Volume 89, Issue 2-3

Scapolite is a metamorphic aluminosilicate mineral that can be described by the general formula (Na,Ca,K) 4 (Al,Si) 6 Si6O 24 (Cl,CO 3 ,SO 4 ). Two common end-members are called marialite (Na 4 ClSi 9 Al 3 O 24 ) and meionite (Ca 4 CO 3 Si 6 Al 6 O 24 ). Variations in scapolite composition can be described by two independent substitutions, NaSi(CaAl) -1 and NaCl(CaCO 3 ) -1 . Twenty eight natural scapolites in the present study exhibit a range of compositions from X EqAn [(Al-3)/3] = 8% and X Me [Ca/(Na+K+Ca)] = 7% to X EqAn = 82% and X Me = 90%. Several coupled exchange reactions can be identified in some inhomogeneous samples (e.g., Na 1.49 SiCl 0.47 [Ca 1.44 Al(CO 3 ) 0.43 ] -1 , Na 1.69 SiCl 0.58 [Ca 1.55 Al(CO 3 ) 0.50 ] -1 , Na 1.91 SiCl 0.79 [Ca 1.75 Al(CO 3 ) 0.69 ] -1 ). The extent of coupling between the two substitutions is controlled by the crystallization environment (P, T, and mineral assemblages). Electron diffraction patterns suggest that the symmetry of scapolite with X Me up to 18% is I4/m, whereas that for intermediate scapolite from X Me = 18% to at least X Me = 90% is P4 2 /n. Under darkfield observation (g = hkl, h + k + l = odd) using a transmission electron microscope (TEM), the P4 2 /n samples have anti phase domains of various sizes, the presence of which provides evidence for an I-P phase transition. A wide compositional range of scapolite solid solutions should have an I4/m symmetry at the time of formation.

Görgeyite, K 2 Ca 5 (SO 4 ) 6 ·H 2 O, is a very rare monoclinic double salt found in evaporites related to the slightly more common mineral syngenite. At 1 atmosphere with increasing external temperature from 25 to 150 °C, the following succession of minerals was formed: first gypsum and K 2 O, followed at 100 °C by görgeyite. Changes in concentration at 150 °C due to evaporation resulted in the formation of syngenite and finally arcanite. Under hydrothermal conditions, the succession is syngenite at 50 °C, followed by görgyeite at 100 and 150 °C. Increasing the synthesis time at 100 °C and 1 atmosphere showed that initially gypsum was formed, later being replaced by görgeyite. Finally görgeyite was replaced by syngenite, indicating that görgeyite is a metastable phase under these conditions. Under hydrothermal conditions, syngenite plus a small amount of gypsum was formed, after two days being replaced by görgeyite. No further changes were observed with increasing time. Pure görgeyite showed elongated crystals approximately 500 to 1000 μm in length. The infrared and Raman spectra are mainly showing the vibrational modes of the sulfate groups and the crystal water (structural water). Water is characterized by OH-stretching modes at 3526 and 3577 cm -1 , OH-bending modes at 1615 and 1647 cm -1 , and an OH-libration mode at 876 cm -1 . The sulfate ν 1 mode is weak in the infrared but showed strong bands at 1005 and 1013 cm -1 in the Raman spectrum. The ν 2 mode also showed strong bands in the Raman spectrum at 433, 440, 457, and 480 cm -1 . The ν 3 mode is characterized by a complex set of bands in both infrared and Raman spectra around 1150 cm -1 , whereas ν 4 is found at 650 cm -1 .

High-pressure experiments were conducted to investigate the phase relations in the binary system Fe-FeSi at a pressure of 21 GPa and temperatures between 1350 and 2200 °C by using a multi-anvil apparatus. The system has a eutectic point at 26 wt% Si and 1820 °C, which is 400 °C lower than the melting point of pure Fe. On the Fe-rich side of the eutectic point, the temperature difference between the solidus and liquidus curves is less than 50 °C and the compositional difference is small between the coexisting solid and liquid. At subsolidus temperatures, homogeneous Fe-Si alloys were quenched with compositions up to 21.7 wt% Si. The more Si-enriched composition, Fe(+25.1 wt% Si), was found to coexist to with CsCl-type FeSi. These results suggest that a large amount of Si could have dissolved into liquid Fe at high pressures during core formation if it occurred under reducing conditions. However, the difference in Si content between the outer core and the inner core would be very small, if the solubility of Si in solid Fe remains large at inner core pressures.

We report the first systematic study of H 2 O speciation in dacitic melt with total dissolved H 2 O ranging from 1.4 to 2.6 wt%. Samples were heated in a one-atmosphere tube furnace at 480-590 °C for certain durations and then quenched in water or air. Species concentrations were measured with Fourier transform infrared (FTIR) spectroscopy. At a given temperature, the equilibrium constant for the reaction, H 2 O m (melt) + O (melt) = 2OH (melt), is roughly constant. At the same temperature and water content, the equilibrium constant for the hydrous species interconversion reaction in dacitic melt appears to be slightly greater than in hydrous rhyolitic melt, reflecting a larger proportion of dissolved H 2 O as hydroxyl groups in dacitic melt. Our results can be applied to model species equilibrium in hydrous dacitic melt, to estimate the apparent equilibrium temperature of natural dacitic glasses, and to investigate viscosity of hydrous dacitic melts.

