Volume 86, Issue 11-12

The theoretical infrared spectrum of kaolinite [Al 2 Si 2 O 5 (OH) 4 , triclinic] was computed using ab initio quantum mechanical calculations. Calculations were performed using the Density Functional Theory and the generalized gradient approximation. The low-frequency dielectric tensor of kaolinite was determined as a function of the light frequency using linear response theory. The IR spectrum was then calculated using a model that takes into account the shape and size of kaolinite particles. A remarkable agreement was obtained between theory and experiment, especially on the position of the stretching bands of OH groups. This agreement provides a firm basis for the interpretation of the IR spectrum of kaolinite in terms of structural parameters

The limiting fictive temperatures (T f ') of 16 multi-component silicate glasses have been derived quantitatively from heat capacity measurements, following the method of Moynihan et al. (1976). These quantitative values of T f ' closely match temperatures corresponding to the onset (T onset ) of the rapid rise in dilatometry heating curves (dL/L vs. T) at the glass transition, obtained on glasses with similar cooling histories. The mean deviation (T f '- T onset ) is 5 K, whereas the maximum deviation is 17 K. These results confirm that the T f ' of a silicate glass can be determined from the T onset of a glass dilatometry curve with an uncertainty that is <20 K. An application of the T f ' measurements includes the precise determination of the specific volumes of supercooled liquids at their respective T f ' values (Lange 1997). By comparison with other measurements in the literature, the accuracy of the T f ' method for determining low-temperature, fully relaxed, supercooled liquid volumes is shown. A comparison of volume-temperature models in the literature shows that a linear model (where thermal expansivity is independent of temperature) provides a superior fit of measured volumes in the SiO 2 -Al 2 O 3 -MgO-CaO-Na 2 O-K 2 O system over very wide temperature intervals (700-1900 K).

Molar enthalpies of formation of natural, near end-member tremolite and talc have been calculated from high temperature lead borate drop-solution calorimetry. Using three mutually consistent reaction cycles for each phase, the critical volatile component H 2 O was treated in fundamentally different ways, namely as the simple “oxide” [i.e., as the enthalpy difference (H TCal - H 298.15 ) only], or as the (OH)-component in both low and high water-content phases, respectively. The molar enthalpies of drop solution of all solid phases participating in the various reaction cycles (brucite, diopside, enstatite, calcite, magnesite, quartz) were independently measured. We corroborate that the dynamic gas flow techniques introduced by Navrotsky et al. (1994) are critically important for very hydrous phases such as brucite, but that this dynamic technique yields the same results as traditional static drop-solution techniques for phases with relatively lower water contents such as tremolite and talc. When these critical differences in H 2 O behavior are taken into account, it can be shown that drop-solution measurements on tremolite and talc over the last twenty years actually lead to similar and consistent results on the enthalpy of formation of these phases; reported differences are due to incorrect assumptions on the final state of H 2 O. Our refined values for the enthalpy of formation from the elements for ideal end-member tremolite and talc range from -12299.2 to -12308.9 kJ/mol and -5892.1 to -5900.2 kJ/mol, respectively, depending on the internally consistent data set used for the reaction cycle components. These values are identical within error to the results refined from various phase-equilibrium experiments and show that these two fundamentally different techniques can lead to a consistent picture on the thermodynamic properties of hydrous minerals, a conclusion that has been strongly questioned in the past. Heat capacity data have been obtained for the same samples of tremolite and talc by differential scanning calorimetry in step-scanning mode in the range 50-500 °C and 50-400 °C, respectively. The following best-fit equations apply to the ideal end-members [T in Kelvin, C P in in J/(mol · K)] in the range 298.15 to 800 K and 298.15 to 650 K, respectively: C P (tremolite) = 1335.77586 - 0.02378 T - 1.00400 10 7 T -2 - 9678.23152 T -0.5 . C P (talc) = 1300.66304 - 0.03007 T - 1.31510 10 7 T -2 - 8778.5468 T -0.5 .

