Volume 87, Issue 5-6

The Lower Williams Lake tailings, which resulted from U-milling operations during the late 1950s and early 1960s, cover an area of 2 ha in a small bog under partial water cover. The tailings are underlain by a sand unit containing decaying organic material above the natural base consisting of sand, till, and gravel. The tailings are composed predominantly of quartz, muscovite, K-feldspar, plagioclase, and clinochlore. Residual pyrite grains, displaying angular to subangular particles measuring less than 1 to 250 μm, occur in concentrations ranging from trace to approximately 6 wt%. Framboidal pyrite has formed within the tailings basin in association with organic-rich material. The appearance of framboidal pyrite in the tailings indicates a reversal of the oxidation process and reprecipitation of Fe sulfides. In addition, the tailings include calcite, calcian siderite, Feoxyhydroxides, and Fe-Si-oxyhydroxides as secondary precipitates and replacement products. Groundwaters in the tailings and the underlying units are saturated with respect to gypsum and siderite. These and the other saturation indices calculated for calcite, goethite, and barite are consistent with the secondary mineralogy of the tailings. Conditions promoting the formation of pyrite can be described as neutral to weakly alkaline and reducing assisted by microbial activity. This study provides the first account of secondary carbonates and framboidal pyrite in the Elliot Lake tailings. The existing environmental conditions at the site are favorable for the desired site rehabilitation.

Hydrous MgSiO 3 akimotoite (ilmenite-type structure), containing 350 ppm wt H 2 O as hydroxyl, was synthesized at 19 GPa and 1300 °C in a multi-anvil apparatus. Polarized Fourier-transform infrared spectra show bands at 3300 and 3320 cm -1 due to hydroxyl groups oriented nearly perpendicular to the c axis as well as a strong band at 3390 cm -1 caused by OH molecules aligned nearly parallel to c. Unpolarized infrared measurements performed up to 13.8 GPa in a diamond anvil cell show that the bands shift to lower frequencies with increasing pressure, which is consistent with hydrogen bonding. At ~9 GPa, the bands initially at 3300 and 3320 cm -1 merge, suggesting that these two bands represent the same type of hydroxyl group. The compression behavior of the infrared bands shows that the pressure derivative of the frequency, dʋ/dP, decreases with increasing initial wavenumber, ʋ io , and this decrease is not linear. These observations are consistent with the compression mechanism of akimotoite, as previously published. The data indicate that the relative compressibility of the hydrogen bonds within a structure (i.e., the sign of dʋ/dP vs. ʋ io ) is governed by a combination of the directionality of the hydrogen bonds and the anisotropy of compression, whereas the rate of compression (i.e., the slope of dʋ/dP vs. ʋ io ) is a function of the bulk modulus.

Using IR radiation from a synchrotron source, high-quality absorbance spectra were obtained from polycrystalline powder of chloritoid (cld) from ambient conditions up to pressures of 10 GPa over 50 to 4000 cm -1 . The idealized chemical composition of the chloritoid group is M 2 Al 4 O 2 (SiO 4 ) 2 (OH) 4 where M = Fe or Mg in our experiments. All of the 42 expected fundamental IR modes were observed. These data, combined with the response of the IR bands to substitutions of Fe for Mg, and of D for H, constrained the band assignments. Heat capacity (C P ) and entropy (S o ) for the triclinic and monoclinic polymorphs of Fe- and Mg-cld were calculated from the Kieffer-type model, using our detailed band assignments. The calculated heat capacity and entropy for the monoclinic and triclinic polymorphs differ negligibly. The results at temperatures above 298 K are described by the following polynomial expressions in J/(mol·K): C P = 7.835 · 10 2 - 5.170 · 10 3 T -0.5 - 1.648 · 10 7 T -2 + 1.934 · 10 9 T -3 for Mg-cld and C P = 7.848 · 10 2 - 5.185 · 10 3 T -0.5 - 1.548 · 10 7 T -2 + 1.783 · 10 9 T -3 for Fe-cld. At room temperature, So = 293 J/mol·K for Mg-cld and 335 J/mol·K for Fe-cld. These values differ somewhat from entropy estimated from various internally consistent databases (-3 to -9% for Mg-cld and -9 to +5% for Fe-cld). However, using our new So and CP values in conjunction with the enthalpy of formation, H f = -7101 kJ/mol for Mg-cld or H f = -6422 kJ/mol for Fe-cld (estimated in this study), we can closely reproduce the experimental data for the reactions Mg-chloritoid + talc = clinochlore + kyanite (Chopin 1985) and Fe-chloritoid = almandine + diaspore + water (Vidal et al. 1994).

