Volume 83, Issue 1

Spontaneous strains for the α ↔ β transition in quartz were determined from lattice parameter data collected by X-ray powder diffraction and neutron powder diffraction over the temperature range ∼5-1340 K. These appear to be compatible with previous determinations of the order parameter variation in α quartz only if there is a non-linear relationship between the individual strains and the square of the order parameter. An expanded form of the 2-4-6 Landau potential usually used to describe the phase transition was developed to account for these strains and to permit calculation of the elastic constant variations. Calibration of the renormalized coefficients of the basic 2-4-6 potential, using published heat capacity data, provides a quantitative description of the excess free energy, enthalpy, entropy, and heat capacity. Values of the unrenormalized coefficients in the Landau expansion that include all the strain-order parameter coupling coefficients were used to calculate variations of the elastic constants. Values of the bare elastic constants were extracted from published elasticity data for β quartz. Calculated variations of C 11 and C 12 match their observed variations closely, implying that the extended Landau expansion provides a good representation of macroscopic changes within the (001) plane of quartz. Agreement was not as close for C 33 , suggesting that other factors may influence the strain parallel to [001]. The geometrical mechanism for the transition involves both rotations and shearing of SiO 4 tetrahedra, with each coupled differently to the driving order parameter. Only the shearing part of the macroscopic distortions appears to show the same temperature dependence as other properties that scale with Q 2 . Coupling between the strain and the order parameter provides the predominant stabilization energy for α quartz and is also responsible for the first-order character of the transition

MgSiO 3 perovskite is shown to be a Debye-like mineral by the determination of specific heat, CV, entropy, S, and thermal pressure, ΔP Th , using the Debye theory up to 1800 K. Sound velocities and bulk moduli at ambient conditions published by Yeganeh-Haeri were used to find the ambient acoustic Debye temperature, Θ ac D . The variation of Θ ac D with T was assumed to be a curve parallel to the Θ ac D vs. T curves previously found for Al 2 O 3 , MgO, and MgSiO 3 , enabling Θ ac D (T) to be given up to 1800 K. To determine C P , the thermal expansivity, α, and the isothermal bulk modulus, K T , are needed. After considering several sets of α(T), the α(T) data of Funamori and his colleagues were chosen. Using the ambient K T and the values of (მK T /]T)P vs. T reported by Jackson and Rigden, K T (T) up to 1800 K was found. Then C P (T) up to 1800 K was found assuming quasiharmonicity in C V . The data behind the C P (T) calculation are also sufficient to find the Grüneisen parameter, γ(T), and the Anderson-Grüneisen parameters, δ T and δ S , up to 1800 K. The value of q = (მ ln γ/] ln V)T was found, and with γ and ρ, ΔP Th vs. V and T was determined. The three sound velocities, v s , v ρ , and v b = √K S /ρ, were then determined to 1800 K. From vs and v ρ , Poisson’s ratio and the isotropic shear modulus, G, were found to 1800 K. MgSiO 3 perovskite is one of a small, select group of Debye-like minerals for which thermoelastic properties and the equation of state (EOS) are calculable from acoustic data.

We present 16 new manometric determinations of H 2 O solubility for a range of natural silicate liquid compositions equilibrated up to 3 kbar of H 2 O pressure. As the threshold temperature of dehydration of the quenched glasses during measurements of the H 2 O content becomes lower as a function both of bulk silicate composition and the dissolved H 2 O content, we measured the H 2 O released on heating over a range of temperature intervals. For example, alkali-rich samples having a dissolved H 2 O content greater than ∼6 wt% start to evolve H 2 O at temperatures less than 150 °C, whereas more mafic samples and silicic samples with less than 6 wt% H 2 O begin to dehydrate at temperatures greater than 200 8C. This behavior is consistent with the concept that alkali-rich liquids can have their glass transition temperatures lowered substantially by dissolved H 2 O and that H 2 O is released only significantly on heating in the supercooled liquid region, rather than in the glass region. Using these new data, in conjunction with previous data from the literature, we refined and extended the empirical H 2 O solubility model of Moore et al. (1995b). The new model works well (2σ = ± 0.5 wt%) between 700-1200 °C and 1-3000 bar and can be applied to any natural silicate liquid in that range. The model may also be used for systems where X H₂O < 1 in the vapor phase.

