Band 87, Heft 2-3

Mounting evidence suggests that partial disequilibrium-meaning disequilibrium for some elements, but not for others-may be a common but rarely detected phenomenon during metamorphic mineral growth, even in ordinary prograde reactions that progress to completion. Detailed examination of compositional variations in garnet crystals over a range of metamorphic grade suggests distinctly different scales of equilibration for common elements, with strong temperature dependence. Under lower greenschist-facies conditions, Fe and Mg may equilibrate at hand-sample scale, whereas Mn and Ca may not equilibrate even at millimeter-scale. Although Mn may achieve hand-samplescale equilibration under upper greenschist-facies conditions, Ca and many trivalent cations (e.g., REEs) may not do so until temperatures exceed those of the middle amphibolite facies. Even in the lower granulite facies, some elements (e.g., Y, Yb) show indications of disequilibrium at sub-centimeter- scales during garnet growth. Analysis and numerical modeling of undisputed disequilibrium textures demonstrate that the most common impediment to equilibration during metamorphism is the sluggishness of intergranular diffusion, the same mechanism known to govern porphyroblast crystallization in many metamorphic environments. Despite this importance to petrology, few quantitative determinations exist of intergranular diffusion rates under metamorphic conditions. Using numerical simulations of coupled intergranular and intracrystalline diffusion processes in coronal textures around partially resorbed garnet crystals from the Llano Uplift, U.S.A., a very precise and relatively accurate estimate is obtained for the rate of intergranular diffusion of Al in fluid-undersaturated systems. This result and an earlier estimate for fluid-saturated systems provide bracketing values for Al diffusivity during metamorphism.

Organic-rich Carboniferous shales with associated coal seams, from the Bacia Carbonífera do Douro-Beira (north Portugal), have been studied by TEM as well as by a variety of other methods. NH 4 and K micas and berthierine form small subparallel packets of a few layers separated by lowangle boundaries. One- and two-layer ordered polytypes, with some spot enlargement typical of minor disorder, occur in the NH 4 micas (i.e., metamorphism tobelites). They exhibit all the characteristics commonly described for sub-greenschist-facies, including a lack of textural and chemical equilibrium. The compositions of the 2:1 layers vary considerably for both K- and NH 4 -micas, and except for Ti, exhibit similar compositional ranges. The most significant compositional variations are explained by phengitic substitution (Si from 3 to 3.25, Fe + Mg from 0.1 to 0.3), but no evidence of an illitic substitution has been found. NH 4 in tobelites was determined by analysis of NH 3 using Nessler’s reagent, basal spacing, and 1-(K + Na). The resulting values indicate NH 4 contents ranging from 30 to 59% of the interlayer site occupancy. At very low temperatures of metamorphism (e.g., North Sea), the intergrowths of NH 4 and K in micas is on the nanometer-scale. With an increase in temperature (e.g., Pennsylvania), NH 4 - and K-micas are present as different minerals. The Douro- Beira samples represent a still higher-temperature case and, therefore, NH 4 and K can be present in the same interlayer sheet but with one cation being dominant. NH 4 - and K-dominated micas have segregated into well-separated packets with scarce intergrowths and almost no mixed-layers. Hence they show a solvus relationship.

Experiments were conducted to determine whether vibrations with frequencies in the sonic to ultrasonic range have any effect on the crystallization of basalt. One-atmosphere melting experiments were done at the QFM buffer on a sample of alkali basalt. A piezoelectric quartz crystal and a function generator were used to generate waves of 1.5 MHz and 150 kHz, and waves were propagated along the platinum quench wire to which the experimental charge was attached. Experiments were conducted at 1182 °C; at this temperature the sample was about 25% crystalline and contained olivine and plagioclase. Compared to the static control experiments, vibrated experiments produced plagioclase crystals that were less euhedral and had lower aspect ratios. Crystal size distributions for plagioclase were steeper for the vibrated experiments, with a higher population density of smaller crystals. The effect on olivine crystal size distributions was less. If the effect on plagioclase extends to lower vibrational frequencies, it is possible that this phenomenon may occur in magmas prior to eruption. The experimental technique developed to produce vibrations is potentially useful for improving homogeneity of melt and increasing likelihood of achieving mineral-melt equilibrium in partitioning experiments

