the H 2 O content of nominally anhydrous mantle minerals now at the Earth’s surface may not always reflect mantle values. It is, therefore, desirable to develop different techniques to estimate mantle H 2 O contents, or values of the activity of H 2 O (ɑ H₂O ) at the conditions of equilibration in the mantle. To examine the potential of amphibole equilibria to determine values of mantle ɑ H₂O , the chemical compositions of co-existing amphibole, olivine, two-pyroxenes, and spinel from a mantle xenolith, sample DH101E of McGuire et al. (1991), were used to estimate values of pressure (P), temperature (T), and ɑ H₂O . A value of ɑ H₂O was estimated from pargasite dehydration equilibria using chemical compositions of minerals as the basis for estimating activities of end-members in the natural phases (e.g., the activity of forsterite in olivine). These calculations were performed with the THERMOCALC software package and, at an estimated maximum T and P of 900 °C and 20 kbar, they yield an estimated value of ɑ H₂O ≈ 0.02 for sample DH101E. The application of oxy-amphibole equilibrium, as described by Popp et al. (2006a, 2006b), using the composition of the amphibole in DH101E yields a value of the log of the hydrogen fugacity (f H₂ ) of -1.37. This value of f H₂ together with the estimated log f O₂ of -9.9 yields a value of ɑ H ₂ O ≈ 0.0005 for sample DH101E. The lower estimated ɑ H₂O compared to that estimated from dehydration equilibria may reflect a slight loss of H from amphibole in the post-formation environment, but both types of amphibole equilibria are consistent with a low value of ɑ H₂O . Values of mantle ɑ H₂O can be used to predict the H 2 O content of mantle olivines. At 900 °C and 20 kbar, the olivine in a sample that equilibrates at ɑ H₂O <0.04, as estimated for sample DH101E, should contain <10 wt ppm H 2 O. This value is consistent with the lower end of the range of measured H 2 O contents of mantle olivines (≈4-400 wt ppm). Thus, estimates of values of ɑ H₂O from amphibole equilibria can produce useful predictions of both the activity of H 2 O as well as the H 2 O content of nominally anhydrous mantle minerals.

Biological apatite, i.e., the major component of teeth and bones, is a widely available source of nanocrystalline apatite. More information is needed about its chemical reactivity in the environment. In the present study, wafers of cross-sectioned dentin and enamel from a modern horse tooth were soaked in phosphate-buffered solutions with NaF concentrations ranging from 0.01 to 2 molar for periods of up to 14 days at about 19.5 °C. The samples were removed at intervals and analyzed by Raman microprobe spectroscopy. Additional, real-time, in-situ Raman spectroscopic analyses were made on some samples during their fluoridation reaction, using an immersible probe. All spectra were deconvolved and their spectral components analyzed for band position, width, and area. Spectral modeling indicates that fluoridation occurred by a kinetically controlled dissolution-reprecipitation process-bioapatite grains partially dissolved and released Ca and P to the F-bearing solution, which caused essentially end-member fluorapatite to nucleate and precipitate, gradually replacing some of the original bioapatite. The replacement process, i.e., fluoridation of the sample, progresses due to the difference in solubility between bioapatite and (highly insoluble) fluorapatite. The bioapatite in (more soluble) dentin reacted much faster and to a much greater extent than that in (less soluble) enamel. These results have implications for paleoenvironmental reconstruction based on the geochemistry of fossil teeth, heavy-metal remediation in soils and water through addition of phosphate phases, and the recommended methods for dental fluoridation in humans.

The thermally induced structural modifications of the natural zeolite barrerite [Na 16 Al 16 Si 56 O 144 ·52H 2 O, a = 13.6239(4) Å, b = 18.2033(5) Å, c = 17.8317(7) Å, V = 4422.3(3) Å 3 , space group Amma, framework type STI] were studied in a temperature-resolved X-ray powder diffraction experiment, using synchrotron radiation, in the temperature range 339-973 K. In the initial stage of heating, up to 508 K, barrerite Phase A (space group Amma) is stable, the unit-cell volume decreases by about 4% and a water release of about 66% is observed. Between 521 and 598 K, a phase transition to the collapsed so-called barrerite Phase B (space group Amma) is observed. During the transition, the rotation of the 4 2 5 4 secondary building units causes a large decrease in cell volume and deformation of the channel system. Phase B, at 611 K, shows the statistical breaking of T-O-T bridges in the 4-rings and the migration of the involved tetrahedral atoms to new “face-sharing” tetrahedra, with a consequent reduction of the free volume of the channels parallel to [100]. The new structure is stable up to 741 K and the total volume decrease is about 16%. A new phase appears from 754 K with cell parameters similar to those reported for the highly deformed barrerite Phase D and is stable up to 910 K, which is the temperature at which the total volume decrease is 22.5%. The material does not undergo amorphization up to the highest temperature investigated.