Rietveld analysis of synchrotron X-ray powder diffraction data for the silica mineral moganite from 100 K to 1354 K has revealed a reversible phase transition from space group I2/a to Imab at approximately 570 K. The thermal expansion behavior of the lattice parameters alters sharply at the transition point, and the monoclinic β angle decreases to 90°. The displacive transition from α- to β-moganite is effected by the rotation of apparently rigid tetrahedra about the [010] axis, and the linear temperature dependence of the volume strain and of the non-symmetry-breaking e 11 and e 22 strains indicates that the character of the transition is second-order. The continuous increase in the b axis over the entire temperature range reveals a concomitant rotation of tetrahedra about [100] that does not affect the overall symmetry. In addition, we present a refinement without structural constraints for α-moganite at room temperature using time-of-flight neutron diffraction data.

KClO 3 thermal decomposition has been studied from 29-76 kilobars using a multianvil highpressure device and in-situ energy-dispersive X-ray diffraction and off-line quenching experiments. The rhombohedral form of KClO 3 was found to decompose to the B2 form of KCl and O 2 via an orthorhombic KClO 4 intermediate over this pressure interval. The decomposition temperature was found to vary only slightly with pressure. The online experiments gave decomposition temperatures between 500 and 580 °C. Further off-line quenching experiments using sealed gold tubes determined the equilibrium decomposition boundary to be 550 ± 15 °C over this pressure range. Unit-cell parameters and volumes were determined for the high-pressure phases of KClO 3 and KCl from the diffraction data. The partial molar volume of O 2 was calculated from the difference in the solid volumes. Oxygen fluid volumes were then calculated along the decomposition boundary and vary from 10.6 ± 0.2 cm 3 /mol at 29 kbar to 9.6 ± 0.1 cm 3 /mol at 76 kbar. These volumes are 30 to 50% less than previous estimates determined from shock wave data, and imply that oxygen can be more easily stored in Earth’s mantle and core than previously believed. The thermal equation of state of the B2 form of KCl was investigated online using NaCl as an internal pressure standard. KCl was then used as an internal pressure calibrant for the online KClO 3 decomposition experiments. The mechanical behavior of the multianvil high-pressure device was also studied and load vs. force characteristics are presented here.

The room-temperature bulk modulus of synthetic end-member clinochlore, Mg 5 AlSi 3 AlO 10 (OD) 8 , has been determined in the pressure range 0 to 6.5 GPa using neutron powder diffraction: K 0 = 75.4(2.7) GPa, K 0 ' = 4; K 0 = 72.3(2.4) GPa, K 0 ' = 5.4(1.0). A structural comparison is made with the related mineral brucite, Mg(OH) 2 . Clinochlore is a much stiffer structure than brucite (K 0 = 41 - 47 GPa). Both minerals have very similar in-plane compressions of their polyhedral sheets and so their very different bulk moduli arise from different compressibilities normal to the structural layering. Rietveld refinements of neutron-diffraction data for clinochlore collected at 0, 1.2, 2.5, and 4.7 GPa reveal that compression normal to the layering is equally partitioned between the interlayer and the 2:1 layer; the octahedral sheets of the brucite-like and 2:1 layers are uncompressed to 4.7 GPa. Increasing pressure strengthens the H3(D3)…O3 hydrogen bond, which contracts linearly from 1.88(2) Å at 0 GPa to 1.77(2) Å at 4.7 GPa, possibly with a concomitant lengthening of the O6-H3(D3) hydroxyl bond from 1.05(2) Å to 1.10(2) Å over the same pressure range.