This study explores the controls of temperature, pressure, and silicate melt composition on S solubility in silicate liquids. The solubility of S in FeO-containing silicate melts in equilibrium with metal sulfide increases significantly with increasing temperature but decreases with increasing pressure. The silicate melt structure also exercises a control on S solubility. Increasing the degree of polymerization of the silicate melt structure lowers the S solubility in the silicate liquid. The new set of experimental data is used to expand the model of Mavrogenes and O’Neill (1999) for S solubility in silicate liquids by incorporating the influence of the silicate melt structure. The expected S solubility in the ascending magma is calculated using the expanded model. Because the negative pressure dependence of S solubility is more influential than the positive temperature dependence, decompression and adiabatic ascent of a formerly S-saturated silicate magma will lead to S undersaturation. A primitive magma that is S-saturated in its source region will, therefore, become S-undersaturated as it ascends to shallower depth. In order to precipitate magmatic sulfides, the magma must first cool and undergo fractional crystallization to reach S saturation. The S content in a metallic liquid that is in equilibrium with a magma ocean that contains ~200 ppm S (i.e., Earth’s bulk mantle S content) ranges from 5.5 to 12 wt% S. This range of S values encompasses the amount of S (9 to 12 wt%) that would be present in the outer core if S is the light element. Thus, the Earth’s proto-mantle could be in equilibrium (in terms of the preserved S abundance) with a core-forming metallic phase.

The incorporation of molecular water into the c-parallel structural channels in end-member beryl VI Al 2 [Be 3 Si 6 O 18 ]·n H2O , where n H2O = 1.00 corresponds to 3.24 wt% H 2 O, was studied experimentally in the temperature-pressure range 400-950 ­­°­C and 1-14.5 kbar (P H2O = P tot ). Time-dependent studies ensured that the incorporation of water by beryl had reached the equilibrium n H2O values in the hydration experiments. Increasing water contents were found to enhance the refractive indices of beryl as follows: n ε = 1.5585 + 0.0080·wt% H 2 O (r = 0.986) n ω = 1.5616 + 0.0086·wt% H 2 O (r = 0.978) These relations were used to determine the water contents attained in the equilibrium experiments, and were calibrated using synthetic beryls, the water contents of which had been determined gravimetrically with an accuracy of ±0.06 wt% H 2 O. The incorporation of water into the structural channels of beryl leaves the hexagonal a-axis unchanged, a = 9.2110(10) Å, c and V increase as: c (Å) = 9.1893 + 0.0020·wt% H 2 O (r = 0.974) V (Å 3 ) = 674.98 + 0.203·wt% H 2 O (r = 0.936) Two isobars at 3.4 and 10.8 kbar with P H2O = P total could be modeled as: wt% H 2 O (P) = a - bT (°C) + cT 2 (°C) 3.4 kbar: a = 4.10, b = 5.90·10 -3 , c = 0.31·10 -5 (r = 0.963) 10.8 kbar: a = 3.31, b = 0.514·10 -3 , c = 0.00 (r = 0.941) Similarly, the two isotherms at 600 and 900 °C could be modeled as: wt%H 2 O (T) = a + bP (kbar) - cP 2 (kbar) 500 °C: a = 0.94, b = 0.261, c = 0.0075 (r = 0.949) 900 °C: a = 0.19, b = 0.331, c = 0.0092 (r = 0.964) These relations together with data points from additional experiments were used to construct a (P H₂O = P tot ) - T field of isohydrons for the system end-member beryl-H 2 O. Comparing n H₂O -P sections at 600 °C for beryl obained in this study with data for Mg-cordierite obtained from the literature, proves nearly the same equilibrium contents of water in both minerals.