The difference in the local environment of CO 3 2- , NO 3 - , SO 4 2- , and ClO 4 - in Mg/Al-hydrotalcite compared to the free anions was studied by infrared and Raman spectroscopy. In comparison to free CO 3 2- a shift toward lower wavenumbers was observed. A band around 3000-3200 cm -1 has been attributed to the bridging mode H 2 O-CO 3 2- . The IR spectrum of CO 3 - hydrotalcite clearly shows the split ν 3 band around 1365 and 1400 cm -1 together with weak ν 2 and ν 4 modes around 870 and 667 cm -1 . The ν 1 mode is activated and observed as a weak band around 1012 cm -1 . The Raman spectrum shows a strong ν 1 band at 1053 cm -1 plus weak ν 3 and ν 4 modes around 1403 and 695 cm -1 . The symmetry of the carbonate anions is lowered from D 3h to C 2s resulting in activation of the IR inactive ν 1 mode around 1050-1060 cm -1 . In addition, the ν 3 shows a splitting of 30-60 cm -1 . Although NO 0 hydrotalcite has incorporated some CO 3 2- the IR shows a strong ν 3 mode at 1360 cm -1 with a weak band at 827 cm -1 , and the ν 4 band is observed at 667 cm -1 , although it is largely obscured by the hydrotalcite lattice modes. The Raman spectrum shows a strong ν 1 mode at 1044 cm -1 with a weaker ν 4 band at 712 cm -1 . The ν 3 mode at 1355 cm -1 is obscured by a broad band due to the presence of CO 3 2- . The symmetry of NO 3 - did not change when incorporated in hydrotalcite. The IR spectrum of SO 4- hydrotalcite shows a strong ν 3 at 1126, ν 4 at 614 and a weak v 1 mode at 981 cm -1 . The Raman spectrum is characterized by a strong ν 1 mode at 982 cm -1 plus medium ν 2 and ν 4 bands at 453 and 611 cm -1 ; ν 3 cannot be identified as a separate band, although a broad band can be seen around 1134 cm -1 . The site symmetry of SO 4 2- is lowered from T d to C 2ν . The distortion of ClO 4 - in the interlayer of hydrotalcite is reflected in the IR spectrum with both ν 3 and ν 4 bands split around 1096 and 1145 cm -1 and 626 and 635 cm -1 , respectively. A weak ν 1 band is observed at 935 cm -1 . The Raman spectrum shows a strong ν 1 mode at 936 cm -1 plus ν 2 and ν 4 bands at 461 and 626 cm -1 , respectively. A ν 3 mode cannot be clearly recognized, but a broad band is visible around 1110 cm -1 . These data indicative a lowering of symmetry from T d to C s .

The effects of Fe oxidation state on the infrared (IR) spectra of dioctahedral smectite were studied using a purified and Na + -saturated fraction of the Garfield nontronite reference clay. The nontronite was first reduced with sodium dithionite for a period of 10 to 240 min to obtain various Fe reduction levels. The reduced samples were then reoxidized by bubbling O 2 through the suspensions for 8 to12 h. IR spectra were collected on the initially unaltered, the reduced, and the reduced-reoxidized samples. After reduction, changes were observed in the spectral regions of O-H stretching, O-H deformation, and Si- O stretching, indicating that the clay structure was significantly modified beyond merely a change in Fe oxidation state. Furthermore, a new component band in the O-H stretching region of the reduced samples exhibited a pleochroic effect, indicating the possible existence of trioctahedral domains. A large (up to 43 cm -1 ) downward shift of the main Si-O stretching band of the reduced samples was also observed. Such a large shift indicates that the change in Fe oxidation state in the octahedral sheet strongly affects the structural properties of the tetrahedral sheet, which might further affect physical and chemical properties of the mineral surface. The spectral differences across all three studied regions between unaltered and reoxidized samples after up to 240 min of reduction indicated that the redox process involving sodium dithionite is in some respects irreversible, even though virtually all structural Fe 2+ can be reoxidized.

The crystal structure of katoite hydrogarnet has been refined at 0.0001, 2.15, 4.21, 5.09, 6.00, 7.09, and 7.78 GPa from single-crystal X-ray diffraction data using a 4:1 methanol:ethanol mixture as pressure medium in a Merrill-Bassett diamond-anvil cell. Below ~5 GPa, the katoite structure has Ia3d symmetry and compresses by bond shortening rather than bond bending, in agreement with recent quantum mechanical calculations. An unconstrained third-order Birch-Murnaghan fit to the unit-cell volumes and pressures for Ia3d symmetry gave the following equation of state parameters: V o = 1987.6(1) Å 3 , K o = 58(1) GPa and K' = 4.0(7). Above this pressure, the structure undergoes a phase transition to space group I4̅3d, a non-centric subgroup of Ia3d. In the I4̅3d structure, there are two non-equivalent (O 4 H 4 ) groups. Both the Ca and Al atoms are displaced along a relative to their positions in Ia3d. It is proposed that compression of the short H-H distance between (O 4 H 4 ) groups destabilizes the structure and may initiate the observed phase transition. Corroboration of this model will require accurate information on the hydrogen atom positions at pressures above 5 GPa.

