Volume 85, Issue 5-6

Allanite, which is a common accessory mineral in a wide variety of rock types, typically contains high concentrations of Th and U; thus, an in-situ method of U-Th-Pb dating of this phase would have broad application. We describe a method to permit Th-Pb ages of allanite to be determined with approximately ±10% accuracy using a high-resolution ion microprobe. Knowledge of the composition and substitution mechanisms of this complex mineral is key to understanding the relative ionization efficiencies of Th + and Pb + . The chemical compositions of three allanite samples used as age standards (Cima d’Asta Pluton, 275.5 ± 1.5 Ma; Atesina Volcanic Complex, 276.3 ± 2.2 Ma; La Posta Pluton, 94 ± 2 Ma) were determined using an electron microprobe, permitting an assessment of matrix effects on ionization. An ion-microprobe calibration curve involving elemental and oxide species of Th and Pb (i.e., 208 Pb*/Th + vs. ThO 2 + /Th + ) yields highly scattered apparent ages when allanite age standards with different Fe contents are used. However, a three-dimensional plot of 208 Pb*/Th + vs. ThO 2 + /Th + vs. FeO + /SiO + improves the accuracy of the calibration to about ±10%. Even though this level of uncertainty is substantially greater than that expected for U-Th-Pb ionmicroprobe analyses of zircon or monazite, Th-Pb ages of allanite can still be used to address important geologic questions. We used this method to date two metamorphic allanite grains from the footwall of the Main Central Thrust, Nepal Himalaya, and an allanite grain from the Pacoima Canyon pegmatite, California. Allanite inclusions in garnet from Nepal yield significantly older ages than the coexisting monazite, indicating that allanite formation in these rocks records a previous metamorphic cycle that predates slip along the fault. The Pacoima Canyon allanite grain yields a younger age than that reported for zircon, implying Pb loss during cooling of the pegmatite.

The Neoproterozoic Florianópolis batholith provides the major clue for unraveling the pre- Brasiliano Cycle evolution of Santa Catarina State, southern Brazil, because it contains septa of basement units such as the Camboriú and Águas Mornas complexes reactivated during the younger Brasiliano orogeny at about 600 Ma. We establish precise U-Pb zircon SHRIMP ages and Nd isotopic evolution of four key samples from the septa. The zircon population displays complex internal structures in three of the four dated samples. The complex crystals have well-preserved magmatic cores and altered rim domains. The core populations were dated at about 2180 Ma (Águas Mornas complex), about 2160 Ma (Camboriú complex amphibolite xenolith), and 2000 Ma (Camboriú complex orthogneiss), whereas the ages of the measured rims and altered domains are about 590 Ma in all three samples. On the other hand, the Presidente Nereu tonalite has a simpler zircon population without core/rim separation, and yields magmatic ages about 2200 Ma. The Nd isotopic analyses of the three samples emphasize the polycyclic evolution of the batholith, which is much more complex than previously realized. The evolution is marked by two successive Paleoproterozoic accretionary orogenic events at about 2200-2175 Ma and by a crustal melting event at about 2000 Ma, which had an Archaean source (Sm-Nd depleted mantle model age, TDM = 2860 Ma). These orogenic events were followed 1.4 billion years later by a Neoproterozoic hydrothermal/ metamorphic overprint at about 590 Ma. This time gap permits us to identify a long intracratonic period-the Atlantica supercontinent. The unraveling of the Proterozoic evolution and the precise characterization of the basement remnants show that the Paleoproterozoic was a crustbuilding, orogenic interval and was the major source of the much later Neoproterozoic granites. The integrated use of back-scattered electron and cathodoluminescence imaging with SHRIMP spot dating and Nd isotopes provide a clear understanding of the timing of crustal generation and deformation in southern Brazil, including strong evidence for a long-lasting supercontinent cycle.

The radioluminescence and thermoluminescence spectra of synthetic zircon crystals doped with individual trivalent rare earth element (REE) ions (Pr, Sm, Eu, Gd, Dy, Ho Er, and Yb) and P are reported in the temperature range 25 to 673 K. Although there is some intrinsic UV/blue emission from the host lattice, the dominant signals are from the rare-earth sites, with signals characteristic of the REE 3+ states. The shapes of the glow curves are different for each dopant, and there are distinct differences between glow peak temperatures for different rare-earth lines of the same element. Within the overall set of signals there are indications of linear trends in which some glow peak temperatures vary as a function of the ionic size of the rare earth ions. The temperature shifts of the peaks are considerable, up to 200°, and much larger than those cited in other rare-earth-doped crystals of LaF 3 and Bi 4 Ge 3 O 12 . The data clearly suggest that the rare-earth ions are active both in the trapping and luminescence steps, and hence the TL occurs within localized defect complexes that include REE 3+ ions.

Water solubility in silica melts was determined at 100-600 MPa, 1200-1350 °C, and at each temperature (T) was found to increase with pressure (P). At P ≤ 250 MPa, the effect of T on water solubility in silica melts is small and within analytical precision (±0.15 wt% H 2 O). A positive correlation with T was observed at 400 MPa. Increasing solubility of water with increasing T was observed when large amounts of water are dissolved in silica and quartzofeldspathic melts (i.e., when molecular water is the dominant species in the glasses at room temperature), as already observed for feldspar melts. Change in water solubility (expressed in mol%) with decreasing SiO 2 content of the melt is nonlinear along the silica-albite join. In the compositional range Ab 100 to Ab 25 (100 to 25 mol% albite, respectively, compositions calculated on an eight-oxygen basis), the solubility of water at 200 MPa decreases only slightly with decreasing Ab content (-0.1 ± 0.01 mol% H 2 O per mol% albite). However, at Ab contents less than 25 mol%, water solubility decreases sharply with increasing Qz content. Similar behavior was observed at 500 MPa. These results suggest that two different incorporation mechanisms of water in quartzofeldspathic melts must be considered: one corresponding to an NaAlSi 3 O 8 -H 2 O mechanism, the other to an SiO 2 -H 2 O mechanism.