This study investigates chlorapatite [ideally Ca 10 (PO 4 ) 6 Cl 2 )] from two spatially separated locations in the Ødegårdens Verk apatite mine, Bamble Sector, SE Norway, which was partially metasomatized under amphibolite-facies conditions. Relative to the original chlorapatite, metasomatized areas in sample DL136 are enriched in OH and F, depleted in Na, Si, and (Y + REE) and contain numerous inclusions (1-15 μm) of monazite and xenotime. In contrast, metasomatized areas in sample TN174 are enriched in F and Si, depleted in Na, show (Y + REE) abundances similar to the original chlorapatite, and host only a few monazite and xenotime grains. In order to define the role of fluids in the formation of (Y + REE)-phosphate minerals associated with apatite, a series of experiments covering wide ranges of temperature (900-300 °C) and pressure (1000-500 MPa) for a series of fluid compositions (pure H2O, a 50/50 molar mix of H 2 O and CaF 2 , and CO 2 with 1-2 wt% H 2 O) were conducted on unaltered chlorapatite from the Ødegårdens Verk. In the H 2 O experiments, features due to metasomatic alteration [depletion in Na, Si, and (Y + REE) as well as numerous monazite and xenotime inclusions], as observed in the natural sample DL136, were fully reproduced. Monazite and xenotime grains were also observed growing on the surface of the apatite in dissolution embayments in metasomatized areas. In the 50/50 CaF 2 -H 2 O experiments, metasomatized regions are strongly enriched in F and Si, depleted in Na, show unchanged (Y + REE) abundances, and contain only a very few, small monazite and xenotime grains similar to what was observed in sample TN174. Natural and experimental observation imply the presence of two coupled substitutions: Na + + (Y + REE) 3+ = 2 Ca 2+ and Si 4+ + (Y + REE) 3+ = P 5+ + Ca 2+ . In the case of DL136 and H 2 O-fluid experiments, Na + and Si 4+ went into solution, whereas the released (Y + REE) contributed to the growth of monazite and xenotime within the metasomatized apatite structure. For sample TN174 and the 50/50 CaF 2 + H 2 O-fluid experiments, Na in apatite became depleted but Si was enriched thereby stabilizing the (Y + REE) in the now recrystallized fluor-hydroxylapatite structure, which retarded the growth of monazite or xenotime inclusions. The basic conclusion of this study is that nucleation of monazite and xenotime in the metasomatized chlorapatite is principally a function of the composition of the infiltrating fluid and, to a much lesser extent, temperature and pressure.

The first direct measurements are reported for the structure of the hydrated La 3+ ion in an aqueous solution (containing 0.007 m La) over a range of temperatures from 25 to 300 °C and pressures up to 1600 bars. The radial distribution of atoms around the La 3+ ion was measured using the X-ray absorption fine structure (XAFS) technique. La L 3 -edge spectra were collected in the fluorescence mode from nitrate solutions in a modified hydrothermal diamond anvil cell using the PNC-CAT Xray microprobe at the Advanced Photon Source, Argonne National Laboratory. Analysis of the XAFS spectra collected at all temperatures indicates that each La 3+ ion has a hydration number of nine and that the solvating waters surround the ion in a tricapped trigonal prismatic arrangement. As temperature is increased from 25 to 300 °C, the bond distance between the equatorial-plane O atoms and the La 3+ ion increases from 2.59 ± 0.02 to 2.79 ± 0.04 Å, whereas the bond distance between La 3+ and the O atoms at the ends of the prism decrease to 2.48 ± 0.03 Å. This study also demonstrates the unique capability of the modified hydrothermal diamond anvil cell for in situ low energy X-ray spectroscopic analysis of elements in dilute aqueous solutions at elevated temperatures and pressures.

Pink fibrous crystals within massive rose quartz from localities in California, South Dakota, Brazil, Madagascar, and Namibia were examined with high-resolution transmission electron microscopy (HRTEM) and analytical electron microscopy (AEM). This study reveals that the nanofibers in all samples are related to dumortierite. Selected-area electron diffraction (SAED) patterns and HRTEM images indicate that the dumortierite-related fibers have a superstructure with a doubled periodicity along the a and b axes of dumortierite, giving cell parameters a = 2a dum = 2.36 nm, b = 2b dum = 4.05 nm, c = c dum = 0.47 nm. Computer simulations suggest that periodic arrangements of two different M1 site occupancies in the octahedral face-sharing chains give rise to the superstructure. One type of M1 site is occupied mainly by Al, whereas the other type is dominated by Ti and Fe. Simulated HRTEM images based on our proposed model match the experimental images. Most of the fibers, elongated along the c axis, are free of defects. AEM analysis shows that the dumortierite-related fibers have a composition similar to well-characterized dumortierite, but that they contain a greater amount of Fe substituting for Al in the M1 sites. Boron was detected in all fibers examined by electron energy loss spectroscopy as well as in sillimanite crystals found as a minor component in one rose quartz from Brazil.