The effect of water (= H 2 O + OH) on the environment of Co 2⁺ and Ni 2⁺ in albite glass was investigated by electronic absorption spectroscopy. The visible spectra of Co 2⁺ -doped glasses change only slightly over the range of water concentrations studied. Co 2⁺ is in a distorted tetrahedral environment, producing a dark blue color in the glasses. Up to about 5 wt% water, Ni 2⁺ -doped glasses are brown and only minor variations in the spectra are seen. At 5.6 wt% water, however, the color of these glasses changes abruptly from brown to light green and a new type of absorption spectrum is observed. Three bands are observed in the visible spectra of the brown glasses. Two bands near 20 500 cm ⁻1 can be assigned to Ni 2⁺ in a distorted octahedral environment. A third band at 15 500 cm ⁻1 could either be due to the distorted octahedral site or to a small amount of tetrahedrally coordinated Ni 2⁺ . The spectra of the green glasses with 5.6 and 5.7 wt% H 2 O resemble closely spectra of aqueous NiCl 2 -solution containing the [Ni(H 2 O) 6 ] 2+ complex. The formation of such a hydration shell around transition metals in hydrous silicate melts should strongly effect partitioning of these elements between silicate melts, minerals, and a metal phase. Consideration of ligand field stabilization energies, suggests that hydrated Ni 2⁺ is stabilized in the melt such that mineral-melt and metal-silicate melt partition coefficients decrease by one to two orders of magnitude relative to a dry melt at 1100 °C.

The reaction talc + forsterite = enstatite + H 2 O has been investigated between 6 and 20 kbar. Previous high-pressure experimental studies suggested various reaction positions, mostly with positive P-T slopes. The new results show that the reaction has a negative slope in the pressure range studied. It was bracketed between 640 and 680 °C at 10 kbar, between 620 and 640 °C at 15 kbar, and between 580 and 600 °C at 20 kbar. The growth of antigorite at high pressures prevented the reaction from being determined above 20 kbar. The reaction position is consistent with the previous low-pressure data of Chernosky et al. (1985), is in very good agreement with the position calculated using the thermodynamic database of Berman (1988), and is in reasonable agreement with the calculation using the thermodynamic database of Holland and Powell (1990). The high reaction temperatures suggested by previous high-pressure experiments may reflect metastable or quench growth of talc in those experiments. Talc in equilibrium with forsterite is stable over a wide P-T range in hydrothermally altered and metamorphosed ultramafic rocks and may be important for carrying H 2 O into subduction zones. However, talc dehydration occurs at too shallow a depth for the H 2 O to contribute directly to subduction zone volcanism.

Synthetic gel and glass of illitic composition, natural kaolinite, and mixed-layer illitesmectite were used as starting materials for hydrothermal synthesis of ammonium illite. Ammonium illite was prepared from synthetic gel by hydrothermal treatment at 300 °C. The onset of crystallization began within 3 h, and well-crystallized ammonium illite appeared at 24 h. Increasing reaction time (up to four weeks) led to many illite layers per crystal. In the presence of equivalent proportions of potassium and ammonium, the gel was transformed to illite with equimolar contents of K and NH 4 . In contrast, synthesis using glass under the same conditions resulted in a mixture of mixed-layer ammonium illite-smectite with large expandability and discrete illite. Hydrothermal treatments of the fine fractions of natural kaolinite and illite-smectite produced ammonium illite from kaolinite but the illite-smectite remained unchanged.