The athermal equilibrium structure, the equation of state, the elastic constants, and O atom charges were calculated for Mg 2 SiO 4 wadsleyite over a range of pressures using a plane-wave pseudopotential method. The zero-pressure volume is 2% lower and the bulk modulus is 4.5% higher than experimentally observed. After correcting for zero point motion and the 300 K temperature difference between theory and experiment, using a Debye model, the calculated zero pressure volume is within 1% of experiment and the bulk modulus agrees within experimental error. The structure compresses anisotropically with linear moduli for the a, b, and c axes of 610 GPa, 599 GPa, and 454 GPa, respectively. The compression is largely taken up by the Mg octahedra M1, M2, and M3 which are much softer than the Si tetrahedra, with polyhedral bulk moduli of 161 GPa, 159 GPa, 157 GPa, and 331 GPa, respectively. The M1 and M3 octahedra were found to compress anisotropically which explains the greater compressibility of the c axis. The geometry of the Si2O7 group is characterized by a small Si-O-Si angle of 121.2°; compression of this group is largely accommodated by shortening of the Si-O bonds, while the inter-tetrahedral angle is almost pressure independent. We find that our results at ambient pressure are consistent with previously established systematics relating bulk modulus to volume, and Si-O-Si angle to Si-O bond length. However the variation of these quantities upon the application of pressure leads to trends that are distinct from the systematics. The calculated zero pressure elastic constants agree to within 10% with available Brillouin scattering data, with the exception of C 12 which is 15% higher than experimentally observed. The calculated isotropically averaged bulk and shear modulus and their pressure derivatives are K 0 = 182 GPa, K 0 ' = 4.23, and G 0 = 116 GPa, G 0 ' = 1.10, respectively. The aggregate velocities and their pressure derivatives are VP = 9.75 km/s, VP' = 0.056 km/s/GPa, and VS = 5.72 km/s, VS = 0.012 km/s/GPa. We find that the elastic anisotropy of wadsleyite is intermediate between the two other Mg 2 SiO 4 polymorphs, forsterite, and ringwoodite. The anisotropy is only weakly pressure dependent and decreases with increasing pressure. The azimuthal and polarization anisotropy for S waves 14.5% and 12.8% respectively, is almost pressure independent, while the azimuthal anisotropy for P waves decreases from 12.5% at ambient pressure to 10.5% in the upper part of the transition zone (14-17 GPa). Our calculated O atom charges suggest that Ol is the most likely hydroxyl site and remains so throughout the stability field of wadsleyite.

Experiments on a MORB (Mid Ocean Ridge Basalt)-derived amphibolite have been conducted in order to determine fluid-absent solidus and supersolidus phase relationships. The solidus has been determined in the range 4-14 kbar by using the incremental heating technique, tested against classical procedures. Peritectic clinopyroxene was produced below 10 kbar and epidote was the stable peritectic phase at pressures greater than 10 kbar for temperatures near the solidus. Garnet was the peritectic phase for temperatures above 800 ∞C and pressures of 10 to 14 kbar. The shape of the solidus was similar to previous estimations but the temperature interval between the high- and low-pressure regions was strongly reduced. With these new estimations, the possibilities for melting of the oceanic crust at P below 10 kbar are increased and, consequently, the number of tectonic scenarios needed for magma generation from amphibolite sources is enlarged. In the absence of free water, the solidus in the pressure region above 10 kbar is at higher T than previous estimations.

Fluorine-rich phlogopite [F content up to ~8 wt%; F/(F + OH) ~0.9] in ultra-high-temperature metapelitic granulites from the Napier Complex, East Antarctica is associated with aluminous orthopyroxene, osumilite, sapphirine, garnet, and quartz. Textural relationships imply that some of the phlogopite is of primary origin and stable under ultra-high-temperature conditions. This is in accord with recent experimental evidence on the stability of F-rich phlogopite. Because the F-rich phlogopite also occurs as rounded inclusions in aluminous orthopyroxene (Al 2 O 3 up to 12.8 wt%), sapphirine, osumilite, and garnet, it is inferred that the ultra-high-temperature mineral assemblages, which includes these minerals formed during prograde partial melting reactions at the expense of phlogopite, at a depth of less than 30 km. Thus the coarse-grained peak metamorphic assemblages formed below 9 kbar, and there is no evidence the rocks underwent any significant degree of decompression during or soon after peak metamorphic conditions. The phlogopite breakdown reactions we suggest on the basis of textural arguments differ from those postulated from experiments on F-free systems.