Pargasite compositions corresponding to the nominal structural formula NaCa 2 (Mg 4 Al 1-x Cr x )(Si 6 Al 2 )O 22 (OH) 2 , with O ≤ x ≤ 1, were investigated at 900 °C and 3 and 20 kbar P H₂O . The aims of this work were to determine the extent of the solid solution, i.e., the maximum x value, as a function of experimental conditions, and to characterize the M 3+ (Al 3+ and/or Cr 3+ ) distribution over the M sites. A first series of experiments at 900 ºC and 3 kbar showed the presence of large amounts of Cr 6+ as sodium chromate (Na 2 CrO 4 ) and dichromate (Na 2 Cr 2 O 7 ) in the final fluid phase, which promotes the extraction of Cr from the bulk solid (up to 45 mol%). The oxidizing agent was the oxygen from water, liberated by diffusion of hydrogen from water through the Pt tube wall. To prevent this oxidation, a new double-chamber tube was designed, one chamber contain mg powdered metallic Cr to trap the excess of oxygen and the second chamber initially containing the starting gel and water. Using this tube, (chroimum)-pargasites were produced with a maximum Cr content around 0.43 apfu from EMP analyses (based on 23 O atoms), close to the maximum Cr solubility observed in naturally occurring pargasites. In FTIR spectra of Cr-free pargasite, two intense OH-stretching bands are observed at 3714 and 3681 cm -1 , corresponding to OH groups adjacent to Mg 3 and to Mg 2 Al, respectively, and pointing toward a filled A-site. A third band is observed at 3656 cm -1 , which can be assigned to OH groups pointing toward vacant A-sites. Along the Al-Cr pargasite join, sig­nificant modifications are observed in the OH-stretching region: a new band appears at 3660 cm -1 , which can be assigned to OH groups adjacent to Mg 2 Cr, and pointing toward a filled A-site. Its relative intensity increases regularly with the Cr content, showing that Cr 3+ enters M site(s) adjacent to an OH group. Combining EMP and FTIR data, we conclude that up to 0.27 Cr per formula unit occupies the M3 site. However, these are more octahedral trivalent cations than expected, and these are balanced by lower tetrahedral Al and lower occupancies of the A and M4 sites. This charge arrangement, with important deviations from ideal stoichiometry, is apparently the only stable one under the conditions applied. Results were confirmed at 900 °C and 20 kbar P H₂O , indicating that increasing pressure does not affect Cr solubility in pargasite.

Thiospinels of Cu-(Fe) and platinum-group elements (PGE) are relatively abundant in the Kirakkajuppiua PGE deposit of the Peuikat layered complex. Finland. In actinolite-clinochlore rock that is nearly base-metal sulfide-free and relatively poor in chromite, the thiospinels occur as subhedral or anhedral grains (up to 0.4 mm). They are members of the cuprorhodsite-ferrorhodsite and cuprorhodsite-malanite series, relatively poor in cuproiridsite. and display considerable grain-to- grain variations in Cu. Fe. Pt. and Rh. Strong negative Fe-Cu. Pt(+Ir)-Fe. Rh-Cu. and Rh-Pt and strong positive Pt(+Ir)-Cu and Rh-Fe correlations in these thiospinels are indicative of a coupled substitution: Fe-for-Cu substitution in the tetrahedral (A) sites causes an excess in formal positive charge, which is compensated by Rh-for-(Pt+Ir) substitution hi the octahedral (B) sites. Probable valence states in the Fe-free and Fe-rich end-members of the solid-solution series at Penikat are Cu + [Rli 3+ (Pt,Ir) 4+ ] S 4 - and (FefoCu 0.5 + )Rh 2 S 4 2- . and these suggest the heterovalent substitution scheme A Fe 3+ + 2 B Rh 3+ → A Cu + + 2 B Pt 4+ (+2 Ir 4+ ) to incorporate Fe in the ferrorhodsite-rich end member.