The kinetics of the α-β transformation in Mg 2 SiO 4 was studied by in situ synchrotron powder X-ray diffraction (XRD) at nine pressure-temperature conditions in the ranges 13.4-15.8 GPa and 850-1100 °C. The transformation from olivine occurred by grain-boundary nucleation and interface-controlled growth mechanisms. Infrared analysis of the recovered samples indicates that a small amount of water, 750 ± 100 ppm by weight, was present in the samples although the experiments were carried out under nominally dry conditions. Nucleation and growth rates were determined by fitting the rate equation for the grain-boundary nucleated transformation to the kinetic data. The activation energy and activation volume for growth were estimated to be 348 (137) kJ/mol and 1.7 (4.5) cm 3 /mol, respectively. The growth kinetics determined in this study bear upon the field of metastable olivine in the subducting slab as the water contents of samples studied are well known. Nucleation rates were estimated to be relatively large, even at small overpressure conditions, which is consistent with the small activation energy for nucleation derived in this study.

The pressure-temperature equilibrium curve of the reaction siderite + hematite = magnetite + CO 2 was determined in the range 5-12 kbar and 480-650 °C by piston-cylinder experiments, with NaCl as a pressure medium. Silver oxalate was used as a CO 2 source and samples were buffered at hematitemagnetite oxygen fugacity. Reaction progress was monitored by extent of CO 2 gas loss and by X-ray diffraction (XRD) analysis. The data define a univariant curve, which is described by P = -14.599 + 0.025T + 0.000027 T 2 with P in kbar and T in °C. Calculations based on these data give ΔH 0 f (298K) siderite = -760.6 ± 0.9 kJ (kilojoules) from the oxides. The formation of siderite requires a specific range of ambient oxygen and carbon dioxide fugacities, dependent upon temperature and pressure. The stable assemblage of siderite and magnetite, at a given temperature and pressure, implies more restrictive ranges of oxygen and carbon dioxide fugacities, defined by reactions among siderite, magnetite, graphite, and hematite. Experimental and thermodynamic investigation of the Fe-C-O system indicates that the formation of magnetite along with Ca-Fe-Mg carbonate globules by inorganic processes is possible and may be relevant to Martian meteorite ALH84001. Decarbonation of the siderite component of the carbonate, either by a transient heating event or by a change in oxygen fugacity of a coexisting fluid, may have formed the observed grains, although this study does not address the size or morphology of magnetite grains formed by this mechanism.

A new calibration was performed for the molecular CO 2 band at 2349 cm -1 in the infrared absorption spectra of rhyolitic glasses. Glasses with varying amounts of CO 2 (730-3900 ppm by weight) and H 2 O (0.5-7.0 wt%) were synthesized in an internally heated pressure vessel at 200-800 MPa and 1100 and 1200 °C. The CO 2 content of the glasses was measured by coulometric CO 2 -titration after thermal extraction at 1200 °C. It is shown that the entire CO 2 content cannot be extracted from a rhyolitic glass at this temperature. Using the Lambert-Beer law, and taking into account the residual CO 2 still present in the glasses after extraction, we have calculated a linear molar absorption coefficient of 1214 ± 78 L⋅cm -1 /mol for the band at 2349 cm -1 . Strictly, this value is a practical absorption coefficient because the IR band intensity is related to the total CO 2 content and not to the molecular CO 2 content. However, no direct evidence for other carbon species such as carbonate groups is found in the IR spectra. Hence, we suggest that the concentration of carbonate in the studied metaluminous rhyolite is negligible and the derived value is close to the real absorption coefficient for molecular CO 2 . The absorption coefficient does not vary noticeably with changing water content of the glasses. With the new calibration, CO 2 concentrations in rhyolitic glasses are 13% higher than data based on the previous calibration of Blank (1993) for water-poor glasses.

A crystalline form of hydrated calcium phosphate, brushite (CaHPO 4 ⋅2H 2 O), has been studied by means of total energy First Principles calculations based on the Density Functional Theory (DFT) approximation. The experimental crystal lattice parameters of this mineral have been reproduced with good agreement. The powder X-ray diffraction pattern simulated from the calculated crystal structure is similar to the experimental one. The intermolecular interactions within the crystal, electrostatic and hydrogen bond interactions, have been reproduced. The calculated crystal structure reproduces the atomic distribution along the (010) surface observed by high resolution scanning force microscope (SFM). The optimized crystal structure describes the different interatomic interactions along the [101], [201], and [001] directions on the (010) plane. Quantum-mechanical calculations of the energy of these surfaces confirm the experimental behavior, justifying the different crystal growth rates of these directions found experimentally.