This paper uses constraints from experiments, thermodynamic modeling, and natural mineral assemblages to argue that Al 2 SiO 5 “triple-point” assemblages, in which all three Al 2 SiO 5 minerals are in stable equilibrium, are not possible in common muscovite(Ms)+biotite(Bt)+quartz(Qtz)-bearing metapelitic rocks because the reactions that first introduce an Al 2 SiO 5 mineral to these bulk compositions occur at higher temperature than the triple point. Less-common, highly aluminous bulk compositions may develop Al 2 SiO 5 minerals at temperatures below the triple point such that stable triple-point assemblages are theoretically possible. The “invisibility” of the triple-point to common Ms+Bt+Qtz-bearing metapelites calls into question most metapelitic triple-point localities reported in the literature, and carries implications for the topology of the metapelitic petrogenetic grid, the bathozone/bathograd scheme of Carmichael (1978), and the possibility of prograde kyanite → andalusite → sillimanite sequences. Re-examination of reported triple-point localities suggests that in most if not all cases, the Al 2 SiO 5 minerals grew at different times in the metamorphic history of the rock

Experiments were carried out in the quartz-saturated part of the system KFMASH at hematitemagnetite buffer conditions in a piston-cylinder apparatus using synthetic biotite, K-feldspar, quartz, and sillimanite 80 -quartz 20 gel. Natural garnet and sillimanite were used as seeds. Three bulk compositions were used: X Mg = 0.81, 0.72, and 0.53, and the experiments were vapor absent. In the bulk composition with the highest X Mg , dehydration-melting of biotite produced osumilite coexisting with spinel solid solution within the P-T window 7-8.5 kbar and 850-1000 °C. However, this mineral pair coexisted with cordierite at P < 7.5 kbar, and with sillimanite above 7.5 kbar. There is no appreciable change in the lower thermal stability of osumilite with lowering of X Mg in the bulk composition, but both the upper pressure and upper thermal stabilities are reduced. With increasing pressure, osumilite + spinel breaks down to an orthopyroxene + sillimanite assemblage. There is no overlap of the stability fields of osumilite and sapphirine. Rather, sapphirine-bearing assemblages are produced at the expense of osumilite with increasing temperature. Sapphirine, however, did not appear in the bulk composition having X Mg = 0.53. Instead, a wide stability field of spinel + cordierite + orthopyroxene + sillimanite is noted at P > 7 kbar and T > 950 °C. Garnet + orthopyroxene + sillimanite become stable at lower pressures with decreasing X Mg in the bulk composition. The effects of additional components in natural assemblages, such as Zn in spinel and F and Ti in biotite, on the KFMASH equilibria are evaluated qualitatively. The experimental results were applied to six high-grade terranes, where osumilite + spinel coexistence at peak metamorphic conditions has been reported. The deduced P-T window for the osumilite + spinel association is entirely consistent with independent PT estimates for the natural occurrences. It is argued that osumilite + spinel assemblage should be characteristic of metamorphism of highly magnesian pelites at high f O₂ along a prograde path of high dT/dP.

Euhedral crystals of purple spinel are associated with calcite and forsterite in the central zone of composite metasomatic veins of the Adamello contact aureole (Northern Italy). In thin section, spinel is generally colorless, but exhibits concentric zones of dark brown color. These zones are aligned parallel to prominent crystal faces and contain abundant anisotropic, lath-shaped inclusions with their long axis oriented in the six <110> directions of spinel. This orientation results from the alignment of {001} Geik and the {111} Sp crystallographic planes. The small (up to approximately 20 × 1 μm) inclusions were identified as geikielite by means of element distribution maps, micro-Raman spectroscopy, and transmission electron microscopy. Electron microprobe data revealed that spinel consists mainly of MgAl 2 O 4 , with only minor amounts of other components. The overall chemical variation within individual crystals is characterized by a continuous increase in FeO tot and ZnO from center to rim (0.6-2.1 wt% and <0.1-1.0 wt%, respectively). Whereas V 2 O 3 is relatively constant (0.15 wt%), the bulk TiO 2 concentration (analysis spot: 20 μm diameter) increases in general from center (0.12 wt%) to rim (1.2 wt%). This variation, however, is marked by several discontinuities that correspond to the zones containing geikielite inclusions. Where geikielite occurs, the spinel matrix is depleted in Ti relative to the inclusion-free zones. Geikielite was analyzed by analytical electron microscopy and has the formula Mg 0.83 Fe 0.15 Mn 0.02 (Ti 0.99 Fe 0.01 )O 3 . The textural relationships between geikielite inclusions and spinel matrix, combined with the analytical data, suggest that geikielite laths represent exsolution features, thus, a miscibility gap exists between Mg-Al-spinel and geikielite. The purple color results from scattering of light by the geikielite inclusions. Spinel with similar features was also observed in metasomatic veins of the Bergell contact aureole (Italy) and in granulite-facies marbles of Southern India.