Sixty axinite samples from 24 localities worldwide were characterized by electron microprobe analysis (EMPA) to define the limits of compositional variation. Three samples are very close to the Mn, Fe, and Mg end-members. Temaiy (Mn,Fe 2+ ,Mg) - compositions occur mostly in the ferroaxinite and manganaxinite fields, and are constrained by the relation Mg ≤ Fe. Core-rim chemical zoning was observed in 20 samples, with systematic enrichment of Fe in the core and Mn in the rim, inde­pendent of sample provenance. The chemical composition (including B, H, and Fe 2+ /Fe 3+ ) of 17 homogeneous samples was in­vestigated using electron-microprobe analysis, thermo-gravimetry (TG), ion microprobe (SIMS), crystal-structure refinement (SREF), and Mössbauer spectroscopy (MS). For all samples except pure manganaxinite, most of the iron is Fe 2+ . The content of Fe 3+ and the Fe 3+ /ΣFe ratio mcrease with Mn content up to 0.31 atoms per formula unit (apfu) and 0.80, respectively. Fe 3+ may substitute for Al or also for divalent cations balanced by the OH deficiency: Fe 3+ + O 2- ↔ Fe 2+ + OH - , or Fe 3+ O(Fe 2+ OH) -1 Boron content ranges from 1.88 to 2.07 apfu (±2.5% relative) and shows an inverse relation with Si content. Direct measurement of the 5-tetraliedron size provides structural confirmation of the Si ↔ B exchange. Hydioxyl deficiency accompanies this substitution and the following coupled mechanism is proposed: Si 4+ + O 2- ↔ B 3+ + OH - , or SiO(BOH) -1 Hydrogen content ranges from 1.7 to 2.1 apfu (±5% relative). The deficiency of OH from the stoichiometric value of 2.0 per formula unit is related directly to the number of trivalent and tetrava­lent cations, as OH content plays a crucial role in charge-balance relations. A revised chemical formula for the axinite-group minerals is proposed: [6] [Ca(Ca 1-x Mn x: )(Mn, Fe 2+ Mg,Zn,Al u ,Fe V 3+ ) Σ=1 (Al 2-y Fe y 3+ ] 2 [4] [(B 1-z Si 2 )Si 8 ]O 30 (OH 1-w 0) 2 , where x≤1, u <1, ν <1, y<1, z≪1, and w=(u+ν+z).

A transmission electron microscope investigation was performed on a series of synthetic clinopyroxenes with compositions between Ca 0.7 Mg 1.3 Si 2 O 6 and Ca 0.5 Mg 1.5 Si 2 0 6 . For samples with Ca content lower than 0.6 atoms per formula unit (apfu) selected-area electron diffraction (SAED) patterns showed the presence of reflections violating the C2/c space group (b type, h+k odd), indi­cating a transition to the P2 1 /c symmetry at room temperature. Antiphase domains induced by the transition could be imaged, with irregular boundaries and size decreasing with increasing Ca con­tent. The antiphase domain size and shape appears unrelated to the non-periodic mottled texture that could be imaged in samples with Ca lower than about 0.65 apfu. A comparison with data reported in the literature suggests a different behavior in the cell param­eters of the P2 1 /c clinopyroxenes along the diopside-eustatite (Di-En) join. For compositions be­tween Di 60 En 40 to about Di 40 Eu 60, cell parameters deviate little with respect to the trend shown by Ca-rich C2/c clinopyroxenes, whereas for compositions richer in Mg than about Di 40 En 60, a signifi­cant deviation is present, which is marked by discontinuities in the c and β cell parameters.

Fourier-transform infrared spectra of a synthetic akaganéite, β-FeOOH, were acquired in transmittance, attenuated total reflectance (ATR) and diffuse reflectance. The transmittance spectra showed a distinct dependence on the mode of sample preparation: measurements taken on pellets prepared by pressing the sample with alkali halides (KBr or CsI) displayed bands at 1096, 1050, and 698 cm -1 that were not observed in spectra of the neat material and must therefore be considered artifacts. Variations in the sampling environment (e.g., water and/or organic volatiles) were also observed to exert a noticeable influence on the development of the IR spectra. Infrared bands due to akaganéite were found at 3480 + 3390 (doublet), 1630, 850 + 820 (doublet), 650, 490, and ~420 cm -1 . Diffuse reflectance spectra dominated by volume scattering (loose), diffuse reflectance spectra comprised of both volume and surface scattering (compacted), ATR spectra (surface only), and transmittance spectra (absorbance only, inverse of ATR) showed compatible trends for the akaganéite features both above and below ~1000 cm -1 . This indicates that the multiple akaganéite measurements are consistent and confirms the band assignments.

One-pulse magic angle spinning (MAS) 29 Si and 1 H- 29 Si cross-polarization (CP) MAS nuclear magnetic resonance (NMR) spectroscopy was performed on minerals with OH groups and on hydrous aluminosilicate glass samples. The silanol-group samples used were krauskopfite, rosenhahnite, thaumasite, ussingite, and KHSi 2 O 5 with known Si to H distances and ellenbergerite, the proton positions of which are not as well defined. The Si-H distances from the minerals can be compared with the cross-polarization time constants (τ SiH ) and the proton spin-relaxation times in the rotating frame [τ 1ρ (H)], and a rough correspondence exists between a mineral’s shortest Si-H distance and its τ SiH value. Also, fast [τ 1ρ (H)] values correspond to large bulk H densities. The CP spectra of the hydrous aluminosilicate glass samples were fitted with two peaks, representing two different Si environments within the glass structure. The contact-time curves of the higher frequency peak imply τ SiH similar to the mineral samples with short Si-H distances, and this suggests that the glasses could contain a large fraction of either Si-OH groups or protonated bridging O atoms.

High-resolution core-loss and low-loss spectra of α-quartz were acquired by electron energyloss spectroscopy (EELS) with a transmission electron microscope (TEM). Spectra contain the Si L 1 , L 2,3 , K, and O K core-loss edges, and the surface and bulk low-loss spectra. The core-loss edges represent the atom-projected partial densities of states of the excited atoms and provide information on the unoccupied s, p, and d states as a function of energy above the edge onset. The band structure and total density of states were calculated for α-quartz using a self-consistent pseudopotential method. Projected local densities of Si and O s, p, and d states (LDOS) were calculated and compared with the EELS core-loss edges. These LDOS successfully reproduce the dominant Si and O core-loss edge shapes up to ca. 15 eV above the conduction-band onset. In addition, the calculations provide evidence for considerable charge transfer from Si to O and suggest a marked ionicity of the Si-O bond. The experimental and calculated data indicate that O 2p-Si d π-type bonding is minimal. The low-loss spectra exhibit four peaks that are assigned to transitions from maxima in the valence-band density of states to the conduction band. A band gap of 9.65 eV is measured from the low-loss spectrum. The structures of the surface low-loss spectrum are reproduced by the joint density of states derived from the band-structure calculation. This study provides a detailed description of the unoccupied DOS of α-quartz by comparing the core-loss edges and low-loss spectrum, on a relative energy scale and relating the spectral features to the atom- and angular-momentum-resolved components of a pseudopotential band-structure calculation.