The crystal chemistry of six crystals of general formula K 2 [M 2+ (H 2 O) 6 ](SO 4 ) 2 , with M 2+ = Mg, Fe, Co, Ni, Cu, and Zn, was investigated by single-crystal structure analysis to determine the effects of the chemical variation of M 2+ on the structural environment surrounding K, M, and S sites. Results indicate that the distortion in the SO 4 tetrahedron and the MO 6 octahedron is very small, except for CuO 6 where it is pronounced because of the Jahn-Teller effect. The KO 8 -octacoordinate polyhedron has the highest degree of distortion, and its idealized shape may be referred to as a bicapped trigonal prism. The SO 4 size is not affected by changes in cation occupancies at the adjacent M site. In contrast, changes in the KO 8 size, which are accompanied by changes in the bond valence sum at K, depend on interaction with the first and second coordination sphere of M. This interaction results by changes in M-O individual lengths, by expansion of the second coordination sphere of M, and by changes in the distribution of the bond strengths over the O atoms coordinated to K. The MO 6 size follows the expected trend from the increased ionic radius at the M site. The latter is also correlated with the unit-cell volume except for the Cu- and Mg-phase, which show a larger cell volume with respect to that expected. Although the relevant octahedral distortion around the Cu 2+ cation explains the volume excess in the Cu-phase, an expansion of the second coordination sphere of Mg 2+ , compared to those of cations of larger ionic radius (such as Zn and Co), explains the excess of the unit-cell volume in the Mg-phase. As the CuO 6 distortion can be caused by the Jahn-Teller effect, the higher ionicity of the Mg atom could be the cause for its anomalous behavior observed in Tutton’s salts. This stereochemical behavior of the Mg atom seems to be consistent with the weakening of the hydrogen bonds in the structure connected to differences in the bonding character of Mg and transition metals when coordinated by water molecules.

The structures, infrared active OH stretching modes, and relative energies of OH-defects in ringwoodite (γ-Mg 2 SiO 4 ) have been studied by first-principles calculations based on density functional theory (DFT). Two types of fully protonated cationic defects in normal spinel were considered at 0 and 20 GPa, i.e., [V Mg (OH) 2 ] x , [V Si OH) 4 ] x defects. In addition, two defects associated with the partial inversion of the spinel structure have been investigated. The first one corresponds to two protons compensating a Mg substituted for Si in tetrahedral site, [Mg Si (OH) 2 ] x , whereas the second defect corresponds to a Mg vacancy located nearby a Mg-Si substitution, [VMg(OH) 2 Mg Si Si Mg ] x . The infrared spectrum and evolution with pressure of these OH-defects make it possible to interpret the major IR absorption bands experimentally observed. The main absorption band at ~3150 cm -1 corresponds to protons located between the O-O pairs shared by 16c and 16d octahedra, instead of OH along the tetrahedral edges as usually proposed in the literature. The large width of this band is most likely related to the association of OH defects with the various cationic configurations related to the partial inversion of a vacancy-bearing spinel structure. The less intense band at ~3675 cm -1 is assigned to hydrogarnet-type defects with a protonation of the tetrahedral edges. This interpretation is consistent with an Mg/Si ratio lower than 2 and its weak variation as a function of water concentration, as experimentally observed. These results emphasize the importance of taking into account the structural relaxation experienced by defects, instead of using empirical correlation, to assign OH stretching bands to specific O-O pairs of the structure.