Aeromagnetic maps of the Egersund Mid-Proterozoic igneous province show a spectacular range of positive and negative magnetic anomalies with a contrast up to 15 600 nT. The positive magnetic anomalies are over magnetite norites and overlying mangerites and quartz mangerites of the Bjerkreim- Sokndal layered intrusion. These rocks are dominated by multi-domain (MD) magnetite. The negative magnetic anomalies are over ilmenite-rich norites of the Bjerkreim-Sokndal layered intrusion, the Tellnes ilmenite norite ore deposit, and massif anorthosites. These rocks are dominated by hemoilmenite and/or by silicates containing fine-grained oxide exsolution lamellae. Electron microprobe analyzes of coexisting Fe-Ti oxides in the layered intrusion confirm earlier observations that oxides in early magmatic rocks are dominated by hemo-ilmenite with minor end-member magnetite, followed by more reduced oxides dominated by titanomagnetite with minor near end-member ilmenite. What is not fully understood is the property of ilmenite with hematite exsolution lamellae, or, even more striking, hematite with ilmenite lamellae, to produce strong remanent magnetization of high coercivity and with a Néel temperature equal to or above the Curie temperature of magnetite. This property makes the rhombohedral oxides an important candidate to explain some high-amplitude deep-crustal anomalies on earth, or strong remanent magnetization on other planets. A remarkable feature in the Egersund province is that primitive magmas produced rocks rich in hemo-ilmenite causing negative magnetic anomalies related to magnetic remanence, and more evolved magmas produced rocks rich in magnetite related to positive induced magnetic anomalies, all in the course of crystallization-differentiation.

Chemistry of the Fe-Ti oxide minerals in the Bjerkreim-Sokndal layered intrusion and in the Tellnes Ilmenite Norite suggests some parallels in magmatic evolution and resultant petrography. Early ilmenites have high ratios of Mg/(Mg + Fe) and Cr/R 3+ and intermediate to low V/R 3+ , and abundant fine hematite exsolution lamellae, in which Cr, V, and Al are concentrated. Magmatic evolution results in lower Mg/(Mg + Fe), lower Cr/R 3+ and higher V/R 3+ . With magma saturation with magnetite, V/R 3+ falls both in ilmenite of rapidly decreasing abundance, and in magnetite, but there is an increase in Mg/(Mg + Fe) of ilmenite, probably due to late ion exchange with much more abundant magnetite. Aside from implications related to magma evolution, these results give information concerning fractionation of elements between the rhombohedral and cubic oxide series, and especially between the ilmenite hosts and their contained hematite exsolution lamellae

The compound Ca 4 Fe 1.5 Al 17.67 O 32 was synthesized at 2.5 GPa and 1250 °C using a piston-cylinder apparatus. The crystal structure, determined from single-crystal X-ray diffraction data collected at 295 K (tetragonal, space group I4̄ - 2d, a = 20.1847(14) Å, c = 5.6203(6) Å, V = 2289.83(3) Å 3 , Z = 4) was refined to a final R-index of 0.024 for 1229 independent observed reflections and 130 parameters. The main building units comprising the compound are 11.8 × 7.2 Å wide spinel-type ribbons running parallel to [001], which are connected via corner-sharing (Fe,Al)O 6 -octahedra. Additional linkage between the spinel units is provided by AlO 4 -tetrahedra residing on the 4̄-axis as well as by Ca cations, in sevenfold coordination with oxygen attached to the spinel-like building units. Refinement of site occupancies reveals that the incorporation of Fe occurs at two of four octahedral and at two of three tetrahedral sites. Apparently, Ca 4 Fe 1.5 Al 17.67 O 32 represents a new structure type exhibiting more pronounced structural modifications relative to the spinel aristotype compared with the socalled spinelloid structures. Furthermore, this synthetic compound contains Fe 2+ in tetrahedral coordination, which is relatively uncommon among inorganic materials.