The crystal structure of jonesite, Ba 2 (K,Na)[Ti 2 (Si 5 Al)O 18 (H 2 O)](H 2 O) n (monoclinic, P2 1 /m, a = 10.618(2), b = 25.918(4), c = 8.6945(14) Å, β = 127.633(3)°, V = 1894.8(6) Å 3 , Z = 4) from the Benitoite Gem Mine, San Benito County, California has been solved by direct methods from a crystal twinned on (001) and refined to R 1 = 0.045 (wR = 0.119, S = 1.028) using 3308 unique observed reflections (|F o | ≥ 4σ F ). The structure is based upon porous double layers of distorted TiΦ 6 octahedra (Φ = O, H 2 O) and TO 4 tetrahedra (T = Si, Al) parallel to (010). The layers consist of two sheets of corner-sharing TiΦ 6 octahedra and Si 2 O 7 groups each. The two adjacent sheets are linked along b by T 4 O 12 tetrahedral rings that are disordered over two positions. The double layer has an open structure characterized by eight-membered tetrahedral rings with apertures (free diameters) of 3.37 × 3.37 and 3.33 × 3.33 Å 2 , for two symmetrically non-equivalent rings. K + cations and H 2 O molecules are located in the pores of the double layers. Ba 2+ cations are between the double layers and provide their linkage into three-dimensional structure. Jonesite is a first example of titanosilicate with porous double-layer structure.

The crystal-chemistry of fifteen samples of talc from various localities and origins having a relatively wide range of Fe, Al, and F contents, have been studied mainly by Fourier-transform infrared (FTIR) spectroscopy, in both the near infrared (NIR) and middle infrared (MIR) regions, and by Mössbauer spectroscopy. For the first time, assignments have been made for the 2νOH bands in talc, and Mg/Al substitutions were revealed using NIR. Less than 0.01 atom of Fe 2+ or Al per half unit cell can be detected from the 2νOH region, whereas the sensitivity is about half in the νOH region. The amounts of F also can be quantified by NIR measurement, using the F inductive effect on the width of the 2νMg 3 OH band. The NIR data show that Fe 2+ and Mg are distributed randomly in their octahedral sheet. Actual structural formulae of natural talcs can be obtained from NIR and Mössbauer spectroscopies only, even for impure talc samples. NIR spectroscopy is simple to use and is a very powerful tool to study the crystal chemistry of talc.

Beryllium is a significant constituent in sapphirine in some metamorphic and pegmatitic rocks, and thus could have a major effect on its stability relationships. Using the stoichiometries of reactions involving sapphirine and associated phases in the MgO-BeO-Al 2 O 3 -SiO 2 (MBeAS) system in conjunction with molar volume data, we have plotted maps of the sapphirine solid-solution field in both μ-μ and μ-P space, where μ is the chemical potential of an exchange component such as (BeSi)(AlAl)-1. These maps give a pressure sequence of stable MBeAS univariant reactions and divariant assemblages that are consistent with experimental data, e.g., they show how Be stabilizes sapphirine + forsterite, which is rare in nature but readily synthesized over a wide P-T range in the presence of Be. We generate a MBeAS petrogenetic grid for sapphirine-bearing assemblages over the approximate range T = 700-900 °C, P = 0-2.5 GPa, identify divariant and univariant assemblages containing sapphirine with maximum Be, and determine the sense of variation of maximum Be content with P. At lower T, maximum Be occurs at the low-P limit of surinamite stability, ca. 0.5 GPa. At higher T, maximum Be increases with P, following the MBeAS univariant reactions involving (sapphirine + surinamite + orthopyroxene + chrysoberyl + forsterite or spinel). Natural assemblages containing sapphirine and its Be-rich near-analog khmaralite from the Napier Complex, Enderby Land, East Antarctica formed at higher T (900-1100 °C) than the experiments and in bulk compositions containing substantial Fe. Associated minerals include garnet, sillimanite, quartz, and magnesiotaaffeite-6N'3S (“musgravite”), whereas forsterite is absent and cordierite is a local, late phase. μ (BeSi)(AlAl)-1 -μ FeMg-1 diagrams show that the stability of magnesiotaaffeite-6N'3S causes the maximally beryllian khmaralite to shift from a magnesian composition in equilibrium with orthopyroxene + surinamite + forsterite + chrysoberyl, as in the MBeAS subsystem, to a more Fe-rich composition associated with garnet + surinamite + magnesiotaaffeite-6N'3S + chrysoberyl. Khmaralite associated with sillimanite + garnet + surinamite + magnesiotaaffeite-6N'3S or chrysoberyl in a Napier Complex pegmatite from Khmara Bay is predicted to be the most Be-rich possible in the presence of sillimanite, whereas the sillimanite + quartz ± orthopyroxene ± garnet associations in quartz granulites requires a sapphirine much lower in both Be and Fe; analyses are roughly in accord with these predictions. The shape of the sapphirine/khmaralite solid-solution field is such that there is a positive correlation between high Be and high Fe 2+ , a chemographic effect independent of any crystal chemical effects due to the clustering of Fe and Be in the crystal structure of khmaralite. The diagram for FMBeAS shows that sapphirine + quartz, which is often cited as evidence for ultrahigh temperatures (e.g., ≥ 1040 °C), is stabilized to lower T and higher P than in the corresponding Be-free system. Hence, this minimum T may be valid only in rocks with relatively abundant sapphirine and/or very low bulk Be content so that what Be is present in the system is not concentrated in sapphirine.