We measured visible reflectance spectra and colors of hematite (α-Fe 2 O 3 ) reagent powders and a natural feldspar grain containing dispersed hematite microcrystals from room temperature up to 800 °C using an in situ, high-temperature visible micro-spectrometer with dark field optics. The spectrum of room-temperature hematite powder is characterized by a nearly constant reflectivity in the range 400-550 nm, a shoulder near 620 nm, and a reflectivity maximum near 750 nm. The reflectance spectrum is similar to the diffuse reflectance spectra measured by a spectrophotometer and a conventional spectrometer with an integrating sphere. This result indicates that the dark field objective is suitable for measuring visible reflectance spectra of hematite powders with the visible micro-spectrometer. The reflectance of hematite powders in the longer wavelength region (>550 nm) decreases gradually with increasing temperature. The shoulder centered around 620 nm and the reflectance maximum near 750 nm also become indistinct at high temperatures. The calculated L* (dark-light), a* (red-green), and b* (blue-yellow) color values decrease with increasing temperature. This means that the red color of hematite becomes black with temperature increase. The calculated absorption intensities (Kubelka- Munk functions) suggest that this temperature dependence of the hematite powder-reflectance spectra can be mainly explained by a change in band gap absorption edges for a semiconductor (Urbach rule). The visible spectra and colors of a natural feldspar grain containing dispersed hematite microcrystals show a similar change with temperature, indicating that the temperature dependence can be observed under natural conditions at high temperatures, such as in volcanic eruptions.

We report the lattice thermal expansion of zircon-(xenotime-)type LuPO 4 and LuVO 4 in the temperature range of 25-1000 °C from high-temperature powder XRD studies. The details of the high-temperature crystal chemistry of both phases have been determined from Rietveld analysis of the powder XRD data. Both the compounds show appreciably higher thermal expansion than analogous zircon-type silicates. Despite isomorphism, the axial thermal expansion of LuVO 4 shows significant anisotropy compared to LuPO 4 . In the studied temperature range, the average axial thermal expansion coefficients of LuPO 4 are α a = 6.0 × 10 -6 and α c = 7.2 × 10 -6 (°C -1 ) and those of LuVO 4 are α a = 3.6 × 10 -6 and α c = 11.8 × 10 -6 (°C -1 ). However, the average volume thermal expansion coefficients are almost identical. The differences in the thermal expansion behavior of the two structures originate from differences in the expansion and distortion of the LuO 8 polyhedra. The LuO 8 polyhedron in LuVO 4 shows about 30% higher thermal expansion than that in LuPO 4 . The overall thermal expansion behaviors of these two structures are predominantly related to the distortion in the LuO 8 unit, inter-cation distances and spatial arrangement of the Lu-O bonds in the structure.

The solubility of H 2 O-CO 2 fluids in a synthetic analogue of a phono-tephritic lava composition from Alban Hills (Central Italy) was experimentally determined from 50 to 500 MPa, at 1200 and 1250 °C. Contents of H 2 O and CO 2 in experimental glasses were determined by bulk-analytical methods and FTIR spectroscopy. For the quantification of volatile concentrations by IR spectroscopy, we calibrated the absorption coefficients of water-related and carbon-related bands for phono-tephritic compositions. The determined absorption coefficients are 0.62 ± 0.06 L/(mol·cm) for the band at ~4500 cm -1 (OH groups) and 1.02 ± 0.03 L/(mol·cm) for the band at ~5200 cm -1 (H 2 O molecules). The coefficient for the fundamental OH-stretching vibration at 3550 cm -1 is 63.9 ± 5.4 L/(mol·cm). CO 2 is bound in the phono-tephritic glass as CO 3 2- exclusively; its concentration was quantified by the peak height of the doublet near the 1500 cm -1 band with the calibrated absorption coefficient of 308 ± 110 L/(mol·cm). Quench crystals were observed in glasses with water contents exceeding 6 wt% even when using a rapid-quench device, limiting the application of IR spectroscopy for water-rich glasses. H 2 O solubility in the ultrapotassic melts (7.52 wt% K 2 O) as a function of pressure is similar to the solubility in basaltic melts up to 400 MPa (~8 wt%) but is higher at 500 MPa (up to 10.71 wt%). At 500 MPa and 1200 °C, the CO 2 capacity of the phono-tephritic melt is about 0.82 wt%. The high CO 2 capacity is probably related to the high K 2 O content of the melt. At both 200 and 500 MPa, the H 2 O solubility shows a non linear dependence on X f H₂ O in the whole X f H₂O range. The variation of CO 2 solubility with X f CO₂ displays a pronounced convex shape especially at 500 MPa, implying that dissolved H₂O promotes the solubility of CO 2 . Our experimental data on CO 2 solubility indicate that the interaction between phono-tephritic magma and carbonate rocks occurring in the Alban Hills magmatic system may result in partial dissolution of CO 2 from limestone into the magma. However, although the CO 2 solubility in phono-tephritic melts is relatively high compared to that in silicic to basaltic melts, the capacity for assimilation of limestone without degassing is nevertheless limited to <1 wt% at the P-T conditions of the magma chamber below Alban Hills.