Acoustic velocities of brucite were measured at room pressure in over 48 directions from Brillouin spectroscopy using a natural sample. These data are supplemented with volume measurements as a function of pressure and temperature that range from ambient conditions to 11 GPa and 873 K using synchrotron X-ray radiation at the National Synchrotron Light Source (NSLS) in a cubic-anvil apparatus (SAM-85) with a synthetic polycrystalline sample. The diffraction patterns are collected during cooling cycles to minimize the effect of deviatoric stress on the measurements. These data yield internally consistent thermoelastic parameters defining the equation of state of brucite along with the single-crystal elastic moduli. The Brillouin spectroscopy measurements are best fit with the following elastic model: C 11 = 156.7(8), C 33 = 46.3(8), C 44 = 21.7(5), C 12 = 44.4(10), C 13 = 12.0(15), and C 14 = 0.2(8) GPa. The resultant linear compressibilities of the a and c axes are 3.8(1) × 10 -3 and 19.6(6) × 10 -3 (GPa -1 ), respectively, with the Reuss bound for the bulk modulus, K R = 36.7(10) GPa and the Hill average, K H = 46(1) GPa. The unitcell parameters (a, c, and volume) determined from the diffraction measurements were fit with a Birch-Murnaghan equation of state, yielding K 0 = 39.6(14) Gpa, Kʹ = 6.7(7), (მK T / ]T) P = 20.0114(16) GPa/K, and α = 5.0(7) × 10 -5 /K. The bulk modulus and linear compressibilities from X-ray diffraction are in agreement with those from Brillouin spectroscopy. The ratio of linear compressibility of the a to c axes is about five times at ambient conditions and reduces to almost unity by 10 GPa. The axial thermal expansions reflect a similar pressure dependence. The ambient shear anisotropy (C 44 /C 66 ) is about 2.5.

Raman spectra were obtained from the (001), (010), and (100) faces of a St. Claire dickite specimen of known orientation. Raman spectra collected from the (010) and (100) faces of dickite are reported for the first time and reveal vibrational features significantly different from the (001) spectra. Variations in intensities of the n(OH) bands in polarized spectra were used to confirm previous band assignments, to determine the shape and orientation of the local Raman tensors for the OH1 and OH3 groups. The most striking polarization effect observed in the n(OH) region of dickite was the behavior of the a(cʹcʹ)a̅ spectrum relative to Raman and IR spectra of other orientations. Unlike previously reported spectra, the dominant feature in this spectrum was the 3643 cm -1 band. This large increase in intensity of the 3643 cm -1 band in comparison with the other n(OH) bands was related to the fact that Raman spectra were recorded from the edge faces of dickite with the electric vector of the incident laser polarized along the c axis. This permitted observation of vibrational modes polarized along the c axis. Raman frequencies of the n(OH) bands assigned to the OH2 and OH4 groups differ from their IR counterparts by 12 cm -1 , suggesting that these groups may be related by a center of symmetry. For comparison, Raman spectra in the n(OH) region were also obtained from individual micro-crystals of kaolinite that were approximately 5 mm across the (001) face.

We report the compositional systematics of the 29 Si MAS NMR spectra of richterite, [A] (Na,K,Rb) 1 (Na,Ca,Sr) 2 Mg 5 Si 8 O 22 (OH,F) 2 ; pargasite, NaCa 2 Mg 4 AlSi 6 Al 2 O 22 (OH) 2 ; and fluor-edenite, NaCa 2 Mg 5 Si 7 AlO 22 F 2 . [A]Na causes (1) splitting of T1 into two sites, while leaving T2 unsplit, and (2) the Q 3 chemical shift to be 2.5 to 3 ppm less negative than Q 3 when the A site is empty as in tremolite and magnesiohornblende. The preferential splitting of T1 is explained in terms of ordering of the A cation at Am. The effects of M4 and Asite chemistry upon the richterite spectra aid the assignment of peaks for pargasite and fluor-edenite. The long-range ordering of Al and Si over T1 and T2 sites in fluor-edenite synthesized at 2 kbar, and 1000 °C and pargasite synthesized at 1 kbar, and 930 °C has been calculated from their 29 Si MAS NMR spectra assuming that Al avoidance operates. The extent of long-range order is calculated from the intensities of the Q 2 (2Al), Q 2 (1Al), and Q 2 (0Al) peaks. An equation is derived that allows the extent of long-range Al-Si order to be calculated from 29 Si MAS NMR spectra of amphiboles. The spectrum of fluor-edenite is consistent with all [4] Al being at T1 with maximal short-range disorder within the constraints of Al avoidance. The pargasite spectrum is more complex, because there is a probable peak coincidence of Q 2 (1Al) and Q 3 (2Al) at 282 ppm that must be considered. The presence of a Q 3 (0Al) peak in the pargasite spectrum indicates unambiguously that some long-range disorder exists, and this implies that Q 3 (2Al) groupings also occur. The calculated extent of long-range disorder in pargasite is 55 ± 10%. This value is consistent with the single-crystal X-ray data of Oberti et al. (1995a) for natural amphiboles extrapolated to [4] Al = 2 apfu at 900 °C. The different long-range Al-Si ordering behavior of flour-edenite and pargasite is explained in terms of the bond-valence requirements of O4. At high temperatures, configurational entropy becomes an important stabilizing factor, and structural distortion around O4 in pargasite and hornblende allows Al into T2, provided that O4 is coupled to Al at an adjacent M2 site, as in pargasite. The results for fluoredenite, which has no [6] Al, show that the low O4 bond-strength sum of the T2 Al M2 Mg M4 Ca configuration cannot be accommodated by sufficient structural relaxation, even at 1000 8C. Coupling between Al at M2 and T2 is an important control on Al-Si long-range orderdisorder.