Single-crystal Raman and FTIR methods have been combined to study the structural OH groups of gibbsite, Al(OH) 3 . According to factor group analysis, six unique OH stretching bands [ν(OH)] bands are expected to occur in both IR and Raman spectra. In this study, six ν(OH) bands were observed in both the Raman and IR spectra. Analysis of the gibbsite crystal structure reveals two distinct types of structural OH groups: interlayer and intralayer hydrogen-bonded OH groups. The ν(OH) bands corresponding to these two types of OH groups were clearly resolved using polarized single-crystal Raman spectroscopy. The interlayer hydrogen-bonded OH groups are oriented along the c axis of the crystal and are represented by three ν(OH) bands at 3433, 3370, and 3363 cm -1 . In contrast, the intralayer hydrogen-bonded OH groups are oriented nearly parallel to the (001) face and are represented by the ν(OH) bands at 3623, 3526, and 3519 cm -1 . Assignment of the ν(OH) bands was based, in part, upon the Lippincott and Schroeder one-dimensional (LS-1D) model of the hydrogen bond. Based upon the known geometry of each OH group, the LS-1D model was used to predict the ν(OH) frequencies corresponding to each OH group. Additional support for the band assignments was obtained by correlation between the single-crystal Raman band intensities and the OH bond orientations obtained from the crystal structure.

The crystal structure and hydrogen positions of serandite, NaMn 2 [Si 3 O 8 (OH)]. have been refined from single-crystal X-ray and time-of-flight neutron diffraction data at ambient conditions. The proton occupies an asymmetric, double-well position between O3 and O4. confirming one of the shortest asymmetric hydrogen bonds known in minerals with d(03.. .04) = 2.464(1) Å (X-ray) and 2.467(1) Å (neutron). The proton position closest to 03 has about 84% occupancy and an O-H distance of 1.078(3) Å. and the position closest to 04 has an occupancy of 16% and an O-H distance of 1.07(1) Å. The d(H.. .0) of these hydrogen bonds is 1.413(3) Å and 1.41(1) Å. respectively Hydrogen bond angles are 164° for H1 and 168° for H2. The Si IV -OH bond length [1.628(1) Å] is intermediate in length among the three other Si-O bonds in the dominantly (84%) hydrated Si1 tetrahedron. These new structure data for a very strong hydrogen bond may be useful for extending spectroscopy-structure correlation diagrams into the region of very low energy O-H stretching.

The substitution mechanisms of Fe and Ti have been determined in phlogopite megacrysts from an ultramafic lamprophyre dyke from the Okenyenya igneous complex, northwestern Namibia. Mica separates were heat-treated from 800 to 900 °C. 1 atm to 10 kbar. and f H2 from that of the IQF solid- state buffer to that of ah. Iron oxidation states and H 2 O contents of the ran products were determined using 57 Fe Mossbauer spectroscopy and a vacuum fusion, U-fumace manometry system, respec­tively. The least-squares fit between the univalent anion content (OH + F) and molar Fe 3+ atoms per formula unit (apfu) has a negative slope with a high correlation coefficient and. at the 95% confi­dence level, is consistent with the Fe-oxy reaction, Fe 2+ + OH - = Fe 3+ + O 2- + 1/2 H 2 By adding O 2- to the univalent anion content in 2:1 molar proportions to the Ti. the total anion content in the OH site of the natural phlogopite is. at the 95% confidence level, close to the theoreti­cal value of 4.0 (O = 24 apfu) for the mica structure. It is proposed that the total H deficiency in the natural phlogopite can be explained by both Fe- and Ti-oxy substitution mechanisms. Principal Com­ponents Analysis carried out on exchange components for the experimentally treated phlogopite confirms the operation of the oxy-substitution mechanisms. Both oxy-substitutions dominate in the compositions of natural igneous micas from a variety of tectonic environments. The dehydrogenation of Fe oxy-components in micas from silicic lavas may be a source of water that can be liberated into crustal melts and play an important role in the mecha­nism for initiating volcanic eruptions.

Synchrotron X-ray powder diffraction study of brucite, Mg(OH) 2 , was carried out in a diamond anvil cell with an imaging plate detector from 0.6 to 18.0 GPa at room temperature using the angular- dispersive technique on beamline BL-18C at the Photon Factory, KEK, Japan. Using Rietveld analysis, unit-cell parameters as well as atomic positions of the O atoms in brucite have been successfully refined, taking into account the effects of preferred orientation. Variation of the c/a ratio with pressure indicates that the compression mechanism changes around 10 GPa, above which the compression behavior is isotropic. Based on the changes of the refined atomic positions of the O atoms with pressure, we conclude that the shortening of the interlayer distance controls compression below 10 GPa, whereas above this pressure compression of the oxygen sublattice is the dominant mechanism. Results of the structural refinements also suggest that the MgO 6 octahedral regularity initially approaches a regular configuration with pressure, which then remains unchanged above 10 GPa.