Dark-brownish, euhedral crystals of an “allanite-like” mineral occur in a hematite-impregnated Mn-silicate rock at Kesebol, Västra Götaland, Sweden, associated with gasparite-(Ce), chernovite- (Y), rhodonite, andradite, manganoan calcite, and quartz. A structural study was carried out on single crystals-untreated, heated in air, and heated under inert atmosphere-combined with Mössbauer spectroscopy and TEM investigation. In all the untreated crystals the mean <M3-O> distance indicates that Me²⁺(Me = Mn, Fe) prevails at this site (<M3-O> in the range 2.169-2.180 Å), in contrast with chemical data obtained by EPMA that yield a simplified formula Ca(REE³⁺ ⅔ ⃞ ⅓ )Me3 ³⁺(SiO 4 )(Si 2 O 7 ) O(OH), when normalized to Si = 3.00 apfu. Moreover, when a crystal is heated in air, all geometrical and structural variations indicate the development of an oxidation-dehydrogenation reaction, thus confirming that M3 is occupied by divalent cations before heating. The corresponding dehydrogenation is confirmed by a dramatic lengthening of the donor-acceptor distance. A crystal was annealed under inert atmosphere to verify possible effects of radiation damage on the polyhedral volumes. After prolonged annealing at 700 °C, a slight decrease of the unit-cell parameters is observed, suggesting restoring of crystallinity from a “partially metamict” state. Nonetheless, even in the annealed crystal, the <M3-O> distance is still consistent with a dominance of divalent cations at the M3 site. For all the examined crystals, structural data point to an octahedral cation population as follows: M1 = (Me³⁺, Al); M2 = (Al, Me³⁺); M3 = (Me²⁺, Me³⁺). This assumption is also in agreement with the Mössbauer spectrum, which was fitted to two Lorentzian quadrupole doublets for Fe³⁺ and one for Fe²⁺. Values of the isomer shifts (0.36 and 0.37 mm/s for Fe³⁺; 1.11 mm/s for Fe²⁺) and the quadrupole splitting (1.96 and 1.02 for Fe³⁺; 1.90 for Fe²⁺) show that Fe²⁺ (~12% of the total iron) is located in M3, while Fe³⁺ occupies M1 and, to lesser extent, M2. TEM-EDS investigations have revealed chemical heterogeneities related to different degree of radiation damage. In particular, areas showing poor crystallinity are relatively enriched in Si and O with respect to the highly crystalline areas, thus suggesting that EPMA chemical data are biased by the presence of metamict areas enriched in SiO 2 and likely in H 2 O. EPMA data were therefore corrected for the excess of silica. The cation population after correction is in keeping with the structural and spectroscopic data. Disregarding minor substitutions, the ideal chemical formula for the epidote-group mineral from Kesebol is CaREEFe³⁺AlMn²⁺(Si 2 O 7 )(SiO 4 )O(OH), which is related to ferriallanite-(Ce) by the substitutional vector M3 (Mn²⁺) → M3 (Fe²⁺).

In this study, we propose a study of dislocations and plasticity in CaSiO 3 perovskite based on the Peierls-Nabarro modeling using the generalized stacking fault (GSF) results as a starting model. The GSF are determined from first-principle calculations using the VASP code. The dislocation properties such as planar core spreading and Peierls stresses are determined for the four possible slip systems: 〈110〉 {11̅0}, 〈100〉 {011}, 〈110〉 {001}, and 〈100〉 {001} and at 0, 30, and 100 GPa. We find that 〈110〉 {11̅0} is the easiest slip system, but more surprisingly, we show that it bears no Peierls friction, even at the higher pressure. The reasons lie in the ability of these dislocations to split into partial dislocations and in the nature of the stacking fault associated with it.