Synthetic Ca-exchanged birnessite (CaBi) was studied by X-ray and selected-area electron diffraction (XRD, SAED). The substructure of CaBi may be described with a fourlayer monoclinic subcell with a = 5.150, b = 2.844, c = 4cʹ = 28.16 Å , and b = 90.38. Two different superstructures of CaBi were distinguished. CaBi type I has cell parameters A = 3a = 15.45, B = 3b = 8.472 Å . The stacking sequence in this unit cell may be described as defect-free OSOS, where successive layers are shifted relative to their predecessors by 0 (O) or b/2 (S) along the b axis. The complete description of stacking involves the structure of the layer itself, the structure of the interlayer, and the shift from this layer to the next one. CaBi type II can be described as a regular interstratification of A P = 3⁄2(a - b), B P = 4b, γ =5 118.98 and A P = 3⁄2(a + b), B P = 24b, γ = 118.98 supercells that are connected by a mirror plane in projection on the a-b plane. Its stacking sequence is a random interstratification of OSOS (90%) and OOOS (10%) structural fragments. Most CaBi crystals appeared to consist of intergrown type I and type II sub-crystals. As in Na-rich birnessite, the A = 3a superstructure arises from the ordered distribution of Mn 3+ -rich rows parallel to [010] and separated from each other along [100] by two Mn 4+ rows. In Mn 3+ -rich rows heterovalent Mn cations are regularly distributed according to Mn 3+ Mn 3+ Mn 4+ (CaBi type I, B = 3b) and Mn 3+ Mn 3+ Mn 4+ Mn 4+ (CaBi type II, B = 4b) sequences. Super-periodicities along the b axis are induced by these regular distributions of heterovalent Mn atoms in Mn 3+ -rich rows and of associated interlayer Ca. No significant amount of layer vacancies was detected. Idealized structural formulae for CaBi type I and II are Ca(Mn 3+ 2 Mn 4+ 7 )O 18 and Ca(Mn 3+ 2 Mn 4+ 10 )O 24 , respectively.

The Inorganic Crystal Structure Database (ICSD) was searched for Si-OH groups with ordered H positions leading to 31 structures with 46 Si-OH groups. The geometrical characteristics of these partly hydroxylated SiO 4 tetrahedra were analyzed. Depending on the condensation of the tetrahedra, the protonization of one tetrahedral apex allows variations in Si-O bond lengths and distorts the tetrahedron. The Si-OH distance decreases with the number of bridging O atoms (Si-O-Si) from average values of 1.668 Å for orthosilicates to 1.604 Å for tetrahedra with three bridging O atoms, whereas the 〈Si-O〉 distances of the non-hydroxylated Si-O bonds remain constant at 1.62 Å. This behavior was modeled by differences in bond strength sums of the tetrahedral O atoms. In the orthosilicates, the non-hydroxylated tetrahedral apices tend to be underbonded and the O of the silanol group is overbonded to satisfy the charge requirements of Si 4+ . In contrast, an Si-OH bearing tetrahedron with three bridging O atoms is characterized by a more regular bond strength distribution consistent with minor bond length distortion.