Spriggite, Pb 3 [(UO 2 ) 6 O 8 (OH) 2 ](H 2 O) 3 , is a new hydrated Pb uranyl oxyhydroxide found near Arkaroola, Northern Flinders Ranges, South Australia. The new mineralʼs name honors geologist and conservationist Reginald Claude Sprigg (1919-1994), founder of the Arkaroola Tourist Station. Together with beta-uranophane, soddyite, kasolite, Ce-rich francoisite-(Nd), metatorbernite, billietite, Ba-bearing boltwoodite, schoepite, metaschoepite, and weeksite, spriggite results from the supergene alteration of U-Nb-REE-bearing hydrothermal hematite breccia. Spriggite forms prismatic crystals up to about 150 μm in length and up to 40 μm across. It is transparent, bright orange in color with vitreous luster, biaxial, n min = 1.807; n max = 1.891 (NaD, 22.5 °C), non-fluorescent, brittle with an uneven fracture. It has a pale orange streak, Mohsʼ hardness ~4, good cleavage along (100) and D calc = 7.64(6) g/ cm 3 . The empirical formula is (Pb 2.77 Ca 0.06 Ba 0.04 ) Σ2.87 U 6 O 19.9 (OH) 2 ⋅3H 2 O, and the simplified formula is Pb 3 [(UO 2 ) 6 O 8 (OH) 2 ](H 2 O) 3 . Spriggite is monoclinic, C2/c, a = 28.355(9), b = 11.990(4), c = 13.998(4) Å, β = 104.248(5); V = 4613(3) Å 3 , Z = 8. The strongest eight lines in the powder X-ray diffraction pattern are [d in Å (I)(hkl)]: 6.92(60)(400), 6.02(30)(112̅;020), 3.46(80)(800), 3.10(100)(204;6̅04;33̅2;5̅32), 2̅.74(30)(4̅40), 2.01(30)(336̅), 1.918(60)(10.04̅;14.04̅;11.3̅2;1̅3̅-.31), 1.738(30)(53̅6;1̅1̅.36). The structure has been solved from a crystal twinned on (001) and refined to R1 = 9.7%. The structure is based upon the [(UO 2 ) 6 O 8 (OH) 2 ] 6+ sheets of uranyl polyhedra of the β-U 3 O 8 anion topology with Pb 2+ cations and H 2 O groups in the interlayer. Billietite and spriggite contain only hexavalent U in the uranyl sheets, whereas the similar sheets in β-U 3 O 8 contain U 5+ and U 6+ , and those in ianthinite U 4+ and U 6+ . Spriggite has the highest Pb:U ratio among the known hydrated Pb uranyl oxyhydroxide minerals.

The dacite of the Pebble Creek Formation (Mount Meager, BC, Canada) is an extraordinary occurrence of lavas containing coarse, sieve-textured plagioclase phenocrysts that appear to have reacted extensively with the melt. We record this unusual occurrence of a naturally reacted crystal/melt pair utilizing Nomarski Differential Interference Contrast (NDIC) imaging and electron microprobe data. Hieroglyphic textures are common in plagioclase of this study, and appear to represent older plagioclase remnants in sharp contact with new plagioclase forming a palimpsest texture outlining the occurrence of previous pools of melt. Throughout the processes affecting the phenocrysts, capillaries of liquid remained open to the external melt.

Raman and infrared active modes of cerussite were assigned according to their symmetry species and compared to other aragonite group minerals. Small satellite bands at 823 and 1031 cm -1 on the low-wavenumber side of the fundamental vibrations ν 2 and ν 1 , respectively, have been assigned to the isotopic substitutions of 13 C and 18 O. The Raman active ν 1 and ν 2 carbonate modes are observed at 1051 and 835 cm -1 . The absence of the B 2g component of the ν 1 and ν 2 vibrations has been explained by the small coupling between the A g and B 2g modes. The Raman active ν 3 carbonate anti-symmetric stretching mode is observed at 1361(A g ), 1376 (B 1g ), 1419 (B 3g ), and 1477 (B 2g ) cm -1 , while the corresponding infrared active bands are observed at 1396, 1432, and 1456 cm -1 . The Raman active ν 4 carbonate bending mode is observed at 673 (A g ), 668 (B 2g ), 681 (B 1g ), and 694 (B 2g ) cm -1 . The corresponding infrared bands are observed at 670, 679, and 698 cm -1 . In both ν 3 and ν 4 the factor group splitting between the B 1g and B 3g modes is 1 to 3 times smaller than the separation of the A g and B 2g modes. Raman active lattice vibrations are detected at 120 (B 3g ), 132 (A g ), 148 (B 1g ), 152(B 2g ), 174 (B 2g ), 179 (B 1g ), 213 (A g ), 226 (B 3g ), and 243 cm -1 (B 2g ). Corresponding infrared active bands are detected at 573, 543, 573, 423, 375, 290, 205, 165, 146, and 134 cm -1 . Raman bands at 949, 966, 989, 1000, and 1104 cm -1 and at 922, 946, 967, 988, 996, and 1007 cm -1 in the infrared spectra are assigned to combination and overtone bands. Raman bands at 1676 (A g ), 1689 (A g ), 1730 (B 3g ), and 1740 (B 1g ) cm -1 are ascribed to combination modes of ν 1 + ν 4 with bands at 2052 and 2092 cm -1 assigned to 2ν 1 . Corresponding infrared bands are observed at 1729 and 1740 cm -1 (ν 1 + ν 3 ). Bands at 2359, 2409, 2471, and 2521 cm -1 are ascribed to ν 1 + ν 3 , with broad bands at 1246 and 1323 cm -1 assigned to 2ν 4 modes.