A high-temperature in situ single-crystal X-ray diffraction study was performed from room T to 1150 °C on two crystals of Fe-free P2 1 /c clinopyroxenes of composition Ca 0.15 Mg 1.85 Si 2 O 6 [cell parameters at room T: a = 9.651(2) Å, b = 8.846(2) Å, c = 5.202(1) Å, b = 108.38(2)°, V = 421.4 (2) Å 3 ] synthesized by isothermal annealing for 624 h at T = 1370 °C, P = 1 atm. A first order P2 1 /c-C2/c phase transition was found slightly below 1000 °C [T c = 926(39) °C]. The transition was revealed by discontinuous changes in intensities and cell parameters. Prolonged heating at high temperature induced a non-reversible increase in the transition temperature up to more than 1150 °C, without apparent changes in the order of the phase transition. Coupling with strain due to incipient exsolution in a formerly almost defect-free sample is suggested to be responsible for increase in T c . TEM observations of a sample from the same starting material after further annealing for 72 h at T = 1050 °C, P = 1 atm are consistent with the proposed incipient exsolution model. Annealing was found to induce the formation of a mottled texture oriented parallel to (101). Results from structure refinement of data collected below the transition at T = 25, 500, 650, 800, and 1000 °C showed only minor changes in the chain configurations, which are highly differentiated up to 1000 °C, confirming the strong first-order character of the transition.

Pseudobrookite-type MgTi 2 O 5 (karrooite) is a synthetic crystalline phase with the Bbmm structure and a component in orthorhombic oxide solid solutions, R 2+ Ti 2 O 5- R 3+ 2 TiO 5 , which are present as accessory minerals in lunar and terrestrial rocks. In this study, we present a model for the molar Gibbs free energy of MgTi 2 O 5 as a function of T, P, and the order parameter s = X M1 Mg - 2X M2 Mg , (-1 ≤ s ≤ 1). We describe the molar Gibbs free energy, (G̅), with the equation: G̅ = g̅ 0 + g̅ 1 ·(1 - s) + 3/2g̅ 2 ·(1 - s 2 ) - T·S̅ config , and take parameter go to represent the molar Gibbs free energy of ordered MgTi 2 O 5 (s = 1), whereas parameters g 1 and g 2 may represent cation-disorder contributions. We used powder and single-crystal X-ray diffraction, and high-temperature relative enthalpy data, to calibrate the disorder contribution to the volume (b o = 7.3822·10 -3 J/bar), and the model parameters g 1 (7370.8 J/mol) and g 2 (3576.1 J/mol), and heat capacity and volume equation coefficients. We also optimized standard state thermodynamic data from the elements for ordered MgTi 2 O 5 , (ΔH 0 = - 2 498 515.28 J/mol, S 0 = 149.55 J/(mol·K), ΔG 0 = -2 362 181.72 J/mol, V 0 = 5.445 J/bar) consistent with the model parameters and equations, the thermodynamic data in QUILF, and phase-equilibrium experiments involving MgTi 2 O 5 , geikelite, rutile, orthoenstatite, and forsterite in the range 973 to 1673 K and 0.0001 to 2.0 GPa. Finally, we investigate theoretically the stability of MgTi 2 O 5 (karrooite) with respect to geikelite, rutile, diopside, enstatite, and forsterite in the CaO-MgO-TiO 2 -SiO 2 system. We find that diopside- and titanite-bearing reactions require extremely high temperatures, and are thus not stable with respect to liquid. The inferred phase relations can be of help in understanding the stability of MgTi 2 O 5 with respect to rutile, geikelite, forsterite, and orthoenstatite, and by extrapolation that of armalcolite relative to rutile, ilmenite, olivine, and orthopyroxene in terrestrial mantle rocks and high-Ti lunar basalts.

Ab initio, molecular orbital calculations were carried out to study the energetic stability and NMR characteristics of “triclusters” of composition [T(XO)(OH) 2 ]O[Al(OH) 3 ][T(OH) 3 ] (X = no atom, H + , Li + , Na + , K + , Mg 2+ , or Ca 2+ ; T = Si 4+ or Al 3+ ). These triclusters consist of a central oxygen atom in trigonal coordination ( III O) surrounded by three tetrahedrally coordinated, network-forming cations (Al and Si). All tricluster configurations modeled are predicted to be dynamically stable except for one containing three Al 3+ cations. Tricluster viability was further tested by computing energetic stabilities relative to Al(OH) 3 and dimers with non-bridging oxygen atoms (NBO). With these species as reactants and T = Si 4+ , calculated reaction energies vary from -87 kJ/mol with X = H + , to -253 kJ/mol with X = Mg 2+ . For the case of X = Ca 2+ , where structures with 1 Si and 2 Al were also modeled, the reaction energy was found to be over 50 kJ/mol more negative in the latter case (- 278 kJ/mol compared to -222 kJ/mol). However, when the energy of the triclusters [Si(Na + O)(OH) 2 ]O [Al(OH) 3 ][Si(OH) 3 ] and [Al(Ca 2+ O)(OH) 2 ]O[Al(OH) 3 ][Si(OH) 3 ] was calculated relative to isochemical chains of three tetrahedra without NBO or III O, the triclusters were predicted to be +124 and +142 kJ/mol less stable than the corresponding trimer chains. Although these latter computed energy changes do not preclude formation of triclusters, they imply that such species should occur only in minor concentrations. To help identify III O within aluminosilicate glasses, predictions of 17 O NMR chemical shifts (⃞ 17 O), quadrupolar coupling constants (QCCs), and asymmetry parameters (⃞) are also presented. When compared with predictions of 17 O NMR parameters in a number of molecules representing Si and Albearing Q 4 species, it is found that the calculated parameters for III O are generally within the range of normal bridging oxygen atoms (BO). Thus, if III O are a minor percentage of oxygen atoms within a glass, we conclude that their identification may be difficult using NMR spectroscopy.