During the preparation of the Mineral Group List Index of the Mineral Powder Diffraction File, it was necessary to develop a consistent scheme for the representation of the structural formulas. Because the fourteen letters B, C, F, H, I, K, N, O, P, S, U, V, W, and Y of the Latin (Roman) alphabet represent chemical elements (element symbols), this system uses twelve letters to represent cations, anions, and molecules occupying various sites of the structure (site symbols). The letters available as structure-site symbols represent cations with decreasing coordination numbers (CN) as follows: D, CN = 9; E, CN = 8 or 7; G, CN = 6; J, CN = 5; Q, CN = 4 planar or 2 linear; T, CN = 4 tetrahedral; and R, CN = 3 planar. These definitions leave sufficient remaining letters for other structure-site symbols such as: A for all cations without regard to the coordination number; L for lone-electron-pair cations; M for neutral molecular units; X for monatomic anions; and Z for polyatomic anions. Structure sites of the same coordination type, yet distinct enough for ordered occupancy of resident cations, may be differentiated by primes (ʹ,ʺ,ʹʺ, etc.). Variable coordination numbers on a structure site and variable site occupancies are indicated by two symbols or subscripts with the intervening symbol ↔. For example, the general structure-type formula of amphibole may be written as A 7↔8 [T 4 X 11 ] 2 X 2 . A specific structure-type formula also allows for a mixture of symbols to represent chemical elements and structure sites such as A 0↔1 [Eʹ↔G] 2 Gʹ 3 Gʺ 2 [Si 4 O 11 ] 2 (OH) 2 whereas an example of a chemical formula is Ca 2 Mg 5 [Si 4 O 11 ] 2 (OH) 2 for the amphibole tremolite.

The high-temperature form of pentlandite (Fe 4.5 Ni 4.5 S 8 ) was found to be stable between 584 ± 3 and 865 ± 3 °C, breaking down into monosulfide solid solution and liquid at the later temperature. The phase is unquenchable and always displays the X-ray pattern of pentlandite (low form) at room temperature. High-temperature X-ray diffraction demonstrated that the high form has a primitive cubic cell with a = 5.189 Å (620 °C) corresponding to a/2 of pentlandite. The high-low inversion is reversible, accompanied by a large latent heat. It is thought to be order-disorder in character. The transition temperature falls with decreasing S content. The high form of pentlandite has a limited solid solution from Fe 5.07 Ni 3.93 S 7.85 to Fe 3.61 Ni 5.39 S 7.85 at 850 °C. However its solid solution extends rapidly toward Ni 3±X S 2 in the Ni-S join with decreasing temperature. High-form pentlandite with Fe = Ni in atomic percent crystallizes first by a pseudoperitectic reaction between monosulfide solid solution and liquid. The high form (Fe = Ni) crystallized from the liquid always has the metal-rich (S-poor) composition in the solid solution at each temperature and coexists with taenite γ (Fe,Ni) below 746 ± 3 °C. This metal-rich high-form Fe 4.5 Ni 4.5 S 7.4 breaks down into pentlandite and g (Fe,Ni) at 584 ± 3 °C (pseudoeutectoid). These results suggest that in geological processes, such as the formation of Ni-Cu ore deposits, pentlandite can crystallize as the high form from liquid (sulfide magma) at the comparatively high temperatures around 800 °C.

Atomistic simulation techniques were used to investigate the surface energies and stabilities of the sphalerite form of ZnS. The results show that for pure ZnS the lowest energy surfaces are type I and all of the form {110} with a calculated surface energy of 0.65 J/m 2 . In addition, we illustrate how type III surfaces, such as {1̅1̅1̅}, can be stabilized with respect to {110} by the introduction of point defects to the surface layer. Such defects lead to changes in stoichiometry and to the valence state of surface species. In general, the results suggest that for Zn-poor surface stoichiometries, the (1̅1̅1̅) surface becomes the most stable, whereas for Zn-rich compositions the (1̅1̅1̅) is stabilized to the greatest extent.