Classical molecular dynamic simulations are performed to investigate the microscopic properties of hexadecyltrimethylammonium (HDTMA) intercalated smectites with and without the cointercalation of acetate groups. Three model smectites with distinct layer-charge characteristics are selected as the frameworks and their HDTMA intercalates correspond to the representative monolayer, bilayer, and paraffin configurations of the aliphatic chains. In the organoclays without cointercalation, all trimethylammonium headgroups of HDTMA are fixed firmly above the center of the surface six-member rings through electrostatic attractions with surface oxygen, whereas the alkyl chains are a little more mobile. It is found that the cointercalations can significantly increase the basal spacings of the pure HDTMA intercalated smectites. The distributions and mobility of the HDTMA headgroups are not affected by the acetate ions because of the overwhelming attractions coming from the clay sheets. The acetate groups are fixed by the HDTMA headgroups through ion pairing. This simulation study provides a molecular level basis to understand the effects of cointercalation on properties of organoclays.

We report data on the composition and crystal structure of the most Ag-rich (15.63 apfu) natural polybasite yet discovered. It shows the -M2a2b2c polytype. The crystal studied was found in a sample (mineralogical collection of the Royal Ontario Museum) from Gowganda, Timiskaming District, Ontario, Canada. Electron microprobe analysis yields the formula [Ag 6 (Sb 1.78 As 0.18 ) Σ=1.96 S 7 ] [Ag 9 (Ag 0.63 Cu 0.43 ) Σ=1.06 S 4 ]. Lattice parameters are a = 26.2625(4), b = 15.1623(5), c = 24.1061(6) Å, β = 90.045(5)°, V = 9599.0(4) Å 3 . The structure was refined in the space group C2/c to R = 0.0581 using 7725 observed reflections [I > 2σ(I)]. The refinement shows that one of the three structural positions of the B module layer usually occupied by Cu is dominated by Ag. Crystal-chemical characteristics are compared with published data on the other members of the pearceite-polybasite group. Some remarks concerning nomenclature are also given.

For the first time, the structure and dynamics of H 2 O in the interlayer of anomalous 11 Å tobermorite have been analyzed based on ab initio molecular dynamics simulations. The simulations provide detailed information on the structure of the hydrogen bonds formed by H 2 O molecules and OH groups. The calculated structural parameters of the tobermorite building blocks are in good agreement with the experimental model of Merlino et al. (1999), which is based on X-ray diffraction (XRD) measurements. However, in contrast to the measurements, the simulations suggest that the W1 and W3 sites are split between two general positions with 50% occupancy. It is proposed that the experimental studies provide only averaged coordinates of these sites due to the limitations imposed by the polytypic structures. Analysis of the H 2 O dynamics at 321 and 506 K suggest the possibility of a temperature induced order-disorder transition associated with the orientation of O6H···W1 and O6H···W3 hydrogen bonds in the structure of anomalous 11 Å tobermorite. The experimental IR and Raman spectra of 11 Å tobermorite are interpreted based on analyses of the vibrational density of states.

Electronic transitions of siderite and rhodocrosite may have played a role in prebiotic synthesis by photochemical reactions in the Archean. In the work described here, I measured the electronic structures of siderite (FeCO 3 ) and rhodocrosite (MnCO 3 ) using O K-edge X-ray absorption and emission spectroscopy. At the same time, I present theoretical calculations of electronic structure using periodic density functional theory with several different exchange correlation functionals. To help verify the reliability of the basis sets and the exchange correlation functionals, I determined the optimized crystal structures of FeCO 3 and MnCO 3 in the antiferromagnetic (AF) rhombohedral unit cell. All three exchange-correlation functionals give static volumes and cell parameters that differ from experiment by <2%. The O K-edge absorption spectra are assumed to measure the O(2p) density of states in the unoccupied orbitals that are primarily Fe(3d) and Mn(3d) in character. Emission spectra show the O(2p) density of states that comprise the occupied bonding orbitals. Using both spectra, I estimate the O 2- → Fe 2+ band gap in FeCO 3 to be 4.4 ± 0.2 eV while the O 2- → Mn 2+ in MnCO 3 is 5.8 ± 0.2 eV. Theoretical calculations of the density of states calculated using the PBE exchange correlation functional seriously underestimate the O(2p)-Mn,Fe(3d) band gaps while those calculated using the hybrid B3LYP exchange correlation functional (with 20% Hartree-Fock exchange) overestimate the band gaps. A modified B3LYP functional with 10% Hartree-Fock exchange, however, gives a band gap in close agreement with experiment for both phases. Simulated O K-edge spectra, obtained by Gaussian smearing of the projected O(2p) density of states, give a reasonable description of the O K-edge emission spectra but a poor description of the experimental O K-edge absorption spectra. However, the O K-edge absorption spectra are reasonably simulated by the total density of states. Excitations to the C-O antibonding states are predicted to lie near 4.6 eV in siderite.