Högbomite is a closest-packed polysomatic mineral composed of spinel, T 2 M 4 O 8 , and nolanitelike, TM 4 O 7 (OH), modules where T stands for tetrahedrally and M for octahedrally coordinated cations. The modules are stacked in an ordered fashion in various ratios. Single-crystal X-ray diffraction for a 24R and a 10T polysome and structure modeling for a 12H and 14T polysome have been applied to characterize different stacking variants. Högbomite from a spinel-phlogopite schist at Corundum Creek (South Australia) with composition Mg 3.8 Fe 3.2 Zn 1.6 Ti 1.0 Al 18.3 O 38 (OH) 2 is a 10T polysome with a = 5.723(1), c = 23.026(4) Å, space group P3̄m1, Z = 1. This polysome with the general formula T 8 M 20 O 38 (OH) 2 is composed of an alternation of spinel (S) and nolanite-like (N) blocks stacked in the sequence NSSNS. Högbomite from a Fe-Ti deposit at Liganga (Tanzania) with composition Mg 13.5 Fe 5.6 (Zn,Mn,Ni) 0.2 Ti 4.7 Al 41.7 (Cr,Ga) 0.2 O 90 (OH) 6 is a 24R polysome with a = 5.7145(7), c = 55.056(5) Å, space group R3̄m, Z = 1. The structure with the general formula T 18 M 48 O 90 (OH)6 is composed of a periodic alternation of two S and two N blocks. The crystal structures of högbomite-12H, T 10 M 24 O 46 (OH) 2 , a = 5.7, c = 27.6 Å, space group P6 3 mc, Z = 1, and högbomite-14T, T 12 M 28 O 54 (OH) 2 , a = 5.7, c = 32.2 Å, space group P3̄m1, Z = 1, were modeled from the stacking principles of the known 6T, 8H, 10T, and 16H polysomes. The 12H and the 14T polysomes have stacking sequences NSSNSS and NSSSNSS, respectively.

Samples of wadsleyite (β-Mg 2 SiO 4 ) containing a range of dissolved water concentrations up to 1.5 wt% were synthesized at 1300 °C and 15 GPa. The samples were studied using 1 H MAS NMR and FTIR spectroscopy to determine the ordering of OH within the structure. As the 1 H NMR chemical shift and O-H stretching frequency are both known to be correlated with the O-H···O distances in silicate and other materials, the spectroscopic data were compared with O­···O distances calculated from the published crystal structure of hydrous wadsleyite. Using this approach we show that the hydroxyl ions in wadsleyite containing 0.8-1.5 wt% H 2 O are highly disordered, occupying at least 14 of the 17 possible O-H···O environments, including some with strong hydrogen bonding. In contrast, for low water concentrations (<0.4 wt%) the hydroxyl ions are less disordered, with four environments being much more abundant than the others are. Three of these environments appear to involve protonation of O1, in agreement with most previous suggestions, and the fourth is probably O2-H···O2. It is probable that the degree of disorder will increase with increasing temperature, so it should be taken into account when predicting phase equilibria involving hydrous wadsleyite and when extrapolating data on density and elastic properties from room temperature measurements.