In-situ high pressure X-ray diffraction of phase E (free of secondary phases) measured up to 14.5 GPa shows that phase E is stable over this pressure range at room temperature. The pressure depen­dence of the lattice parameters are a = 2.967 - 0.011P + 0.0001P 2 andc = 13.886-0.054P + 0.001P 2 (P is in GPa). A least-squares fit to third-order of Eulerian strain theory yields a bulk modulus K′ T0 for phase E of 93 (±4) GPa and pressure derivative K′ T0 of 5 (±1). The bulk modulus obtained by this study is about 10% lower than that obtained by Brillouin scattering. Phase E appears to have the lowest bulk modulus among DMHS. The OH stretching frequency of 3613 cm -1 indicates weak, if any, hydrogen bonding. The associated O-O distance of phase E is estimated to be 3.00-3.10 Å.

High-pressure crystal structures are reported for two synthetic wadsleyite crystals, β-Mg 2 SiO 4 (Fe00) and P-(Mg 0.75 Fe 0.25 ) 2 SiO 4 (Fe25), at six pressures to 10.12 GPa. In both compositions, bulk compressibilities are equal to the average compressibility of divalent cation octahedra. Individual silicate tetrahedra, by contrast, are relatively rigid, though the Si-O-Si angle between tetrahedra in Si 2 O 7 dimers decreases systematically with pressure. Wadsleyites display anisotropic compression, with the c axis approximately 40% more compressible than a or b. This behavior results from differ­ential compression of (Mg, Fe)-O bonds; in each of the structure’s three symmetrically independent octahedra, the longest and most compressible bonds are roughly parallel to the b axis. Although the linear compressibilities of FeOO and Fe25 are similar, details of structural changes with pressure differ. Iron-enriched M1 and M3 octahedral sites in Fe25 are significantly less compressible than analogous Mg sites in Fe00.

The crystal structure of Fe 3+ -wadsleyite, (Fe 1.67 2+ Fe 0.33 3+ )(Fe 0.33 3+ Si 0.67 )O 4 , was determined by single­crystal X-ray techniques at six pressures to 8.95 GPa. The isothermal bulk modulus is K T0 = 173(3) GPa [K′ T0 = ∂K T /∂P = 5.2(9)], which is identical within error to bulk moduli observed for normal wadsleyites [β-(Mg 5 Fe) 2 SiO 4 ]. Compression of Fe 3+ -wadsleyite is significantly more isotropic than for β-(Mg 5 Fe) 2 SiO 4 because Fe 3+ substitutes into both Si 4+ tetrahedral sites and (Mg 5 Fe 2+ ) octahedral sites. Ferric iron thus reduces the contrast between tetrahedral and octahedral compressibilities, which in turn reduces the compressional anisotropy. Bond distance analysis and octahedral compressibilities of the three symmetrically distinct octahedral sites reveal that Fe 3+ orders prefer­entially into M1 and M3, while M2 occupancy is close to pure Fe 2+ .

The pressure and compositional dependence of element partitioning were examined in the pseudo- ternary section anorthite-titanite-perovskite of the system CaO-Al 2 O 3 -TiO 2 -SiO 2 . Preliminaiy phase relations in that section were experimentally determined at pressures from 1 atm to 1.5 GPa and at temperatures from 1300 to 1500 °C. using a double-ellipsoid minor furnace and a piston cylinder apparatus. Further experiments were earned out with trace element extended compositions. All run products represent near-equilibrium parageneses consisting of crystals with their coexisting melt: the pairing is desired for element partitioning studies. At pressures above 1.0 GPa a new Ti-rich compound, having the ideal formula Ca 3 TiSi 2 (ALTi-Si) 3 O 14 , appears as the dominant phase in this paragenesis. The composition and homogeneity of the phases were checked with electron microprobe and scanning electron microscopy. Determination of the structure of this compound was performed with a conventional four circle diffractometer on a single crystal, cut from a larger specimen of the equi­librium paragenesis. The compound crystallizes in space group P321 with cell dimensions a = 7.943(1) Å and c = 4.930(1) Å. Its structure consists of layers stacked parallel to (001) such that sheets of two types of comer-linked tetrahedra (T1 = Si; T2 = Al, Ti, Si) at z = ½ alternate with sheets centered at z = 0 containing [6] Ti- and irregular eightfold-coordinated Ca-sites. A statistical distribution for Al, Si, and Ti atoms at one of the two tetrahedral sites in the ciystal structure was confirmed by X-ray diffraction data. The Ti-octahedron is distorted, with Ti-O = 1.952(3) Å and O-Ti-O bond angles between 85.5° and 102.4°. which results in an angle bend of the vertex O atoms of 167.6°.