Atomistic computer modeling methods are used to investigate the surface structures and stabilities of magnesite MgCO 3 and dolomite MgCa(CO 3 ) 2 in comparison with calcite, CaCO 3 . The surfaces are generally more stable than magnesite, but less so than calcite. Upon hydration, the magnesite {101̅4} surface becomes more stable than either calcite or dolomite, due to the less favorable interactions between dolomite surface ions and adsorbing water molecules. The high affinity of the surface magnesium ions for adsorbing water molecules, shown by the large average hydration energy over all surfaces for magnesite (134 kJ/mol), is negated in the case of dolomite by the inaccessibility of the surface magnesium ions, due to surface relaxations and rotation of the carbonate groups. In addition, the larger lattice spacing of dolomite compared to magnesite disrupts the extensive network of hydrogen bonding between the adsorbed water molecules that are present on the latter, leading to similar average hydration energies for calcite, dolomite, and magnesian calcite surfaces (80-101 kJ/mol). Not only are the dolomite planes under aqueous conditions less stable than their calcite counterparts, dolomite-like surfaces of disordered magnesian calcites are calculated to be more stable than either pure calcite or dolomite surfaces. Extensive molecular dynamics simulations of the growth of CaCO 3 and MgCO 3 at two experimentally observed calcite steps show that, on thermodynamic grounds, Mg is easily incorporated into a growing calcite crystal, but will then inhibit incorporation of further Ca. Furthermore, although the calculations suggest a small energetic advantage (~10 kJ/mol) to ordering of the Ca and Mg ions on the obtuse growth step, a disordered arrangement is favored at the acute growth step and incorporation of Mg is always energetically more favorable than Ca, in accord with the occurrence of highly magnesian calcites in nature.

The Hutter Mine locality, Pittsylvania County, Virginia, is a metamorphosed magnetite deposit, with substantial development of subsidiary manganoan marble, that occurs within Latest Precambrian or Early Paleozoic sillimanite-grade pelitic schists. Manganese oxides and spinels at the Hutter Mine include manganosite (MnO) (coexisting with hausmannite and jacobsite) as well as spinels rich in jacobsite (FeMn 2 O 4 ), magnetite (Fe 3 O 4 ), and galaxite (MnAl 2 O 4 ), and a variety of intermediate solid solutions between these three end-members. Several samples contain spinels that exhibit substantial miscibility along the jacobsite-galaxite and jacobsite-magnetite joins. Magnetite-galaxite solid solution is, by comparison, very limited. Coexisting manganoan spinels within the jacobsitegalaxite- magnetite ternary system include jacobsite-rich varieties with galaxite <65 (normalized to glx + mag + jac = 100) that coexist with Mg-Zn-bearing galaxite-rich spinel with galaxite >75. However, the wide range of spinel compositions at the Hutter Mine largely reflects compositional variability in the host rock. In a skarn reaction zone between Fe-rich, quartz-bearing amphibolites and Si-poor, Mn-rich marbles, the galaxite content of spinel drops from 60% to near zero as silica activity increases over a 5 mm interval. In this same reaction zone, magnetite content of spinel increases from about 10 to 95%, but over a narrower interval (about 2 mm). Total variation in spinel composition in this reaction zone is nearly the same as that seen over the entire suite of Hutter Mine samples. Both regional metamorphic geology and thermobarometry on local pelite samples indicates that T max at the Hutter Mine was 550-600 °C. Manganosite formed by the decarbonation of Mn-rich carbonate in the presence of a CO 2 -poor (X CO2 ≤ 0.01) fluid having log a SiO2 < -3.0. Oxygen fugacity in the manganosite-bearing sample was buffered by coexisting manganosite and hausmannite, placing f O2 within the magnetite stability field at peak T. This result is consistent with the occurrence of magnetite as the principal ore at Hutter. The extensive miscibility observed along the jacobsite-galaxite join requires reexamination of miscibility gaps proposed in previous studies. We suggest that the wide compositional gaps found in previous studies reflect a variety of chemical factors of which silica activity is the most critical. In particular, the large range of silica activities observed in Hutter Mine rocks stabilizes spinels with a wide range in galaxite content. The crests of both the jacobsite-galaxite and jacobsite-magnetite two-phase regions appear to occur at relatively low temperatures, probably below 600 °C.

We document the first reported igneous occurrence of the orthoamphibole, gedrite, from a troctolitic cumulate in the Eriksberg gabbro, Sweden. The gedrite is sodic and coexists in textural equilibrium with plagioclase, orthopyroxene, Ca-amphibole, and Na-rich phlogopite. The gedrite crystallized from evolved interstitial liquid that was buffered by cumulus olivine. Geothermometry indicates that the assemblage equilibrated at around 900 °C and 4-6 kbar. Gedrite shows a convex- downward REE-profile, with a positive Eu anomaly. The rare earth element concentrations are ~0.2 to 0.8 times chondritic abundances and much lower than in coexisting Ca-amphibole. Laser ablation ICP-MS analyses show that the gedrite has much lower concentrations of Zr, Nb, Th, and U, and lower concentrations of Sr, Ba, Rb, and Pb relative to co-crystallizing Ca-amphibole. However, nickel and cobalt partition preferentially into gedrite. Comparison of Zr and Nb concentrations in early crystallized Ca-amphibole with gedrite and associated Ca-amphibole reveals that X mg of amphibole is not the best criteria to use for estimates of (D Ampf/L Nb/Zr ) for cumulates where olivine is buffering the Mg/(Mg + Fe).