Fast reactions in the medium from which a crystal is growing (e.g., H + + HCO - 3 ↔ H 2 CO 3 ) occurring near the crystal surface can conflict with the stoichiometry of crystal formula-unit creation (e.g., CO 2- 3 is readily incorporated in calcite but H + is not). As a result, a thin fast-reaction boundary layer (FRBL) is created within which the fast reactions are out of equilibrium even though they are maintained at equilibrium far from the crystal surface. The FRBL corrects the fluxes of components to and from the surface of the crystal; not taking the FRBL into account leads to violation of mass conservation. This “paradox” is resolved by showing how the equations of crystal growth can be solved in the fast-reaction limit without violating mass conservation. As this paradox was not understood in the literature, models and geological conclusions drawn should be re-examined.

Electron microprobe analyses of major- and minor-element oxide components for two glassed samples of natural KLB-1 peridotite are presented. One glass was made with the aid of a phosphate flux, and the second glass was made by laser melting of aerodynamically levitated spheroids resulting in homogeneous silicate glass beads. For unknown reasons, the silicate-phosphate glass yields compositions that are incompatible with the composition of KLB-1 peridotite. However, analysis of the glass bead formed by laser synthesis is believed to give an accurate representation of the composition of KLB-1 peridotite, except for minor loss of Na 2 O owing to volatilization. The new data resolve conflicting FeO, CaO, and TiO 2 values from two older measurements present in the literature. Mass-balance calculations using the new composition measurement combined with new analyses of the mineral compositions in KLB-1 result in a lower sum of squares of the residuals than those using the older measurements. There are appreciable differences in calculated modes from partial-melting experiments of KLB-1 when calculated using older KLB-1 analyses or our new analysis.

Spinel single crystals of 19 compositions along the magnetite-ulvöspinel join were synthesized by use of a flux-growth method. To obtain quantitative site populations, the crystals were analyzed by single-crystal X-ray diffraction, electron-microprobe techniques, and Mössbauer spectroscopy. All results were processed by using an optimization model. The unit-cell parameter, oxygen fractional coordinate, and tetrahedral bond length increase with increasing ulvöspinel component, whereas the octahedral bond length decreases marginally. These changes result in sigmoidal crystal-chemical relationships consistent with cation substitutions in fully occupied sites. As a first approximation, the Akimoto model T (Fe 3+ 1-X Fe 2+ x )M(Fe 2+ Fe 3+ 1-X Ti X )O 4 describes the cation substitutions. Deviations from this model can be explained by an electron exchange reaction T Fe 2+ + M Fe 3+ = T Fe 3+ + M Fe 2+ , which causes M Fe 2+ ≠ 1 and T Fe 2+ /Ti ≠ 1. The resultant S-shaped trends may be related to a directional change in the electron exchange reaction at Ti ≈ 0.7 apfu. In general, variations in structural parameters over the whole compositional range can be split into two contributions: (1) a linear variation due to the T Fe 3+ + M Fe 3+ = T Fe 2+ + M Ti 4+ chemical substitution and (2) non-linear variations caused by the internal electron exchange reaction. In accordance with bond-valence theory, strained bonds ascribable to steric effects characterize the structure of magnetite-ulvöspinel crystals. To relax the bonds and thereby minimize the internal strain under retained spinel space group symmetry, the electron exchange reaction occurs.