The structural behavior of sodalite, ideally Na 8 [Al 6 Si 6 O 24 ]Cl 2 , at room pressure and from 28 to 982 °C on heating, was determined by using in situ synchrotron X-ray powder diffraction data (λ = 0.92007(4) Å) and Rietveld refinement. The sample was heated at a rate of about 9.5 °C/min and X-ray spectra were collected at intervals of about 15 °C. The cubic unit-cell parameter for sodalite increases smoothly and non-linearly to 982 °C. The percent volume change between 28 and 982 °C is 4.8(2)%. Between 28 and 982 °C, the Al-O and Si-O distances are constant, while the Al-O-Si angle increases from 138.29(1) to 146.35(2)° by 5.06(2)°. Simultaneously, the angle of rotation of the AlO 4 tetrahedron, φ Al , decreases from 22.1 to 16.9°, a difference of 5.2°, while the angle of rotation of the SiO 4 tetrahedron, φ Si , decreases from 23.6 to 18.0°, a difference of 5.6°. Moreover, the [Na 4 Cl] 3+ clusters expand with increases in the Na-Cl bond length by 0.182(4) Å, and corresponding increases in the short Na-O bond length by 0.093(2) Å, and decreases in the longer Na-O* distance by 0.108(1) Å. Large displacement parameters occur for the Na and Cl atoms, and as the weaker Na-Cl bond expands with temperature, the Na atoms move toward the plane of the framework six-membered rings, which causes the framework tetrahedra to rotate and results in a relatively high rate of expansion of the structure. The framework TO 4 tetrahedra distort slightly with temperature. If the Na atom reaches approximately the plane of the six-membered ring, the expansion will be retarded, but sodalite melts before this occurs. Sodalite melts at about 1079 °C and begins to lose NaCl. The NaCl component is lost in two stages: about 4.5 wt% of NaCl is lost slowly at about 1150 °C, and about 7.0 wt% of NaCl is lost at a faster rate at about 1284 °C.

Metahohmannite, Fe 3+ 2 [O(SO 4 ) 2 ]⋅4H 2 O, is a hydrated sulfate of ferric iron that occurs in sulfate deposits in the desert areas of Northern Chile. The compound used for this study was obtained as a dehydration product of hohmannite, Fe 3+ 2 [O(SO 4 ) 2 ]⋅(4+4)H 2 O. Intensities for the structure analysis were collected from a powdered sample using in situ synchrotron X-ray powder diffraction at ESRF (Grenoble, France). The structure was solved ab initio by profile deconvolution and the application of standard Patterson and difference Fourier maps. The structure was refined to R p = 5.46% using the Rietveld method. Metahohmannite crystallizes in the triclinic system, space group P1̅ with unit-cell parameters a = 7.3484(5) Å, b = 9.7710(6) Å, c = 7.1521(5) Å, α = 91.684(5)°, β = 98.523 (5)°, γ = 86.390 (5)°, and Z = 2. The structure consists of four Fe 3+ octahedra and four sulfate tetrahedra, which share vertices and edges to form a complex building block of Fe 3+ 2 [O(SO 4 ) 2 ]⋅8H 2 O composition. Such blocks are connected to form chains running parallel to the c axis. A complicated system of hydrogen bonds connects adjacent chains into a three-dimensional network. Finally, the crystal structures of metahohmannite, hohmannite, and amarantite are compared and the geometrical features discussed in detail.

The lattice parameters of natural spessartine and uvarovite garnets were measured as a function of pressure up to 25 and 35 GPa, respectively, by synchrotron radiation powder diffraction. The bulk modulus and its first and second derivatives were determined by fitting three equation of state (EoS) models to the pressure-volume data: the Birch-Murnaghan EoS, the Vinet EoS, and the Poirier- Tarantola logarithmic EoS. The third-order Birch-Murnaghan EoS, assuming V 0 at its experimental value, yields K 0 = 171 (1) GPa and K 0 ʼ = 5.4 (2) with V 0 = 1573 Å 3 for spessartine, and K 0 = 160(1) GPa and K 0 ʼ = 5.8(1) with V 0 = 1697.5 Å 3 for uvarovite. The results obtained by the EoS models are consistent with each other within the experimental uncertainties. A comparison is made with earlier results from the literature.

Tugtupite [Na 8 (Al 2 Be 2 Si 8 O 24 )Cl 2 ] is an extremely ordered mineral containing 27 Al, 23 Na, and 9 Be quadrupolar nuclei. The quadrupolar interaction parameters (QI), C q and η, for these nuclei were calculated from the electronic structure by full-crystal linearized augmented plane-wave (FC LAPW) or ab initio molecular orbital (MO) calculations. The orientation of the EFG tensors, calculated using FC LAPW correspond with the crystallographic axes a1, a2, and c, and η at the Al site is equal to 0.00 as required by the symmetrical electron density distribution in the a-c planes. The electron density of states (DOS) shows the EFG to originate from the valence band of the p-p, p-s electrons. The agreement between the calculated QI values and experimental results for tugtupite shows that FC LAPW is very effective to evaluate the electronic structure of an ordered mineral, and aid in the interpretation of MAS NMR spectra. The combination of computational efficiency and accuracy of ab initio cluster MO calculations were examined by using different cluster sizes, basis sets, theoretical levels, and charge backgrounds using the Gaussian 98W program. A small cluster embedded in a point charge background gave sufficiently accurate 27 Al EFG tensors, with much greater computational efficiency than with WIEN2k. The MO cluster calculation became closer to the experimental results as the number of atoms in the clusters increased, but the calculation cost increased as well. The point-charge background was found to contribute about 10% to the EFG. MO calculations, with B3LYP level and basis set 6-311G+d** for the cluster of two atom shells around the central Al nucleus, gave quadrupole interaction values within 13% of that found by the FC LAPW method using the WIEN2k program. Therefore, MO cluster calculations can provide a more efficient but less accurate alternative to FC LAPW for quantum calculations of mineral structures.