Eleven clinopyroxenes from the Sludyanka Crystalline Complex in Russia belonging to the ternary join NaVSi 2 O 6 -NaCrSi 2 O 6 -CaMgSi 2 O 6 (natalyite-kosmochlor-diopside) were studied by means of X-ray single crystal diffractometry and electron probe microanalysis. The crystal chemical data show that the T site is almost completely occupied by Si, so that the Na (V 3+ ,Cr 3+ ) → Ca Mg substitution mechanism ensures charge balance. Changes in M1 site geometry are explained by the aggregate ionic radius, and are influenced by Mg occupancy and V 3+ /(V 3+ + Cr 3+ ) ratio. The M2 site geometry depends both on Na content and on the (V 3+ ,Cr 3+ ) → Mg substitution in M1 site. Changes in M2-O3c1 bond length are mainly related to Na content, whereas the longest M2-O3c2 bond lengths are significantly affected by the V 3+ /(V 3+ + Cr 3+ ) ratio of the M1 site. The T site geometry is affected by chemical and geometrical variations at the M1 and M2 sites, principally the M1 site occupancy.

Single crystals of a synthetic apatite-like phase with composition Na 6.9 Ca 3.1 (SO 4 ) 6 (OH) 1.1 were grown under hydrothermal conditions. This compound crystallizes in the hexagonal space group P6̅ (a = 9.4434(13) Å, c = 6.8855(14) Å, Z = 1). The structure was solved by direct methods, and subsequently refined using 655 independent reflections (R1 = 0.0542). The chemical composition and the unit cell parameters indicated a close structural relationship with the mineral cesanite. A reinvestigation of the mineral showed that the natural and the synthetic phases are isostructural. Small differences result from the incorporation of both H 2 O and (OH) - into the structure of cesanite. Observed systematic absences revealed that the space group P6 3 /m allocated to cesanite in earlier studies is incorrect. The crystal structure of a cesanite with composition Na 7.0 Ca 3.0 (SO 4 ) 6 (OH) 1.0 (H 2 O) 0.8 was successfully refined in space group P6̅ (a = 9.4630(8) Å, c = 6.9088(5) Å, Z = 1, R1 = 0.0468 for 720 independent reflections [I > 2σ(I)]). The symmetry reduction can be attributed to ordering of the Na and Ca atoms among four symmetrically independent cation sites.

The crystal structure of leightonite, K 2 Ca 2 Cu(SO 4 ) 4 ·2H 2 O, C2/c, a = 11.654(2), b = 7.497(1), c = 10.097(1) Å; b = 125.21(1)°, V = 720.8(2) Å 3 , Z = 2 has been solved by direct methods and refined to R = 3.90% for 1564 F o > 4σ(F o ), using MoKα X-ray data from a crystal twinned on {201̅}. Structural sub-units [Ca(SO 4 ) 2 ] 2 - formed by one CaO 8 polyhedron and two opposite-sided SO 4 tetrahedra are linked by edge sharing. These sub-units are linked to each other by corner sharing to form a three-dimensional framework with channels, where the Cu atoms are located. The framework of CaO 8 polyhedra and SO 4 tetrahedra exhibits a perfect orthorhombic symmetry whereas the copper atoms located at the Cu1 and Cu2 sites are not equivalent because they have different partial occupancies (0.37 and 0.13, respectively). Both Cu1 and Cu2 are coordinated by O atoms to form two rhombically elongated octahedra (2 + 2 + 2 Jahn-Teller distortion). K and Ow are disordered on the same site. The crystal structure of leightonite closely resembles that of the triclinic polyhalite, K 2 Ca 2 Mg(SO 4 ) 4 ·2H 2 O, with the main difference being the different distribution of Cu in leightonite with respect to Mg in polyhalite.