A combination of atomic force microscopy, scanning election microscopy, transmission electron microscopy, energy dispersive spectroscopy, electron diffraction, and X-ray diffraction were used to study reactions of 0.5-500 mg/L aqueous Pb with Ca 5 (PO 4 ) 3 OH, hydroxylapatite (HAP), at pH 6 and 22 °C. Following 2 h reaction time, concentrations of Pb aq ([Pb aq ]) decreased from 500 mg/L to ,100 mg/L, and from 0.5-100 mg/L to ,15 mg/L. This loss of Pb aq from solution (i.e., sorption) resulted partially from simultaneous dissolution of HAP and precipitation of Pb 5 (PO 4 ) 3 OH, hydroxypyromorphite (HPY), or another solid Pb phase. The initial saturation state with respect to HPY (defined as the ratio of the ion activity product to equilibrium solubility product) influenced strongly precipitation processes. At a high degree of saturation (initial [Pb aq ] > 100 mg/L), small nuclei or aggregates of poorly crystalline HPY precipitated homogeneously in solution. At intermediate saturation (initial [Pb aq ] ∼ 10-100 mg/L), large, euhedral needles of HPY precipitated homogeneously in solution. At a low degree of saturation (initial [Pb aq ] , 10 mg/L), a needle-like Pb-containing phase grew heterogeneously on HAP. These results agree well with concepts derived from nucleation and growth theories and demonstrate that initial saturation state influences strongly the sorption process.

Lead feldspar single crystals were annealed at T = 1050 and 1000 °C, starting from a disordered metastable configuration (PbF H , Q od = 0) and from an ordered configuration (PbF L , Q od = 0.89). Single-crystal data collection and refinement in space group I2/c show that the degree of Al-Si order increases to Q od = 0.42 after annealing the disordered PbF H at 1050 8C and decreases to Q od = 0.70 after annealing the ordered PbF L sample at 1000 °C. This suggests that the equilibrium Q od is between 0.70 and 0.42 for temperatures between 1000 and 1050 °C, where anorthite or strontium feldspar are almost completely ordered. A residual in the difference-Fourier map because of positional disorder was observed near the Pb site in all the refined crystals. The average y/b Pb coordinate changes with increasing Al-Si disorder, as Pb approaches the glide plane. A significant decrease in the intensity of b-type reflections was consequently observed. A spontaneous strain, with the main axis almost parallel to the a axis, is associated with Al-Si ordering. Pb polyhedral deformation related with Q od accounts for the observed strain. A calibrating equation, Q od = [(8.427(2) - a) / 0.048(3)] 1/2 , has been calculated and applied to the unit-cell parameters obtained from subsequent thermal treatments and from Bruno and Facchinelli (1972) to define the evolution of the Q od vs. the treatment temperature. The thermal behavior of the Q od could then be bracketed, suggesting T c between 1150 and 1200 °C for the I2/c-C2/m phase transition induced by the Al-Si order-disorder process.

Ferri-clinoholmquistite is a new amphibole species from episyenites in the East Pedriza Massif (Central System, Spain), where it is associated to albite, augite-aegirine, titanite, andradite, magnetite, and apatite. It is black, vitreous, translucent, non-fluorescent, and brittle. It shows gray streak, H(Mohs) = 6, splintery fracture, perfect {110} cleavage, (001) parting, D means = 3.19, and D cal = 3.25. Crystals are prismatic, elongated on [001]. In plane-polarized light, it is strongly pleochroic: X = yellow green, Y = indigo blue, Z = green blue, with absorption X < Y ≤ Z; Z = b, Y - c = 10(2)°, X - a ~ 228 (in β obtuse). Ferri-clinoholmquistite is biaxial positive, α = 1.699(2), β = 1.703(2), γ = 1.708(2), 2V Z (meas) = 72(7), 2V Z (calc) = 84(6), r < v. It is monoclinic, space group C2/ m, a = 9.472(4), b = 17.844(6), c = 5.276(6) Å , β = 101.97(9)8, V = 872(1) Å 3 , Z = 2. X-ray powder-diffraction pattern data were determined. Analysis by a combination of electron microprobe and flame photometry gives the following formula, calculated assuming OH 1 F = 2 and T sites fully occupied by Si: A (Na 0.43 K 0.03 ) B (Li 1.66 Na 0.30 Ca 0.04 ) C (Fe 3+ 1.54 Fe 2+ 1.35 Mg 1.21 Li 0.49 Al 0.20 Ti 4+ 0.12 Mn 2+ 0.07 Zn 0.02 ) T (Si 8 )O 22 (OH 1.58 F 0.42 ). From crystallographic refinements the M4 site is split, implying ordering of Li and Na, and within the A cavity, Na occupies the Am position.