The major contributions to the pyrite valence band have been identified experimentally by monitoring XPS valence band signal intensity as a function of photon (X-ray source) energy. The S-S σ and σ* orbitals are located at about 13 and 16 eV binding energy and are almost exclusively derived from S 3s-S 3s mixing; there is no experimental evidence for a S 3p contribution to these molecular orbital (molecular orbital) values as required for sp 3 hybridization. Fe-S σ (Fe 3d eg -S 3p z derived) and Fe-S π and π* (Fe 3d t2g -S 3p x,y derived) molecular orbitals are identified at Binding Energies (B.E.) less than 8 eV, and are consistent with energies calculated by Eyert et al. (1998). Fe-S π molecular orbital contributions are located in the valence band between about 1.8 and 3 eV and Fe-S π* molecular orbitals contribute between about 0.7 to 1.5 eV binding energy. Contributions observed at the top of the valence band and concentrated between the Fermi level and about 0.6 eV B.E. may be derived from Fe 3d eg (non-bonding) orbitals. The experimental data are consistent with these non-bonding contributions being derived from surface Fe species (Fe electronic surface states), and perhaps represent the Fe 3d eg “dangling bonds” produced by rupture of Fe-S σ bonds during fracture. The conformity of spectral assignments based on experimental data with band theoretical calculations (of Eyert et al. 1998, using a basis set appropriate to pyrite) indicates that the essential aspects of the pyrite valence band are understood and energetically quantified. Additional study is required to identify positively.

Raman spectra were collected for synthetic gypsum (CaSO 4 ⋅2H 2 O) powder between 9 and 373 K under atmospheric pressure with special emphasis on the temperature dependence of the OH-stretching modes. The stretching bands of the water molecules in gypsum were found to shift in opposite directions as a result of the different degree of intermolecular hydrogen-bonding between nonequivalent water H atoms and the O atoms of nearby SO 4 ions. The anharmonic parameters of the OH-stretching modes are calculated using the temperature derivatives measured from the present investigation and existing pressure derivatives. These parameters are -4.7 × 10 -6 K -1 and -0.6 × 10 -6 K -1 for the 3407 and 3494 cm -1 bands, respectively. The dehydration of gypsum into γ-CaSO 4 and the subsequent rehydration of γ-CaSO 4 into hemihydrate are clearly identified in the Raman spectra by the observed variation in Raman shifts of the OH and ν 1 (SO 4 ) bands. The latter increases as the mineral becomes increasingly anhydrous (1007 cm -1 in gypsum; 1014 cm -1 in hemihydrate; 1026 cm -1 in γ-CaSO 4 ), which can be used as a fingerprint for the remote detection of these minerals on planetary surfaces.

A new structure type (L) is reported for disilicates of the middle and heavy rare earth elements (REE) quenched from 10 GPa, 1600-1700 °C. Crystal data are: triclinic, space group P1̅ , Z = 6; Dy 2 Si 2 O 7 : a = 6.5971(3), b = 6.6504(2), c = 18.0582(6) Å, α = 83.791(2), β = 88.653(2), γ = 88.498(2)°, V = 787.2 Å 3 , and D x = 6.242 g/cm 3 ; Tm 2 Si 2 O 7 : a = 6.5499(2), b = 6.5876(2), c = 17.8916(7) Å, α = 83.828(1), β = 88.368(1), γ = 88.152(2)°, V = 766.9 Å 3 , and D x = 6.574 g/cm 3 ; Lu 2 Si 2 O 7 : a = 6.5240(2), b = 6.5553(2), c = 17.7909(6) Å, α = 83.977(2), β = 88.074(2), γ = 87.846(2)°, V = 755.8 Å 3 , and D x = 6.830 g/cm 3 . The type-L structure comprises linear trisilicate [Si 3 O 10 ] 8- and orthosilicate [SiO 4 ] 4- ions cross linked by REE 3+ in one sevenfold and five eightfold coordinated positions (to 3.0 Å), and is assembled from alternating (001) strips of the type-B structure of REE2Si2O7 and a sheared structure containing structural elements found in the type-B structure but with Si distributed with 50% occupancy over two face-sharing tetrahedra. The geometry of these half-occupied tetrahedra is consistent with decomposition of a SiO 5 trigonal bipyramid during quenching of the pressure.