Phase X, ascribed by Luth (1995) to a hydrous K-rich silicate formed from the breakdown of Kamphibole at high pressures, was synthesized at 1250-1300 °C and 10-16 GPa in four different compositions: Na 1.78 (Mg 1.89 Al 0.13 )Si 2.02 O 7 (anhydrous sodic phase X), Na 1.16 K 0.01 (Mg 1.89 Al 0.14 ) Si 2.02 O 7 H 0.65 (sodic phase X), K 1.85 Mg 2.06 Si 2.01 O 7 (anhydrous phase X), and K 1.54 Mg 1.93 Si 1.89 O 7 H 1.04 (phase X). A general chemical formula for these phases can be expressed as A 2-x M 2 Si 2 O 7 H x , with A = K and/or Na, M = Mg and/or Al, and x = 0-1. Structure determination from single-crystal X-ray diffraction data shows that anhydrous sodic phase X is trigonal with space group P3̄̄-1m, whereas the other three have an identical structure with space group P6 3 cm. Both P3̄̄1m and P6 3 cm structures are characterized by MgO 6 octahedral layers that are stacked along the c axis and inter-linked together by Si 2 O 7 tetrahedral dimers and K or Na cations. Within the MgO 6 layers, each MgO 6 octahedron shares three edges with neighboring MgO 6 octahedra to form brucite-like layers with one out of three octahedral sites vacant. Large K or Na cations are situated right below and above each occupied octahedron in the MgO 6 layers, whereas the Si 2 O 7 groups are located below and above each vacant octahedron in the layers. The two types of structures, however, differ in the relative orientation of MgO 6 octahedral layers, the coordination of K or Na, and the configuration of SiO 4 tetrahedral dimers. By comparison, the Na 2 Mg 2 Si 2 O 7 phase synthesized by Gasparik and Litvin (1997) appears to have the stoichiometry identical to anhydrous sodic phase X. Hence, these two highpressure phases are likely to possess the same structure, or at least are closely related to each other structurally

Fluoro-edenite, ideally NaCa 2 Mg 5 (Si 7 Al)O 22 F 2 , was found both as prismatic or acicular crystals of millimetric size and as fibers in the rock cavities in gray-red altered benmoreitic lavas occurring at Biancavilla (Etnean Volcanic Complex, Catania, Italy). It is associated with feldspars, quartz, clino- and orthopyroxene, fluoro-apatite, ilmenite, and hematite, and probably crystallized from late-stage hydrothermal fluids. Fluoro-edenite is transparent, intense yellow, non-fluorescent, has vitreous to resinous luster, and gives a yellow streak parallel to the c axis; Mohs’ hardness 5-6, D calc = 3.09 g/cm 3 , perfect cleavage on {110}, and conchoidal fracture. In plane-polarized light, fluoro-edenite is birefringent (1 st order), biaxial negative, α = 1.6058(5), β = 1.6170(5), γ = 1.6245(5), 2V calc = 78.09°, Y ≡ β ⊥ (010), and γ:Z = 26°. No pleochroism is observed. Fluoro-edenite is monoclinic, space group C2/m, a = 9.847(2) Å, b = 18.017(3) Å, c = 5.268(2) Å, β = 104.84(2)°, V = 903.45 Å 3 , Z = 2; the ten strongest X-ray diffraction lines in the powder pattern are [d(I, hkl)]: 3.125(10, 310), 8.403(6,110), 3.271(5,240), 2.807(4,330), 2.703(3,151), 1.894(2,5̄10), 2.938(2,221), 1.649(2, 461), 3.376(2,131), 1.438(2,6̄61). IR analysis showed absorption bands at 1066, 991, 791, 738, 667, 517, 475 cm -1 , and no bands in the OH-stretching region. Structure refinement allowed determination of cation site-preference and ordering. Microprobe analysis of the refined crystal gave SiO 2 52.92, TiO 2 0.29, Al 2 O 3 3.53, FeO t 2.50, MnO 0.46, MgO 22.65, CaO 10.83, Na 2 O 3.20, K 2 O 0.84, F 4.35, Cl 0.07 wt%, and the crystal-chemical formula obtained by combining all the available data is: A (Na 0.56 K 0.15 ) B (Na 0.30 Ca 1.62 Mg 0.03 Mn 0.05 ) C (Mg 4.68 Fe 2+ 0.19 Fe 3+ 0.10 Ti 4+ 0.03 ) T (Si 7.42 Al 0.58 ) O 22 O3 (F 1.98 Cl 0.02 ) 2 .