The crystal-structure, crystal-chemistry, and low-temperature behavior of a natural phillipsite-Na from the “Newer Volcanic Suite,” Richmond, Melbourne district, Victoria, Australia [K 0.75 (Na 0.88 Ca 0.57 ) Σ1.45 (Al 2.96 Ti 0.01 Si 5.07 )Σ 8.04 O 16 ·6.2H 2 O (Z = 2), a = 9.9238(6), b = 14.3145(5), c = 8.7416(5) Å, β = 124.920(9)°, and V = 1018.20(9) Å 3 , space group P2 1 /m], have been investigated by means of in situ single-crystal X-ray diffraction, thermogravimetric analysis, and electron microprobe analysis in the wavelength dispersive mode. Two accurate structural refinements have been obtained on the basis of single-crystal X-ray diffraction data collected at 298 and 100 K, with: R 1 (F)298K = 0.035, 3678 unique reflections with Fo > 4σ(Fo) and 195 parameters, and R 1 (F)100K = 0.035, 3855 unique reflections, Fo > 4σ(Fo) and 195 parameters. In both refinements, the residuals in the final difference Fourier maps are <1 e - /Å 3 . A configuration of the extra-framework population different from that reported in previous studies is found at room temperature, with two possible sites for potassium (K1 and K2), one sodium/calcium site (Ca), and seven independent sites partially occupied by water molecules (W1, W2, W3, W4, W4′, W5, and W6). The low-temperature refinement shows that the framework component of the phillipsite structure is maintained within the T-range investigated. However, a change in the configuration of the extra-framework content occurs at low temperature: the occupancy of site K2 drastically decreases, while that of site K1 increases, the Ca site is split into two sub-sites (Ca1 and Ca2) and the number of water molecule sites decreases to six (W1, W2, W3, W4, W5, and W6). The rearrangement of the extra-framework population at low temperature is likely due to the change in shape (and size) of the micropores by tetrahedral tilting. The evolution of the “free diameters” with temperature shows that an “inversion” of the ellipticity of the eight-membered ring channel along [010] occurs. The evolution of the unit-cell parameters with T (measured at 298, 250, 200, 150, and 100 K) shows a continuous and linear trend, without evident thermo-elastic anomalies. The axial and volume thermal expansion coefficients (αj = l j -1 ⋅∂l j /∂T, α V = V -1 ⋅∂V/∂T) between 100 and 298 K, calculated by weighted linear regression, yield the following values: α a = 1.8(1) × 10-5, α b = 1.2(1) × 10-5, α c = 1.1(1) × 10-5, and α V = 3.7(1) × 10-5K-1. The thermal expansion of phillipsite is significantly anisotropic (α a :α b :α c = 1.64:1.09:1).

The OH speciation of 18 palygorskite samples from various localities were evaluated by near infrared spectroscopy (NIR) and compared to the corresponding octahedral composition derived from independent, single-particle analytical electron microscopy (AEM). NIR gives evidence for dioctahedral-like (AlAlOH, AlFe 3+ OH, Fe 3+ Fe 3+ OH) and trioctahedral-like (Mg 3 OH) species. Therefore, palygorskite can be approximated by the formula yMg 5 Si 8 O 20 (OH) 2 ·(1 - y)[xMg 2 Fe 2 ·(1 - x)Mg 2 Al 2 ] Si8O20(OH)2, where x is the Fe content of the dioctahedral component, and y is the trioctahedral fraction. The values of x estimated from the NIR data are in excellent agreement with the Fe/( VI Al + Fe) ratio from AEM (R 2 = 0.98, σ = 0.03), thus suggesting that all octahedral Al and Fe in palygorskite participate in M 2 M 2 OH (dioctahedral-like) arrangements. Furthermore, y values from AEM can be compared to NIR (R 2 = 0.90 and σ = 0.05) after calibrating the relative intensity of the Mg 3 OH vs. (Al,Fe) 2 OH overtone bands using AEM data. The agreement between the spectroscopic and analytical data are excellent. The data show that Fe 3+ for Al substitution varies continuously in the analyzed samples over a broad range (0 < x < 0.7), suggesting that fully ferric dioctahedral palygorskites (x = 1) may exist. On the other hand, the observed upper trioctahedral limit of y = 0.50 calls for the detailed structural comparison of Mg-rich palygorskite with sepiolite.