A total of 30 synthetic samples of the Ca 2 Fe 2-x Al x O 5 , 0.00 ≤ x ≤ 1.34 solid solution series have been investigated by single crystal X-ray diffraction at 25 °C. Pure Ca 2 Fe 2 O 5 and samples up to x = 0.56 have space group Pnma, Z = 4, whereas samples with x > 0.56 show I2mb symmetry, Z = 4. The substitution of Fe 3+ by the smaller Al 3+ cation decreases unit-cell parameters and average octahedral and tetrahedral bond lengths and induces distinct changes in the O-atom coordination of the interstitial Ca atom. Discontinuities in the structural parameters vs. the Al 3+ tot content and changes in slope of these quantities are associated with the phase transition. The essential difference between the two modifications is the cation-O atom-cation angle within the planes of corner sharing octahedra, which is close to 180° in I2mb, but ≈184° in the Pnma phase, and the existence of two different orientations of the tetrahedral chains in Pnma as opposed to one in I2mb. At low overall Al 3+ concentrations Al 3+ preferentially enters the tetrahedral site until ≈2/3 of it is filled. Additional Al 3+ cations, substituted for Fe 3+ , are equally distributed over octahedral and tetrahedral sites. At high temperature pure Ca 2 Fe 2 O 5 transforms to a body-centered structure at 724(4) °C. Substituting Al 3+ for Fe 3+ linearly decreases the transition temperature by 15 °C per 0.1 Al 3+ down to 623(5) °C for x = 0.65.

Single-crystal X-ray data were collected from a natural fibrous mordenite crystal of composition K 2.99 Ca 1.85 Na 1.06 Al 7.89 Si 40.15 O 96 ⋅28H 2 O and from a platy synthetic mordenite crystal of composition Na 6 Al 6.02 Si 42.02 O 96 ⋅19H 2 O. Diffraction data were measured with a point detector using a sealed X-ray tube and an image plate using synchrotron radiation, respectively. Both structures exhibit the same defect features visible in difference-Fourier maps. Domains of the entire Cmcm framework structure are reproduced by a non-crystallographic (001) mirror plane at z = 0 and z = 1/2. An identical description is a shift of framework domains 1/2 along the c axis. The concentration of this defect domain is 2.7(2) and 3.1(1)% for the natural and synthetic mordenite crystals, respectively. Reproductions of reciprocal layers from synchrotron image-plate data reveal diffuse scattering for hkl layers with l = 2n + 1. The diffuseness of these layers is not homogeneous but concentrates in the form of halos around selected reflections allowed for C-centering. Diffuse features in electron diffraction patterns of natural and synthetic mordenite have been described before and were interpreted either as evidence of c/2 faults or intergrowth with different mordenite-related structure-types. We have modeled a (100) defect layer that is modified from the mordenite characteristic puckered sheet of six-membered rings and allows coherent intergrowth of identical structural subunits shifted by c/2. These defect domains do not influence or obstruct the 12-membered ring-channels characteristic of this zeolite. The major difference in Si,Al distribution between the two samples is that the natural crystal has Al strongly enriched at T3, which is part of the four-membered rings. We suggest that a synergetic effect between extraframework cations and Si, Al ordering during crystal growth is responsible for Al enrichment in natural mordenite with ca. 2 Ca p.f.u. close to T3.

Compositional reversals of orthopyroxene solid solution (Opx) on the enstatite (En, MgSiO 3 )-ferrosilite (Fs, FeSiO 3 ) join equilibrated with hematite (Hm, Fe 2 O 3 ), magnetite (Mt, Fe 3 O 4 ), and quartz (Qz) were obtained in piston-cylinder apparatus in the pressure range 10-25 kbar at temperatures of 800 and 950 °C. A double-capsule technique with the Hm-Mt external buffer was employed to ensure the presence of both Fe oxides in the run products. Starting materials were crystalline mixtures of synthetic Opx with X Fs [=Fe/(Fe+Mg)] = 0.05 and 0.45, reagent Mt, Hm, and natural Qz. Run products were carefully examined using optical microscopy, back-scattered (BSE) imaging, and wavelength dispersive electron microprobe analysis. Despite the differences in the rates of forward and backward reactions Fs (in Opx) + Hm = Mt + Qz (1), very tight compositional brackets (less than 0.01 X Fs ) were obtained at each P-T. At 800 °C, midpoints of the compositional brackets are: 0.195/10 kbar, 0.24/15, 0.29/20, and 0.34/25; at 950 °C: 0.21/15 kbar, 0.25/20, and 0.295/25 kbar. Activities of Fs, recalculated from these compositions using either Berman and Aranovich (1996) or Holland and Powell (1998) data sets for the pure end-member reaction (1), show negative deviation from ideal mixing. Strongly temperature dependent enthalpy of mixing is required to make these activity values consistent with the configurational entropy of the solid solution deduced from the site occupancy measurements. A simple mixing model with a linearly varying temperature-dependent W term, similar to that of Berman and Aranovich (1996), adequately reproduces the experimental data.