Jacobsite-rich Fe-Mn ores of probable Dogger age fill paleokarst pockets in the Triassic marbles of the Barrhorn Unit under Pipjigletscher in the Turtmanntal, Valais, Switzerland. These ores and embedding rocks underwent Tertiary metamorphism under upper greenschist facies conditions. Some of these jacobsite ores contain minor amounts of a yellow micaceous mineral, which appears to be a new Mn-Mg silicate-vanadate-arsenate that was named “turtmannite” with respect to the type locality. Turtmannite flakes up to 200 μm in length occur parallel to the main schistosity, or fill thin discordant veinlets. Turtmannite is rhombohedral R3̄c, with a H = 8.259(2) and c H = 204.3(3) Å in the hexagonal setting. The corresponding primitive rhombohedral cell has a R = 68.31 and α R = 6.92°. HRTEM images indicate that turtmannite is perfectly ordered along c. The structure of turtmannite has been solved to a final R1 of 12.4% on a Siemens Smart CCD diffractometer with MoKα X-radiation, and a detector to sample distance extended to 12 cm. The structure consists of 84 oxygen layers stacked along c, with twelve close-packed layers followed by two non-close-packed layers. This sequence is repeated six times. The structure contains eight symmetrically distinct cation layers. Three different occupational variants have been recognized leading to the following hypothetical end-member formulae and approximate abundances: I [IV] Mn 1.5 [IV] Mg 3 [VI] (Mn,Mg) 21 [(V,As)O 4 ] 3 [SiO 4 ] 3 O 5 (OH) 20 , 50% II [IV] Mn 1.5 [VI] (Mn,Mg) 21 [(V,As)O 4 ] 3 [SiO 4 ] 3 [AsO 3 ](OH) 21 33% III [IV] Mn 1.5 [VI] (Mn,Mg) 21 [(V,As)O 4 ] 3 [SiO 4 ] 2 [SiO 3 OH](OH) 25 16% The simplified chemical formula for turtmannite can be written as: (Mn,Mg) 22.5 Mg 3-3x [(V,As)O 4 ] 3 [SiO 4 ] 3 [AsO 3 ] x O 5-5x (OH) 20+x . The unit cell of turtmannite is similar to those of mcgovernite, a Mn-Mg-Zn arsenate from Sterling Hill, New Jersey, and an unnamed “mcgovernite-like” Mn-Mg arsenate from the Kombat Mine, Namibia. The crystal structure of turtmannite is close to the model predicted for mcgovernite by Moore and Araki (1978)

The crystal structure of the low temperature form of natural melanophlogite, 46SiO 2 · 6M 14 · 2M 12 (M 14 = N 2 , CO 2 ; M 12 = CH 4 , N 2 ), was determined using single-crystal X-ray diffraction data at room temperature. The structure is tetragonal with space group P4 2 /nbc and unit cell a = 26.818(2) and c = 13.365(1) Å, which is the (2 × 2 × 1) superstructure of high-temperature cubic melanophlogite and includes four formula units. The structure with 335 variable parameters including anisotropic temperature factors (or atomic displacement factors) was refined to R = 0.0288 for 2706 observed reflections. The main silica framework consists of nearly regular SiO 4 tetrahedra forming large internal voids represented by distorted tetrakaidecahedra and pentagondodecahedra, which accommodate CO 2 or N 2 and CH 4 or N 2 guest molecules, respectively. This low temperature form is a displacive variant of the cubic high-temperature form. The mean bond length is 1.588(4) Å for Si-O, and the bond angles for Si-O-Si are distributed over a large range from about 145 to 171° with a mean value of 159.4(3)°. The thermal vibrations of Si are nearly isotropic with amplitudes approximately equal to the average mean square displacement of 0.0119(4) Å 2 . The thermal vibrations of the O atoms are highly anisotropic with a wide range of mean square displacements. There is a positive correlation between the Si-O-Si bond angles and the mean-square displacements of the O atoms.