Carmichaelite [MO 2-x (OH) x , where M includes Ti, Cr, Fe, Mg, and Al] is a new hydrous titanate from Garnet Ridge, an ultramafic diatreme on the Colorado Plateau in Arizona. It is named in honor of Ian S.E. Carmichael, Professor of Geology at the University of California, Berkeley, in recognition of his many contributions to petrology, especially his studies of Fe-Ti oxides in volcanic rocks. The new mineral occurs as platy, cinnamon-brown inclusions in mantle-derived pyrope crystals, which also contain inclusions of rutile, srilankite, ilmenite, minerals of the crichtonite group, spinel, and olivine. Carmichaelite is commonly in contact with rutile and occasionally with srilankite. The average composition is TiO 2 = 62.16, Cr 2 O 3 = 18.43, Al 2 O 3 = 1.88, V 2 O 3 = 0.87, Nb 2 O 5 = 0.37, FeO = 7.61, MgO = 2.80, H 2 O (calc) = 5.76, total 99.89 wt%, giving an empirical formula close to Ti 0.62 Cr 0.19 Fe 0.09 Mg 0.06 Al 0.03 V 0.01 O 1.5 (OH) 0.5 . A cation (non-H) to (O+OH) ratio of 1:2 was confirmed by microprobe analysis of oxygen, and the presence of the hydroxyl component was supported by IR data. Carmichaelite is monoclinic, space group P2 1 /c, with a = 7.706(1), b = 4.5583(6), c = 20.187(3) Å, β = 92.334(2)°, V = 708.5(3) Å 3 , and Z = 22. The calculated density is 4.13 g/cm 3 . The strongest diffraction lines [d(I, h k l)] from a calculated powder pattern are 2.842 (100, 1̅15), 3.773 (94, 013), 2.664 (70, 213), 1.688 (54, 1̅22), 1.679 (44, 226), 1.661 (44, 1̅28), and 1.648 (34, 1.1.11). The crystal structure, which has been determined from single-crystal X-ray diffraction data, consists of stacked chains of cation octahedra within layers of hexagonal closest-packed anions. The arrangement of cation octahedra within a single chain has the same basic structural unit of five edge-sharing octahedra found in olivine, the humite group, and leucophoenicite. Carmichaelite is inferred to have co-precipitated with its pyrope host and other titanates in the presence of a fluid or melt phase (i.e., mantle metasomatism) at temperatures of 650-750 °C and pressures of 20-25 kbar. It contains the highest OH content reported for titanates of mantle origin and offers a potential storage site for water and high field strength elements in the upper mantle. The new mineral is probably stabilized by high pressure and the high concentration of Cr. Certain features of the carmichaelite structure raise the possibility of a similar high-pressure phase in the Si-Al- O-H system.

The structure of parsonsite, Pb 2 [(UO 2 )(PO 4 ) 2 ], Z= 2, triclinic, space group P1̅, a = 6.842(4), b = 10.383(6), c = 6.670(4) Å, α = 101.26(7)°, (β = 98.17(7)°, γ = 86.38(7)°, V= 459.8(7) Å 3 , has been solved by direct methods and refined to R = 6.0 and a goodness-of-fit (5) of 0.92 using 1187 unique observed reflections (|F|≥4σ F ) collected with MoKα X-rays and a CCD-based detector. The single unique U 6+ cation is present as a (UO 2 ) 2+ uranyl ion (Ur) and is coordinated by five additional atoms of O arranged at the equatorial corners of a pentagonal bipyramid capped by the O Ur atoms. Uranyl polyhedra share an edge-forming dimers, wh|ich are cross-linked by edge- and vertex-sharing with two distinct phosphate tetrahedra, resulting in a new uranyl phosphate chain. Two symmetrically distinct Pb 2+ cations are coordinated by nine and six oxygen atoms, and link adjacent uranyl phos­phate chains. Parsonsite is the first uranyl phosphate mineral structure that is based upon chains of polymerized polyhedra of higher bond-valence; others contain sheets that are either based upon the autunite or phosphuranylite anion-topologies.

The crystal structure of pinalite, Pb 3 WO 5 Cl 2 , is orthorhombic Amam, with a = 11.073(2), b = 13.067(3), c = 5.617(1) Å, V = 812.81(3) Å 3 , and Z = 4. It has been refined to an R index of 0.049 on the basis of 593 unique, observed reflections. There are two crystallographically distinct Pb cations with different coordinations: Pb1 has eightfold-coordination with ligands of 4 Cl atoms, 4 O atoms, and one stereoactive lone-pair, whereas Pb2 has ninefold-coordination with ligands of 4 Cl atoms and 5 O atoms. The W cation has fivefold-coordination with 5 O ligands. The pinalite structure is layered parallel to (100) with Pb-O-Cl layers interleaved with Pb-W-O layers. Pinalite belongs to the bismuth oxychloride group of structures. The Pb-O-Cl layer of pinalite duplicates the basic structure of thorikosite and the Pb-W-O layer of pinalite is able to conform to this fundamental cell through differences in the two Pb coordinations and a unique W coordination. Similarly the structures of nadorite and kettnerite are also shown to replicate the basic bismuth oxychloride structure.

The crystal structure of philolithite, Pb 12 O 6 Mn(Mg,Mn) 2 (Mn,Mg) 4 (SO 4 )(CO 3 ) 4 Cl 4 (OH) 12 , P42/nnm, a = 12.627(9), c = 12.595(9) Å, V = 2008(2) Å 3 , Z = 2 has been solved by Patterson difference- Fourier syntheses and refined to R = 0.053 for 814 Fo > 4σ Fo using MoKα X-ray data. In the structure, MnO 6 octahedra form straight chains parallel to [110] and [11̅ 0] by sharing opposite octahedral edges. Octahedra within the chains are further linked by sharing free corners with MnO 4 and SO 4 tetrahedra and CO 3 triangles. The MnO 4 and SO 4 tetrahedra also form bridging struts between octahedral chains, connecting them in the [001] direction into an open framework. The Pb, Cl, and nonframework O atoms occupy the open spaces within the framework. The 10- and 12-fold coordinations of the Pb atoms exhibit the lone-pair effect. Pb atoms link via short bonds to non-framework O atoms to form chains parallel to [110] and [ 1̅10]. When viewed down [111], [ 1̅1̅̅ 1], [1̅1̅̅1], or [11̅1̅̅], [ 1̅11̅] the framework (less the CO 3 groups) is seen to be based upon cubic closest-packing of anions. This open framework of composition [ [6] (Mn,Mg) 12 2+ [4] (SO 4 ) 2 2- [4] (Mn 2+ O 4 ) 6- 2 O 8 (OH) 24 ] 32- , referred to as a closestpacked trellis, is the fundamental unit for the structure.