Efforts to simulate the extreme toughness of the polycrystalline diamond variety known as carbonado typically entail the sintering of diamond powders in the presence of metal solvent-catalysts. In this study, we have attempted to duplicate the carbonado microstructure by sintering diamond powders without catalysts in a multi-anvil press at pressures of 6 to 9 GPa, temperatures of 1200 to 1800 °C, and times up to 6 h. The resultant microstructural defect assemblages for each experimental condition were characterized by transmission electron microscopy (TEM). Despite the absence of catalysts, sintered compacts were successfully produced for all runs, though intergranular porosity was significantly higher than that observed in natural carbonado. Primary grain sizes were reduced by more than 50% from their original dimensions in some experiments due to surface fracturing and abrasion, and aperiodic slip planes rigorously parallel to {111} consistently emerged in high densities, with lamellar spacings of 3 to 30 nm. In addition, sintering over all conditions produced polysynthetic spinel twinning in close association with the partial slip defects. Compacts compressed at 8 GPa produced some euhedral crystals with very low dislocation densities surrounded by grains in which dislocation densities were quite high. In addition, curviplanar defects loosely constrained to {111} were visible within some specimens sintered at the highest pressures. These textures resembled the polygonalization fabrics and defect microstructures observed in natural carbonado (De et al. 1998), and the appearance of these features suggests that our experiments at their most extreme pressure and temperature parameters reproduced carbonado-like defect assemblages. The formation of such textures in quasi-hydrostatic experiments suggests that shock metamorphism is not required to produce the periodic defect lamellae observed in carbonado.

Tourmaline has recently been shown to incorporate large amounts of substituent B at the tetrahedral site. To characterize the response of the tourmaline atomic arrangement to differing amounts of substitution of B for Si, five samples were separated from a core-to-rim (∼3 mm) section of an Fe-bearing olenite with a dark green core and a nearly colorless rim from Koralpe, Austria. Crystal structures of the five samples were refined to R values <0.018 using three-dimensional X-ray methods, and the compositions of the crystals were determined by electron microprobe, secondary ion mass spectrometric, and Mössbauer analyses. From core to rim, [4] B increases monotonically from 0.35 to 0.65 apfu, whereas the mean T-O distance decreases from 1.621 to 1.610 Å. Optimized formulae using chemical and structural data range from X (Na 0.632 Ca 0.145 ⃞ 0.223 ) Y (Al 1.320 Fe 2+ 1.202 Li 0.190 Mg 0.086 Ti 0.028 Mn 2+ 0.024 ⃞ 0.150 ) Z Al 6.00 B 3.00 T (Si 5.525 B 0.333 Al 0.130 Be 0.012 ) O 27 [(OH) 3.19 O 0.81 ] (core composition) to X (Na 0.408 Ca 0.290 K 0.002 ⃞ 0.300 ) Y (Al 2.338 Li 0.365 Fe 2+ 0.084 Mn 2+ 0.009 Mg 0.005 Ti 0.005 ⃞ 0.194 ) Z Al 6.00 B 3.00 T (Si 4.989 B 0.615 Al 0.362 Be 0.034 ) O 27 [(OH) 3.41 O 0.59 ] (rim composition). The variation of chemistry and structure, coupled with short-range order constraints, demonstrates that (1) the average tetrahedral bond length (<T-O>) reflects the substitution of [4] B, (2) tourmaline samples with relatively high Fe 2+ contents (ca. 1 apfu Fe 2+ ) and <T-O> distances up to 1.621 Å can contain significant amounts of [4] B (up to ca. 0.3 apfu), (3) the presence of substantial [4] B is limited to, or more common in Al-rich tourmalines, (4) the presence of [4] B substituents favors OH at the O3 site, (5) the presence of Ca or Na at the X site is not simply correlated with occupancy of [4] B in the adjacent tetrahedral ring, and (6) no two B-substituted tetrahedra will link through bridging O atoms.

We present the discovery of a novel dense silica polymorph retrieved from shock-wave and diamond- anvil cell experiments. This polymorph is the first observed silicate composed of face-sharing polyhedra and it has a density similar to stishovite. Sterical constraints on the bond angles induce an intrinsic disorder of Si positions, such that the Si-coordination is transitional between four- and sixfold. The structure provides a mechanism for this coordination change in silica and other silicates at high temperature that is fundamentally different from mechanisms at 300 K. The new polymorph also illustrates how the face-sharing polyhedra, naturally occurring along previously proposed compression mechanisms for dense silicate melts, can be constructed without inferring unphysically small bond angles.

Crystallographic relationships between exsolved phases and their hosts are typically characterized using transmission electron microscopy (TEM) or single crystal X-ray diffraction (XRD). In this investigation, electron backscatter diffraction (EBSD) was used to determine the epitaxial relationships of exsolved laths of magnetite in clinopyroxenes from three sampling sites in the Cretaceous Messum Complex of Namibia. Two orientations of magnetite inclusions are found with their long axes subparallel to [100] and [001] of the host clinopyroxene. Inclusions subparallel to [100] c have [1̅10] m // [010] c , (1̅1̅1) m // (1̅01) c , and [112] m // [101] c . Inclusions subparallel to [001] c have [1̅10]m // [010] c , (111) m // (100) c , and [1̅1̅2] m // [001] c . The EBSD-derived orientation relationships agree well with previous TEM and XRD studies on similar materials. The crystallographic relationships obtained with EBSD are used in conjunction with optimal phase boundary theory to determine the exsolution temperature of the magnetite inclusions, which is of importance to paleomagnetic studies. For one sample, this temperature (840 ± 50 °C) can be compared with that (865 ± 25 °C) derived from a more widely used cation exchange geothermometer. Thus it appears clear that exsolution occurred well above the Curie temperature of pure magnetite (580 °C).