Neustädtelite and cobaltneustädtelite, two new minerals related to medenbachite, were found on samples from the dumps of the Güldener Falk mine near Schneeberg-Neustädtel, Saxony, Germany. The general appearance of the two new minerals is very similar: small tabular crystals up to 0.2 mm in diameter, transparent to translucent, with a brown color and a light brown streak; the lustre is adamantine. Both minerals are optically biaxial negative, 2V = 65(5)°, n x = 2.02(2), n y = 2.09 (calc.), n z = 2.12(2); pleochroism is strong with X = brown to opaque, Y = yellow, Z = pale yellow. Mohs’ hardness is 4.5. The cleavage parallel to {001} is good. The chemical compositions were derived by means of electron-microprobe analyses. Average contents for neustädtelite/cobaltneustädtelite are (in wt%): Bi 2 O 3 52.58/51.54, PbO 0.08/0.08, CaO 0.15/0.32, Fe 2 O 3 13.92/10.90, Al 2 O 3 0.29/0.07, CoO 3.35/5.47, NiO 0.34/1.61, ZnO 0.09/0.39, CuO 0.07/0.00, As 2 O 5 26.82/25.91, P 2 O 5 0.23/0.43, H 2 O (calc.) 2.56/3.01, total 100.48/99.73. Mössbauer spectra of cobaltneustädtelite and medenbachite confirmed that all of the iron is trivalent. Based on 12 O atoms, the empirical formulae for the neustädtelite and cobaltneustädtelite type materials are (Bi 1.94 Ca 0.02 ) ∑1.96 Fe 1.00 (Fe 0.50 Co 0.38 Ni 0.04 Al 0.05 Zn 0.01 Cu 0.01 ) ∑0.99 [(OH) 2.44 O 1.40 ] ∑3.84 [(AsO 4 ) 2.01 (PO 4 ) 0.03 ] ∑2.04 and (Bi 1.91 Ca 0.05 ) ∑1.96 Fe 1.02 (Co 0.63 Fe 0.16 Ni 0.19 Zn 0.04 Al 0.01 ) ∑1.03 [(OH) 2.88 O 1.12 ] ∑4.00 [(AsO 4 ) 1.95 (PO 4 ) 0.05 ] ∑2.00 , respectively. As derived from chemical analyses and crystal-structure investigations the ideal end-member compositions are Bi 2 Fe 3+ Fe 3+ O 2 (OH) 2 (AsO 4 ) 2 (neustädtelite) and Bi 2 Fe 3+ Co 2+ O(OH) 3 (AsO 4 ) 2 (cobaltneustädtelite). Extensive solid solution is observed between these two minerals. Neustädtelite and cobaltneustädtelite crystallize in space group P1̅; the cell parameters refined from powder data are a = 4.556(1)/9.156(1), b = 6.153(2)/6.148(1), c = 8.984(2)/9.338(1) Å, α = 95.43(2)/83.24(1), β = 99.22(2)/70.56(1), γ = 92.95(3)/86.91(1)°, V = 246.9/492.2 Å 3 , Z = 1/2, density (calc.) 5.81/5.81 g/cm 3 . Structure investigations were performed using single-crystal X-ray data. In both minerals edge-sharing alternating Fe 3+ Ø 6 and (Fe 3+ ,Co 2+ )Ø 6 /(Co 2+ ,Fe 3+) Ø 6 octahedra (Ø = O,OH) form chains parallel to [010] that are corner-linked by arsenate tetrahedra to layers parallel to (001). The Bi atoms are linked by O atoms to form columns parallel to [100]; these are sandwiched between layers of composition [6] M 2 (OH) 2 (AsO 4 ) 2 (M = Fe 3+ ,Co 2+ ). In neustädtelite the Bi atoms are site disordered; in cobaltneustädtelite half of the Bi atoms are ordered and half are on a split position. The partial ordering is induced by the presence of three OH groups, as compared to two in neustädtelite. A structural reinvestigation of medenbachite, Bi 2 Fe 3+ (Cu 2+ ,Fe 3+ )(O,OH) 2 (OH) 2 (AsO 4 ) 2 , proved isotypy with cobaltneustädtelite; the new cell parameters for medenbachite (refined from X-ray powder data) are: a = 9.162(2), b = 6.178(1), c = 9.341(2) Å, α = 83.50(2), β = 71.04(2), γ = 85.15(2)∞, V = 496 Å 3 , Z = 2.