Reductive dissolution of 7 A bimessite [Mi 2+ 0.05 Mn 3+ 0.25 Mn 4+ 0.7 O 1.7 (OH) 0.25 ] by selenious acid (H 2 SeO 3 ) produces Mn 3+ and Mn 2+ surface reaction products (here represented as S-MnOOH and S-MnO. respectively) and Mn 2+ , 3+ -selenite surface complexes at the solution-mineral interface. Mn2p 3/2 , Se3d. and O1s X-ray photoelectron spectra of reacted surfaces reveal that Mn 4+ of birnessite is reduced simultaneously to Mn 3+ and Mn 2+ while Se 6+ is oxidized to Se 4+ according to the probable stoichiometric reactions: 2MnO 2 + H 2 SeO 3 + H 2 O → 2S-MnOOH + H 2 SeO 4 MnO 2 + H 2 SeO 3 → S-MnO + H 2 SeO 4 There is no XPS evidence for reduction of surface Mn 3+ to Mn 2+ in the presence of selenite. Whereas this reaction proceeds in the presence of arsenite. selenite apparently inhibits reduction of Mn 3+ . perhaps through formation of a strong Mn-selenite surface complex. The rate of release of Mn 2+ to dilute selenious acid (1.5 × 1O -3 M) is considerably lower than the rate of release to aerated, distilled water at similar pH. This behavior suggests that adsorbed selenite complexes impede the proton promoted dissolution of the soluble Mn 2+ component of birnessite.

The structural transformation of high pH Na-rich buserite (NaBu) to H-exchanged hexagonal birnessite (HBi) at low pH was studied by simulation of experimental X-ray diffraction patterns. Four HBi samples were prepared by equilibration of NaBu at constant pH in the range pH 5-2. The samples differ from each other by the presence of one (at pH 2 and 3) or two (at pH 4 and 5) phases, and by the structural heterogeneity of these phases which decreases with decreasing pH. The sample obtained at pH 5 is a 4:1 physical mixture of a 1H phase (a = 4.940 Å, b = a/√3 = 2.852 Å, c = 7.235 Å, β = 90°, γ = 90°) and of a 1M phase (a = 4.940 Å, b = a/√3 = 2.852 Å, c = 7.235 Å, β = 119.2°, γ = 90°) in which successive layers are shifted with respect to each other by +a/3 along the a axis as in chalcophanite. Both the 1H and 1M phases contain very few well-defined stacking faults at pH 5. At pH 4, the sample is a 8:5 physical mixture of a 1H phase containing 15% of monoclinic layer pairs and of a 1M phase containing 40% of orthogonal layer pairs. Any further decrease of the pH leads to the formation of a single defective 1H phase. This 1H phase contains 20% and 5% of monoclinic layer pairs at pH 3 and 2, respectively. Independent of pH, all phases contain 0.833 Mn layer cations, 0.167 vacant layer sites, and 0.167 interlayer Mn cations located either above or below layer vacancies per octahedron. A structural formula is established at each pH. The origin of the observed phase and structural heterogeneities has been analyzed. 1H and 1M phases are assumed to inherit their specific structural and crystal chemical features from the two distinct NaBu modifications. NaBu type I, with a high proportion of Mn 4+ layer cations, is thought to be responsible for the monoclinic layer stacking because this configuration allows Mn cations from adjacent layers to be as far as possible from each other, thus minimizing the electrostatic repulsion between these high charge cations. In contrast, NaBu type II has a high interlayer charge induced by Mn 3+ layer for Mn 4+ layer substitutions. Consequently, the 1H phase has a high amount of interlayer protons and achieves compensation of the unfavorable overlap of layer and interlayer Mn cations, in projection on the ab-plane, by the presence of strong hydrogen bondings between layers. The higher proportion of defined stacking faults in both 1H and 1M phases at pH 4 compared to pH 5 can be attributed to the increase in reaction rate with decreasing pH. At lower pH (3 and 2) the formation of strong hydrogen bonds between adjacent layers controls the layer stacking mode and leads to the formation of a unique 1H phase. The proportion of well-defined stacking faults in this phase decreases from pH 3 to 2.

A new approach for modeling anisotropic peak broadening with Rietveld analysis has been tested on a biologically grown soil greigite, whose morphology (as determined by TEM, Stanjek et al. 1994) had a symmetry lower than cubic. The anisotropic morphology of this greigite determined with single peak fitting was more evident than with the Rietveld analysis in the cubic setting, because in Rietveld analysis {hhh} reflections are treated as one reflection, although two different peak breadths are required: one for modeling two (hhh) reflections perpendicular to the [111] axis of elongation and one for modeling the remaining six (hhh) reflections. For assessing the lower morphological symmetry, the cubic structure was transformed into the tetragonal and into the trigonal system. In both crystal systems, the axis of shape anisotropy is [001]. Due to splitting of peaks, the anisotropic morphology became then measurable. Contributions from other likely sources of anisotropic peak broadening, such as stacking faults and chemical inhomogeneities could be ruled out. The cell-edge length a = 9.868(4) Å agrees with previous values (Stanjek and Murad 1994). The isotropic temperature factors of B (Fe 8a ) = 1.16(6) Å 2 , B (Fe 16d ) = 1.7(2) Å 2 and B (S 32e ) = 1.10(8) Å 2 are much smaller than the overall factor B = 3.2(4) Å 2 determined by Skinner et al. (1964), whereas the anion positional parameter u = 0.2535(3) is identical with 0.2505(38).