Phase-X, a potassium di-magnesium acid disilicate, is a high-pressure synthetic compound-a potential K-bearing silicate in the mantle-with space group P6 3 cm (no. 185), a = b = 5.028(2) Å, c = 13.216(3) Å, V = 289.34 Å 3 , Z = 2. The structure has been determined with 1521 CCD measured intensities and refined by the least-square method to R = 0.0187. The structure is built up of octahedral MgO sheets and layers containing disilicate groups, Si 2 O 7 , (with distinct Si1 and Si2 tetrahedra linked by the apical O2 atom) alternating along the c axis. The octahedral sheet is based on a hexagonal closest-packed array of two layers of non-equivalent O atoms, O1 and O3; two-thirds of all edge-sharing M octahedra are filled. Within the framework of the Si 2 O 7 groups are channel structures parallel to [100], [010], and [110] that contain K atoms disordered in the middle of a large trigonal cavity (the A site). The FTIR spectrum in the OH stretching region shows a sharp peak at 3602 cm -1 due to OH- ordered in one anion site; the position of hydrogen, which operates in a charge-balancing substitution for the partial occupancy of the A site (K 1-x □ x ) A ↔ H x □ 1-x ) H , is undetermined. Densification in phase-X is affected by the greater compression of the empty octahedra in the octahedral layer and by constraining the trigonal A cavity containing the K atom to the size of the Si 2 O 7 disilicate group. This dense packing contributes to the relatively high zero-pressure calculated density of 3.38 g/cm 3 .

Two suites of garnets were examined by infrared absorption spectroscopy and depolarized Raman spectroscopy to determine if the Raman signal could be used as a quantitative measurement of OH content. To avoid the problems of determining the absolute Raman signal intensity, the integrated intensity of the OH bands was ratioed to the integrated intensity of silicate Si-O stretching bands and used as a proxy for the OH intensity. These were compared to the OH contents, expressed as wt% H 2 O, independently determined from infrared absorption. A somewhat useful trend developed between OH contents determined by Raman and IR for some grossular garnets with H 2 O contents less than 0.5%. Many colorless to near-colorless grossular garnets do not provide a useful Raman signal due to fluorescence, although other samples with H 2 O contents between 0.5 and 1.3 wt% and are strongly colored, fall far below the trend of the samples with lower H 2 O-contents. Spessartine-almandine garnets from the Rutherford no. 2 pegmatite all respond to the Raman experiment, but produce a confusing trend when compared to H 2 O. Coincidentally, a relatively smooth but decreasing trend was observed when the spessartine Raman OH intensity ratio was compared to the iron content. These observations suggest that Raman measurements by this method are not suitable, in general, for the determination of OH in garnets.

Raman spectra of syngenetic majoritic garnets were obtained from the Brazilian Sao-Luiz diamond suite. These garnets have a low majoritic content (<50%). Spectra were collected from (1) inclusions embedded in the diamonds and (2) inclusions extracted from the diamonds. From the latter, a clear relationship exists between the Raman spectra and the chemical analysis. All the majoritic garnets (Si > 3) have a characteristic signature in the frequency region of the SiO 4 -SiO 6 stretching vibrations. This vibration gives a broad peak between 800 and 900 cm -1 just before an intense band (900-930 cm -1 ) classically assigned to Si-O stretching vibrations of the SiO 4 tetrahedra in all known garnets. This broad band is also present in non-majoritic garnets (Si = 3) that result from the transformation, within the diamonds, of majoritic garnets into low-pressure garnet + pyroxene intergrowths. We show that the presence of significant TiO 2 content in these normal garnets leads to Raman spectra mimicking those of majoritic garnets. The occurrence of a broad shoulder between 800 and 900 cm -1 as well as broad bands near 960 and 1030 cm -1 in the Raman spectra are no longer diagnostic features indicative of a majoritic garnet when the TiO 2 content of the garnets is high (>1 wt%) as is the case for the Sao Luiz diamonds. The full width at half maximum (FWHM) of the strong Raman mode near 910 cm -1 is the only reliable signature that discriminates Si = 3 from Si > 3 garnets. The frequency of the diamond Raman mode used as an in-situ piezometer shows that the actual pressure of the garnet inclusions is between 0.5 and 0.8 GPa.