Mullite-type alkali aluminates (KyNa 1-y ) 0.67 Al 6 O 9.33 were synthesized from amorphous Al and alkali nitrates by sol-gel techniques. Rietveld refinements of six members of the solid solution series (y = 0.0, 0.2, ..., 1.0), together with Fourier syntheses and grid search analyses show that the Na and K atoms reside in the vacant Oc sites, with K at 1/2, 0, 1/2 and Na on a split site off the special position. The number of alkali atoms is restricted to 2/3 atoms per unit cell due to crystal chemical constraints. Consequently, unlike the aluminosilicate mullites, alkali mullites do not form a solid solution series with varying oxygen composition. All compounds studied here crystallize in space group Pbam with lattice constants ranging from a = 7.6819(4) Å, b = 7.6810(4) Å, c = 2.91842(8) Å for the Na aluminate to a = 7.6934(3) Å, b = 7.6727(3) Å, c = 2.93231(7) Å for the K aluminate mullite

Heat-induced color changes of fossilized Miocene mastodon ivory (13-16 Ma) have been known at least since the Middle Ages. Cistercian monks are believed to have created odontolite, a turquoise- blue “gemstone,” by heating mastodon ivory found in Miocene geological layers next to the Pyrrenean chain, France, to use it for the decoration of medieval art objects. This material has been the object of investigations of famous European naturalists and gemmologists, among them Réaumur (1683-1757). Although vivianite [Fe 3 (PO 4 ) 2 ·8H 2 O] is the commonly accepted coloring phase supposed to appear when heating fossilized mastodon ivory, our previous spectroscopic studies using PIXE/PIGE and TEM-EDX demonstrated that the chemical composition of collection odontolite and heated mastodon ivory corresponds to well-crystallized fluorapatite [Ca 5 (PO 4 ) 3 F] containing trace amounts of Fe (230-890 ppm), Mn (220-650 ppm), Ba (160-620 ppm), Pb (40-140 ppm), and U (80-210 ppm). No vivianite has been detected. To provide new insights into the physico-chemical mechanism of the color transformation of fossilized ivory, we used the combination of UV/visible/near-IR reflectance spectroscopy, time-resolved laser-induced luminescence spectroscopy (TRLIF), and X-ray absorption near-edge structure (XANES). Contrary to what had formerly been described as the color origin in odontolite, our study has conclusively identified traces of Mn 5+ by UV/visible/near-IR reflectance spectroscopy, TRLIF, and XANES inside the fluorapatite. Thus, odontolite owes its turquoise-blue color to Mn 5+ ions in a distorted tetrahedral environment of four O 2- ions. XANES also demonstrated oxidation of disordered octahedral Mn 2+ ions to tetrahedral Mn 5+ species in apatite during the heat process. So we give the first evidence of the real color origin in odontolite.

The single-crystal elastic moduli of natural samples of both calcite (CaCO 3 ) and rhodochrosite (MnCO 3 ) have been measured by Brillouin spectroscopy under ambient condition. Based on the trigonal unit cell, the elastic constants C 11 , C 33 , C 44 , C 12 , C 13 , and C 14 are 149.4(7), 85.2(18), 34.1(5), 57.9(11), 53.5(9), -20.0(2), and 223.9(15), 132.6(41), 44.5(9), 93.4(21), 76.0(23), -17.3(6) GPa for CaCO 3 and MnCO 3 , respectively. Our data for calcite are in good agreement with earlier data obtained by ultrasonic experiments. The off-diagonal elastic constants (C 12 , C 13 , and C 14 ) for rhodochrosite have systematically larger values than the trend defined by other isostructural carbonates, in all of which the divalent cations are alkaline-earth metals. This is a distinctive signature of transition- metal-bearing oxides, which is present in silicates and simple oxides as well.

Neutron diffraction results are linked to the pre-diffraction, morphological crystallography of E.S. Dana to show that ikaite is the precursor of “glendonite” and “thinolite” and that some of the unusual features of the morphology can easily be explained by an accidental near-equivalence of the unit cell of ikaite, CaCO 3.6 H 2 O. This relationship between the morphology of the pseudomorphs and the crystallography of ikaite gives strong justification for the use of “glendonite” as a paleothermometer, representing near freezing conditions for water.