Serrabrancaite is a new manganese phosphate from the Alto Serra Branca pegmatite near Pedra Lavrada, Paraiba, Brazil. The mineral occurs as dark brown to dark greenish black isometric crystals up to 0.3 mm size. The crystals have an adamantine luster and are translucent. Chemical analysis yields (in wt%) Mn 2 O 3 = 46.85, P 2 O 5 = 42.72, H 2 O = 9.8, total = 99.37. The empirical formula is Mn 0.98 P 1.00 O 3.98 ·0.90 H 2 O, simplified to MnPO 4 ·H 2 O. Serrabrancaite is monoclinic, space group C2/c, with unit-cell parameters a = 6.914(2), b = 7.468(2), c = 7.364(2) Å, b = 112.29(3)°, V = 351.8(1) Å 3 . The measured density is 3.17(1) g/cm 3 , the calculated density for Z = 4 is 3.16 g/cm3. Serrabrancaite is isostructural with the sulfates of the kieserite group. The mineral is an alteration product from a phosphate pegmatite and is accompanied by vernadite and phosphosiderite.

A X-ray photoelectron Fe 2p 3/2 spectrum of a pristine pyrite fracture surface was collected using synchrotron radiation with the source tuned to 800 eV. Comparison of this highly surface sensitive Fe 2p spectrum with Fe 2p spectra collected by conventional means (1487 eV AlKa source) reveals that the high binding energy tail of the pyrite Fe 2p 3/2 line results primarily from Fe surface state contributions. The three major contributions to the spectrum are interpreted to be: (1) Fe 2+ resident on bulk sites; (2) Fe 2+ resident on surfaces, edges and corners; (3) Fe 3+ surface states produced during fracture by an auto-redox reaction involving Fe and S. The intense main peak is ascribed to the bulk state, whereas the high binding energy tail of the spectrum is composed primarily of Fe 2+ and Fe 3+ surface state contributions. Fe 2+ on bulk sites is octahedrally coordinated (O h symmetry). All valence electrons of Fe on bulk sites are paired (diamagnetic) and a singlet photopeak at 707 eV is consequently produced. Fracture produces Fe 2+ surface states with lower coordination than bulk sites. Fe 2+ located at surfaces, edges and corners experiences modified Ligand Field Stabilization Energies (LFSE) which results in stabilization of the d z 2 orbital and destabilization of the dxy orbital. Promotion of a dxy electron to the d z 2 orbital makes surface Fe 2+ surface states paramagnetic resulting in multiplet splitting of their associated photopeaks. The Fe 3+ surface state is necessarily paramagnetic and its photoemissions are consequently multiply split. Analysis of photopeak structures and binding energy splittings of Fe 2+ and Fe 3+ surface states demonstrates that they are located at the appropriate binding energies, and span the appropriate energy range, to satisfactorily explain the high binding energy tail on of the Fe 2p 3/2 spectrum.

Early exposure cap microdolomites (<50 °C) and low-temperature hydrothermal (<177 °C) calcian dolomite crystals from the Latemar buildup, Dolomites, northern Italy, are microstructurally heterogeneous. Selected-area electron-diffraction patterns contain incommensurate superstructure reflections from domains that are approximately 2.5 nm wide. These domains are quasiperiodic with wavelengths of 5 to 20 nm. The scale of these modulations varies with composition of the bulk crystal. Previously described c-domain superstructures in dolomite are reported as commensurate with the host structure; c-diffraction spots are situated exactly halfway between the principal (a and b) reflections. Re-examination of published diffraction patterns reveals that some are actually incommensurate. Individual c-domains are monoclinic and have been interpreted to be more calcium-rich than the host dolomite. Newly recognized c-domains resemble previously described c-domains but are distinct because they are incommensurate with the dolomite host and produce diffraction spots with irrational indices. In samples from the Latemar buildup, the c-axes of the host and modulations are parallel, but the [110] directions are not parallel. Further, the length of c* dolomite ≈ c* superstructure , but the lattice spacing of d 110 superstructure is more than twice d 110 dolomite . The superstructure phase is metrically monoclinic and is interpreted to incorporate more calcium than dolomite. Domains are, however, too small to analyze directly. Excess calcium may account for the more than doubling of the domains’ α-dimension relative to that of dolomite. Inclination of the domain lattice relative to the host lattice may vary as a function of calcium content.

We present here high-resolution, 19 F NMR spectra, collected with sample spinning rates to 25 kHz, for sodium and calcium silicate and aluminosilicate glasses. Several distinct fluoride ion sites are well resolved and can be assigned to various coordination environments based on clear similarities to crystalline model compounds. In aluminosilicates, fluoride with one Al and several Na or Ca neighbors predominate, but silicon-fluoride groups may also be significant. Small concentrations of the latter can also be detected in sodium silicate glasses, suggesting a possible role in reduction of viscosity. Fluoride sites with no Si neighbors are, however, predominant in Al-free sodium silicates and probably in Ca silicates as well.

A new method of determining the water content of melt inclusions using confocal laser Raman microprobe spectroscopy is described. The water content of melt inclusions can be determined in the concentration range of 0 to 20 wt% with a high spatial resolution (~2 μm). Because the method works in reflection, minimal sample preparation is necessary. The method is fast, has good accuracy and precision (±0.25 wt%), and has the potential to become a useful, high resolution spectroscopic tool for melt inclusion studies.