Two new minerals of the hellandite group were found within alkali-syenitic ejecta enclosed in pyroclastic formations of the Roman Comagmatic Province (Latium, Italy). Mottanaite-(Ce) [ideally X Ca 4 Y (CeCa) Z Al T Be 2 (Si 4 B 4 O 22 ) W O 2 ] and ciprianiite [ideally X Ca 4 Y [(Th,U)(REE)] Z Al T ⃞ 2 (Si 4 B 4 O 22 ) W (OH, F) 2 ] occur as transparent, brown-colored, tabular euhedral crystals in miarolitic cavities and voids of the ejecta, which consist mainly of sanidine and plagioclase (An ranging from 20 to 80%), with minor amounts of feldspathoid, clinopyroxene and/or clinoamphibole, magnetite, titanite, and zircon. Locally, accessory minerals include britholite-(Ce), baddeleyite, phosphate to silico-phosphate phases close in composition to the brabantite-cheralite series, thorite, fluorite, danburite, and vonsenite. The genesis of the new hellandite end-members can be related to late-stage post magmatic hydrothermal fluids enriched in Zr, Ti, REEs, and actinide elements. Both mottanaite-(Ce) and ciprianiite have a vitreous luster and are non-fluorescent. Cleavage is absent in mottanaite-(Ce), fair to good in ciprianiite, {100}. Twinning is frequently observed in ciprianiite. Due to the strong intra-crystalline chemical zoning and twinning, physical properties could be measured only for mottanaite-(Ce). D meas is 3.61(4) g/cm 3 , D calc is 3.88 g/cm 3 . Mottanaite- (Ce) is biaxial negative, with α = 1.680(5), β = 1.694(2), γ = 1.708(5); 2V meas ~90°. Both minerals are monoclinic, space group P2/a, Z = 2. Unit-cell parameters for the crystals studied are: a = 19.032(9) Å, b = 4.746(3) Å, c = 10.248(5) Å, β = 110.97(5)°, V = 864.3(8) Å 3 for mottanaite-(Ce), and a = 19.059(5) Å, b = 4.729(1) Å, c = 10.291(4) Å, β = 111.33(2)°, V = 864.0(5) Å 3 for ciprianiite. Single-crystal structure refinement confirmed the presence of a further distorted tetrahedral site which was first detected in a hellandite-(Ce) sample from Latium (Oberti et al. 1999). This site is occupied by Be (±Li) in stoichiometric mottanaite-(REE), whereas it mainly hosts hydrogen (bonded to the O5 oxygen atom) in ciprianiite and hellandite-(REE); solid solution between the end-members is possible, as shown by the studied samples. The chemical composition of the refined crystals was obtained by combining EMPA (for medium-Z elements) and SIMS analyses (for low- and high-Z elements); their results are in excellent agreement with the chemical information obtained from the structure refinements. The crystal-chemical formulae of the crystals studied, recalculated on the basis of 24 anions, are M3,4 Ca 4 M2 [REE 1.45 Ca 0.37 (Th,U) 4+ 0.17 Y 0.01 ] ∑=2 M1 (Al 0.50 Fe 3+ 0.38 Mg 0.03 Ti 4+ 0.07 ) ∑=0.99 T (Be 1.18 Li 0.02 ⃞ 0.37 )B 3.99 Si 3.98 O 22 O5 (O 2- 1.04 F 0.53 OH 0.43 ) Σ=2 for mottanaite-(Ce); and M3,4 Ca 4 M2 [REE 3+ 0.72 (Th,U) 4+ 0.66 Ca 0.60 Y 0.02 ] Σ=2 M1 (Al 0.48 Fe 3+ 0 .38 Ti 4+ 0.10 Mg 0.05 Mn 3+ 0.02 ) Σ=1.03 T (Be 0.82 ⃞ 0.60 Li 0.04 ) B 4.00 Si 4.00 O 22 O5 (O 2- 0.97 OH 0.54 F 0.49 ) Σ=2 for ciprianiite.

Detailed X-ray single-crystal structure refinements and complete (SIMS + EMPA) microprobe chemical analyses of a series of non-metamict samples of hellandite recently found in Latium (Italy), together with a critical re-evaluation of the existing data on hellandite, allow a better understanding of the crystal-chemistry of the hellandite group. Relative to the crystal structure determined by Mellini and Merlino (1977) for a Y-rich sample from Predazzo (Italy), a new tetrahedrally coordinated site has been detected; this may be fully, or in part, occupied by Li and Be. These cations occur at the center of the tetrahedral cavity where the H atom, which is bonded to the O5 O atom, protrudes; thus the H content in hellandite is constrained to values ≤ 2 - (Li + Be + F). A new general formula for hellandite-group minerals is proposed: X 4 Y 2 ZT 2 [B 4 Si 4 O 22 ]W 2 , where X = Na, Ca, Y, LREE 3+ at the eightfold-coordinated M3 and M4 sites; Y = Ca, Y, HREE 3+ , Th 4+ , U 4+ at the eightfold-coordinated M2 site; Z = Al, Mn 3+ , Fe 3+ , Ti 4+ at the octahedral M1 site; T = ⃞ (vacancy), Li, Be at the new tetrahedrally coordinated site; and W = OH, F, O 2- at the O5 site. Eight root end-member compositions were identified; four of these correspond to known compositions: hellandite-(REE), tadzhikite-(REE), and two new minerals, mottanaite-(Ce) and ciprianiite, which are described in a companion paper (Della Ventura et al. 2002). The root-name tadzhikite must be used for samples with Ti 4+ > 0.5 apfu, with no reference to the OH content. Hellandite is an REE mineral. Rare-earth elements occur at two distinct sites (M2 and M4), with a strong preference for M2 (particularly HREE). Thus, one or two Levinson modifiers can be added to the root name to correctly describe the species. A sequence of incorporation based on crystalchemical arguments is provided and allows evaluation of the site populations for nomenclature purposes. This redefinition of the hellandite group and the new nomenclature rules have been approved by the IMA Commission on New Minerals and Mineral names (code 00-F).