Raman spectra of Fe 2 O 3 were measured to 62 GPa in a diamond anvil cell. All group theoretically predicted Raman-active phonon modes were detected to 54 GPa. In addition, some high-pressure spectra show an IR-active E u mode (~660 cm -1 ), possibly induced by surface defects or stress. This mode is related by a factor of two to a mode at 1320 cm -1 . The assignment of the 1320 cm -1 mode has been controversial (two-magnon scattering or two-phonon scattering), and our observation supports the phonon assignment. All Raman-active phonons show nonlinear pressure-induced shifts. The mode Grüneisen parameters and their logarithmic volume dependences for two low-frequency phonons, A 1g and E g , become negative and infinite, respectively, near 50 GPa as a result of the instability of the corundum structure at this pressure. Using the observed Raman-active phonons together with acoustic phonons previously measured by ultrasonics, and Kieffer’s model, we calculate the phonon contribution to the thermodynamic parameters of hematite. Comparison with experimentally measured values allows an estimation of an upper bound to the magnon contribution to the heat capacity at ambient pressure. This increases continuously above the Morin temperature and reaches a maximum at the Néel temperature (~37%). The Raman spectra change dramatically at pressures greater than 54 GPa as indicated by the appearance of new peaks and a significant increase in background. Although direct structural analysis is not possible due to the low signal-to-background ratio and the lack of polarization information, we were able to examine the consistency of our Raman observation with the corundum-to-perovskite phase transformation using the results for an analog system: MgSiO 3 ilmenite (ordered corundum type) and perovskite. This analysis shows that observed new features in Fe 2 O 3 Raman spectra may not be consistent with the GdFeO 3 perovskite structure.

The (010) surfaces of forsterite (Fo 100 ) and natural forsteritic olivine (Fo 90 ) single crystals have been studied by Rutherford backscattering spectrometry (RBS). Spectra were collected either in channeling geometry with the incident beam parallel to the [010] zone axis, or in a random orientation. In both materials, two surface preparations were examined: (1) mechanical polishing and (2) chemical etching by hydrofluoric acid after mechanical polishing. Composition profiles extending to several hundreds of nanometers below the surface were probed by RBS in random mode. Simulations of all the spectra indicated constant major element compositions, equal to the bulk compositions of the crystals. Characteristics of the few top atomic layers were probed by RBS in channeling mode. The crystalline quality of the surface of chemically etched samples has been evaluated quantitatively and is shown to be much better than that of mechanically polished samples. Given the energy resolution, we estimate that the bulk crystalline quality is perturbed for more than 40 nm below the surface in the case of mechanical polishing whereas disorder is limited to a topmost layer thinner than 1 nm in the case of chemical etching. On average, for the chemically etched samples, less than one atom per [010] atomic row is displaced from its mean crystallographic position. The bulk stoichiometry is preserved in the topmost layer of pure forsterite whereas a slightly higher Fe/Mg ratio is detected at the surface of chemically etched Fo 90 . A method for the quantitative assessment of surface quality of olivine is thus proposed, constituting a useful preliminary step before any study of surface modifications of olivine subjected to various geological conditions.

The isothermal equation of state (EoS) of phase A, Mg 7 Si 2 O 8 (OH) 6 , has been determined using high-pressure single-crystal X-ray diffraction. A third-order Birch-Murnaghan EoS fit to pressurevolume data collected from room pressure and temperature to 7.6 GPa results in V 0 = 512.56(3) Å 3 , K 0 = 97.5(4) GPa, and K' = 5.97(14). Compression of the hexagonal (P6 3 ) structure is anisotropic with the c axis, which is perpendicular to the distorted close-packed planes of anions, approximately 23% less compressible than the a axis: K a = 90.1(5) GPa with K a ' = 5.4(2) and K c = 116.8(9) GPa with K c ' = 7.5(3). The bulk modulus of phase A is intermediate between those of brucite (Br) and forsterite (Fo) and less than those of hydroxylclinohumite and hydroxylchondrodite, in a manner that is entirely consistent with its water content and density in relation to the Fo-Br series of minerals.

Although it has been well accepted that F-Al linkages dominate in F-bearing alkali aluminosilicate glasses, the possibility of F-Si linkages remains unresolved. Recent 19 F NMR results have been interpreted as indicative of not only the presence of the Al-F-Na(n) linkages but also of the presence of Si-F-Na(n) linkages. High level ab initio NMR calculations were performed to investigate the nature of possible species contributing to the 19 F spectra. The B3LYP/6-31G* level was used to optimize structures and a scaling technique applied to the calculation of NMR properties using both HF and B3LYP methods at the 6-311+G(2df,p) level. The applicability of this scaling method was checked by comparing calculated results against experimental data on several crystalline materials; the results are improved over HF or DFT methods alone. The calculation results show that none of the F-Al species investigated can produce the -147 ppm peak of Schaller et al. (1992). However, several species in which F is bonded to fourfold-, fivefold-, and sixfold-coordinated Si produce a 19 F chemical shift of approximately -147 ppm. These results verify the assignments of this peak by Zeng and Stebbins (2000) to tetrahedral Si-F and NaSiF 6 -like species, and rule out many other possible species. Several additional species such as a fivefold coordinated Si-F species may further contribute to the -147 ppm 19 F peak.