Bamfordite from the abandoned W-Mo-Bi mines at Bamford, Queensland, Australia, is a new hydrated iron molybdate with a unique structure. This mineral formed by oxidation of molybdenite, MoS 2 , in the presence of strongly acidic solutions. It occurs as microcrystalline aggregates of tabular triclinic crystals between 0.005 and 0.05 mm long. The aggregates are apple-green with an earthy luster and greenish yellow streak. Crystals are transparent, with pale to moderate yellow-green pleochroism. They show principal forms {001}, {100}, {010}, {110}, {11̄0}, and prominent (100) cleavage traces. The Mohs hardness is 2-3 and the measured density is 3.620 g/cm 3 (calculated density is 3.616 g/cm 3 ). Crystals are biaxial negative and length slow, with RIs of α = 1.91, β = 2.03, and γ = 2.11, and 2V≈ 908. Chemical analysis yielded an empirical formula of Fe 3+ 1.00 Mo 2.01 W 0.03 P 0.02 O 10 H 4.62 , calculated on the basis of ten O atoms. The simplified formula is Fe 3+ Mo 2 O 6 (OH) 3 ·H 2 O, chosen on the basis of crystal-structure determination and Mössbauer spectroscopy results. Unit-cell parameters calculated both from the X-ray powder and singlecrystal diffraction data are a = 5.889(5), b = 7.545(5), c = 9.419(5) Å ; α = 71.46(4)°, β = 83.42(4)°, γ = 72.78(4)°; V 5 378.9(4) Å 3 ; Z = 2; P1 or P1̄. The crystal structure was solved in P1 using direct methods and Fourier techniques. The final refinement based on 1486 observed reflections [I > 2.00 σI] converged to R 5 0.05 and R w 5 0.038. The bamfordite crystal structure contains groups of four MoO 6 octahedra, linked by edgesharing, which in turn are linked through corner-sharing to pairs of FeO 6 octahedra thereby forming infinite sheets parallel to (100). These sheets are stepped and linked by hydrogen bonding. No other molybdenum oxides have this or a similar structure, instead molybdates such as wulfenite, PbMoO 4 , are based on tetrahedrally coordinated molybdenum.

Six cordylites from the four known localities [Narssarssuk (Greenland), Mont St. Hilaire (Canada), Bayan Obo (China), and Kola Peninsula (Russia)] were investigated by electron microprobe and X-ray single-crystal structure determination, and a seventh sample was investegated by X-ray methods only. The material studied included type cordylite. The idealized formula for cordylite is redefined as (Na 1-x Ca x/2 )BaCe 2 (CO 3 ) 4 F, 0 < x < 1, with Ce for the sum of REE; the SrO content may reach about 5.7%. All of our structure refinements (P6 3 /mmc; a ≅ 5.10, c ≅ 23.10 Å ) agree very well among themselves and with the published structures of ‘‘baiyuneboite-(Ce)’’ and unnamed (Ca 0.5 ⃞ 0.5 )BaCe 2 (CO 3 ) 4 F. Cordylite has a sheet structure formed of (001) layers of [Ba] [CO 3 ] [Ce,CO 3 ] [Na,F] [Ce,CO 3 ] [CO 3 ] [Ba] stacked along [001]. The interatomic distances are as expected, with the exception of unshielded Na-F distances of 2.94Ä; unlike the carbonate groups in basnäsite and synchysite, the carbonate groups in cordylite-(Ce) are parallel (001). Investigation of type cordylite showed that the formula proposed by Flink, i.e., BaCe 2 (CO 3 ) 3 F 2 , is to be modified to that given above, with x ≅ 0 for Flink’s material. Baiyuneboite-(Ce), a mineral previously approved by the IMA CNMMN but later withdrawn because of potential similarities with cordylite, is confirmed here as being essentially identical to type cordylite.