Baumstarkite is a new mineral found coating miargyrite from the San Genaro mine, Huancavelica Department, Peru. It is triclinic and the third naturally occurring modification of AgSbS 2 besides monoclinic miargyrite and cubic cuboargyrite. The composition is usually close to the ideal formula. However, some grains of baumstarkite show zoned lamellae with As contents up to 11.5 wt% and accords to Ag 3 (Sb,As) 2 SbS 6 . Baumstarkite is isotypic with aramayoite [end-member composition Ag 3 Sb 2 BiS 6 ; solid solutions require the extended formula Ag 3 Sb 2 (Bi,Sb)S 6 ]. Single-crystal X-ray structure investigations were performed for baumstarkite [type locality, a = 7.766(2), b = 8.322(2), c = 8.814(2) Å, α = 100.62(2), β = 104.03(2), γ = 90.22(2)°, Z = 2{Ag 3 Sb 3 S 6 }, space group P1̅, R1(F) = 0.057, wR2(F 2 ) = 0.128], aramayoite [Armonia mine, El Quevar, Argentinia: a = 7.813(2), b = 8.268(2), c = 8.880(2) Å, α = 100.32(2), β = 104.07(2), γ = 90.18(2)°, Z = 2{Ag 3 Sb 2 S 6 }, space group P1̅, R1(F) = 0.034, wR2(F 2 ) = 0.084], and miargyrite associated with baumstarkite type material [a = 12.862(3), b = 4.409(1), c = 13.218(3) Å, β = 98.48(2)°, Z = 8{AgSbS 2 }, space group C2/c, R1(F) = 0.031, wR2(F 2 ) = 0.082]. The space-group symmetries of aramayoite and miargyrite were revised, and the refinements unambiguously showed that the three investigated minerals are centrosymmetric. In baumstarkite and aramayoite each three atomic sites are occupied by Ag and M = As, Sb, Bi, respectively. The Ag atoms have two short bonded ligands (Ag-S is 2.51 to 2.58 Å). The M1 and M2 sites are [3 + 3] coordinated and are predominantly occupied by (Sb, As) atoms (M-S = 2.44 to 2.54 Å and > 3.09 Å). The [2 + 2 + 2] coordination of the M3 atom differs in the two mineral species: the two shortest bond lengths in baumstarkite are smaller (2.51 Å) than in aramayoite (2.64 Å) to allow for the different sizes of the Sb and Bi atoms, respectively; the medium bond lengths are similar (2.75 to 2.82 Å) and the longest bond lengths are > 3.02 Å. Considering only the nearest-neighbor environments, baumstarkite and aramayoite feature zigzag chains parallel to [010], which are linked together to form layers parallel to (001). In miargyrite [2 + 2] and [2] coordinated Ag atoms are linked by SbS 3 pyramids to form a three-dimensional network.

The fractionation of B and its isotopes between aqueous fluid and silicate melt has been studied from 550 to 1100 °C and 100-500 MPa. Fluid-melt partition coefficients are <1 for basaltic melt and >1 for rhyolite melt. This shows that B is not always strongly extracted from melts into hydrous fluids. Boron isotopic fractionation is large compared with the carbon and oxygen stable isotopic systems (especially at high T) and is most simply explained by differences in coordination (trigonal vs. tetrahedral) among coexisting phases. Combined with earlier measurements on illite-water (300- 350 °C), B isotopic fractionation defines a temperature-dependent trend from 300 to 1100 °C. Because of the large magnitude and apparent low sensitivity to bulk composition, B isotopic fractionation can be readily applied to studies of diagenesis, hydrothermal alteration of planetary bodies, subduction- zone processing and arc magma generation, and magma chamber evolution.

We present an experimental study of synthetic peraluminous granite doped with H 2 O, B, P, and F, which confirms that aluminosilicate melt, hydrous fluid, and hydrosaline melt (high-temperature brine) can stably coexist at 450-900 ∞C and 0.1-0.2 GPa in bulk compositions similar to those of natural granitic pegmatites. Hydrosaline melt is not quenchable, unstable at room conditions, and requires special techniques for synthesis and preservation. Raman spectroscopy and electron microprobe analyses of hydrosaline melt synthesized in our experiments show that it is composed of H 3 BO 3 , Na 3 AlF 6 , AlPO 4 , H 2 O, and aluminosilicate components. Aluminosilicate melt saturated in both hydrosaline liquid and hydrous fluid at 850 °C and 0.2 GPa contains 3.6 wt% F, 4.2 wt% P 2 O 5 , and 4 wt% B 2 O 3 . Natural hydrosaline melts have previously been found as inclusions trapped in rock-forming minerals. They are not restricted to granites and can be effective agents for enhanced crystal growth, metasomatism, and ore formation. In addition, hydrosaline melts may account for many characteristic features of rare-element and miarolitic pegmatites, such as giant size and perfect shapes of crystals in pegmatite cores, diverse mineralogy, and strong enrichment in rare elements.

Rapid and accurate determination of the surface area of three kaolinite clay standards, taking into account the complex microtopography of the particles, was achieved using atomic force microscopy images and computerized image analysis. All surface areas were determined to within 3%. Edge surface area is 18.2-30.0% of the total surface area depending on the particular kaolinite standard. Specific surface areas agree to within 4% of published values determined by the BET method. The approach can be applied to clay and nanoparticle samples too small in quantity for BET analysis, since it requires ~11 orders of magnitude less material.

A new high-pressure phase of FeSi with the CsCl structure has been synthesized by high-temperature reaction of Fe and Si mixtures at 24 GPa. Powder X-ray diffraction measurements yield a cubic unit cell of a = 2.7917(1) Å for the CsCl-FeSi phase which has composition Fe 0.52 Si 0.48 by electron microprobe. The transition from e-FeSi to the high-pressure phase occurs at 1950 ± 50 K at 24 GPa and has a negative Clapeyron slope.