The crystallographic and magnetic structures of Fe 0.99 O at 10 K have been determined by highresolution neutron powder diffraction. Fe 0.99 O is found to be monoclinic, space group C2/m, with unit-cell dimensions a = 5.2615(1), b = 3.0334(1), c = 3.0602 (1) Å, and β = 124.649(2)°. The Fe-O distances in the distorted FeO 6 octahedron are 2.154 Å × 4 and 2.165 Å × 2. The magnetic unit cell is obtained by doubling one of the crystallographic axes, a (magn) = a m , b (magn) = b m , and c (magn) = 2c m . The refined magnetic components at 8 K are M x = 2.7(1) μB, M y = -0.9(2) μB, and M z = 4.77(2) μB, with resultant M = 4.03(2) μB.

An examination of synthetic crystalline material by single crystal X-ray diffractometry reveals a unit-cell dimension that was previously ascribed to tincalconite. However, the best fit to the reflection data obtained at 291(2) K is obtained with composition Na 6 [B 4 O 5 (OH) 4 )] 3 ·8H 2 O. This formula is different from the Na 2 [B 4 O 5 (OH) 4 )]·3H 2 O previously assigned to tincalconite from an earlier crystallographic study. Additionally, our model fits best with non-crystallographic measurements and agrees with the analysis from a more recent low temperature crystallographic analysis. Tincalconite is best formulated as Na 6 [B 4 O 5 (OH) 4 )] 3 ·8H 2 O. It is trigonal, R32, a = 11.1402 (11), c = 21.207 (3) Å, Z = 3, R = 0.020 for 512 F o > 4σ (F o ) and 0.021 for all 529 data at 291(2) K.

Despite the importance of the chloride ion in magmas and the fluids that separate from them, very little is known about atomic-scale structural environments for Cl- in silicate glasses. We present here the first solid-state 35 Cl NMR data for Cl in silicate and aluminosilicate glasses, made possible by the availability of very high (14.1 to 18.8 Tesla) magnetic fields. We find that 35 Cl has a wide range in chemical shifts that correlate well with cation-Cl distance and thus contain considerable structural information. In general, Cl is coordinated primarily by network-modifying alkali or alkaline earth cations, and we see no evidence for Al-Cl bonding.

Significant amounts of boron in both trigonal and tetrahedral coordination have been found through 11 B magic-angle-spinning (MAS) nuclear magnetic resonance (NMR) spectroscopy of natural olenite (aluminum-rich tourmaline) from Stoffhütte, Koralpe, Styria, Austria. The 11 B MAS NMR spectrum consists of a superposition of two quadrupolar MAS peaks. A broad peak with δ iso = 18.3 ppm, ν Q = 1410 KHz, and η = 0.11 demonstrates the trigonal-planar environment of the BO 3 group (relative area ratio = 80%). The narrow symmetrical peak (δ iso = 0.0 ppm, ν Q = 300 KHz, η = 0.00) represents tetrahedral BO 4 groups (relative area ratio = 20%). An infrared spectrum shows hydroxyl stretching bands at low wavenumbers indicating that underbonded O atoms of the hexagonal ring (due to a partial replacement of [4] Si 4+ by [4] B 3+ ) form relatively strong hydrogen bonds with the protons of the hydroxyl groups. A 29 Si MAS NMR spectrum shows a peak consisting of a main signal at -90 ppm and a shoulder at about -85 ppm. The main signal originates from Si atoms connected (via oxygen bridges) to two other Si atoms in the hexagonal ring, and the minor signal is from Si atoms connected to one Si and one B atom. No signal corresponding to [6] Si was detected in this natural olenite sample.