Volume 93, Issue 4

Guanacoite with high Mg content from the type locality (El Guanaco mine, Chile) was investigated to elucidate the crystal-chemical characteristics of this arsenate mineral. In this study, WDS X-ray intensity maps revealed that guanacoite single crystals from the investigated specimen show distinct and characteristic compositional zoning, which varies cyclically along [100], which is in the direction of the elongation of the prismatic crystals (Witzke et al. 2006); the boundaries separating different compositional regions lie perpendicular to the prism faces. The crystal structure of a selected fragment [a = 5.459(2), b = 16.808(9), c = 6.917(3) Å, β = 100.44(1)°, V = 624.1(5) Å 3 , P2 1 /c, and Z = 2], was solved using direct methods and refined to an R index of 3.09% for 1385 observed [I o > 4σ(I o )] reflections measured at 103 K using MoKα X-radiation. The chemical formula based on the refinement is (Cu 0.85 Mg 0.15 ) 2 Mg 2 (Mg 0.83 Cu 0.17 )(OH) 4 (H 2 O) 4 (AsO 4 ) 2 . The structure model obtained is in good agreement with the previous structure determination, although the occupancies of the three M sites (M = Mg or Cu) are distinctly different, i.e., the sites are considerably richer in Mg, especially M1 and M3. These structural results suggest that guanacoite’s chemical formula should be enlarged from the Curich composition Cu 2 Mg 2 (Mg,Cu)(OH) 4 (H 2 O) 4 (AsO 4 ) 2 to (Cu,Mg) 2 Mg 2 (Mg,Cu)(OH) 4 (H 2 O) 4 (AsO 4 ) 2 to include the Mg-rich members (Mg up to 3.395 apfu). It appears that the chemical variability of guanacoite is mainly due to Mg-Cu substitution on the M1 and M2 sites.

Production of acid mine drainage (AMD) is in large part due to pyrite oxidation. The tie-in between environmental remediation and pyrite oxidation requires understanding pyrite oxidation in aqueous systems. In this study, ex-situ measurements using vertical scanning interferometry (VSI) were utilized to investigate pyrite surface reactivity under AMD conditions, such as pH 1 (HCl), an O2-saturated atmosphere, and room temperature, including (1) ex-situ measurements using vertical scanning interferometry (VSI) and (2) solution chemistry measurements using a flow reactor. In the former, two fragments were immersed in the acidic solution for 27 days at undersaturation with respect to pyrite. Weathered surfaces of pyrite that showed surface history (e.g., existence of terraces, steps, etch pits, and non-uniform surface roughness) were selected to examine surface topography changes with time. Based on the VSI measurements, the reaction mechanism includes the formation and coalescence of etch pits leading to overall surface retreat. This result is consistent with the stepwave model that predicts that, under sufficiently high undersaturation (i.e., ΔG < ΔG crit ), pit opening and generation of stepwaves are mechanisms that control mineral dissolution. Surface reactivity was not uniform over the entire surface, yielding surface regions with a considerable variety of dissolution rates that ranged from 1.7 × 10 −7 mol/(m 2 ·s) to lower than 2 × 10 −11 mol/(m 2 ·s). The overall pyrite dissolution rate calculated over the explored surface was 2.8 × 10 −9 mol/(m 2 ·s), which agrees very well with the absolute rate measured [3.1 × 10 −9 mol/(m 2 ·s)]. Based on the release of Fe in the flow experiment and normalizing with the geometric surface area, the steady-state pyrite dissolution rate obtained was 7.2 ± 1.5 × 10 −9 mol/(m 2 ·s), i.e., a factor of about 2 higher than that determined by VSI. Collectively, these rates also agree with pyrite dissolution rates obtained by bulk dissolution measurements. These results highlight (1) “local” dissolution rates vary widely over mineral surfaces at the short range scale from nano- to millimeter lengths; (2) inferred dissolution rates that are usually based on topographic changes on rather small (micrometer-scale size) surface regions are not necessarily representative of rates under field conditions; and (3) BET surface area and geometric surface area associated with reactive surface area and used to normalize the pyrite dissolution rates yield a variation in rate that ranges from 6.3 × 10 −11 to 7.2 ± 1.5 × 10 −9 mol/(m 2 ·s).

Scanning transmission electron microscopy images of the 00l white mica planes in crystals from central Switzerland and Crete, Greece, reveal that domains of paragonite, margarite, and muscovite are ordered within the basal plane. Energy dispersive X-ray analyses show that both cations in the interlayer and in the 2:1 layer have ordered on the scale of tens to hundreds of nanometers. Domain boundaries can be both sharp and crystallographically controlled or diffuse and irregular. A model outlining the domain formation process is presented that is consistent with X-ray powder diffraction and transmission electron microscopy data. The domain model incorporates aspects of a mixedlayered and a disordered compositionally intermediate phase models. The main feature of the model is the formation of mica species that segregate within the basal plane and contradict the notion of homogeneous layers within mixed-layer phases. Implications for the formation of all diagenetic and very low-grade metamorphic 2:1 sheet silicates are discussed

Molecular dynamics simulations and first-principles electronic structure calculations are used to study the structural behavior of orthoenstatite, MgSiO 3 , at high pressures. The calculations suggest two possible high-pressure polymorphs of orthoenstatite, one with P2 1 ca and the other with Pbca symmetry. Both polymorphs are structurally related to orthoenstatite. Molecular dynamics simulations reveal the displacive nature of the phase transitions between the three phases. Electronic structure calculations predict a phase transition from orthoenstatite to the metastable P2 1 ca structure at 9 GPa, which may explain the anomalies in elastic and vibrational properties observed experimentally. A second metastable transition from the P2 1 ca to the high-pressure Pbca structure may be observable above 20 GPa.

The Fe(III)-reducing bacterium Geobacter sulfurreducens produces nanometer-sized magnetite by transferring electrons from organic matter or hydrogen to ferrihydrite, resulting in the reductive transformation of the ferrihydrite to magnetite, and the generation of ATP for growth. Electron transfer can occur by direct contact between the cell surface and the mineral or via a soluble “electron shuttle,” for example a quinone-containing humic species. The minerals produced at different stages of ferrihydrite reduction during two experiments, one with and one without the humic analog anthraquinone-2, 6-disulphonate (AQDS), were measured using high-resolution synchrotron powder X-ray diffraction. Amorphous 2-line ferrihydrite converts to goethite, then to a mixture of goethite and magnetite, and finally to magnetite. Samples with and without AQDS showed the same general mineralogical trends, and the rate of reaction was faster in the presence of AQDS. In addition, two transient minerals structurally similar to goethite and magnetite were observed to form as intermediates between ferrihydrite and goethite and goethite and magnetite, but only in samples produced in the absence of the electron shuttle. These transient minerals were named proto-goethite and proto-magnetite. Proto-goethite has a shorter c-axis [4.467(20) Å] than crystalline goethite, a function of size (<2 nm) where quantum properties prevail. Proto-magnetite is identified by long tetrahedral (2.113 Å) and short octahedral (1.943 Å) Fe-O bonds compared to stoichiometric magnetite, possibly indicative of a coordination crossover caused by charge density [Fe(II)] migration to tetrahedral sites. Fe(II) in solution or sorbed to the mineral surface is considered to be the catalyst causing the mineral transformations. The Fe(II) is thought to form predominantly from the reductive dissolution of 2-line ferrihydrite by G. sulfurreducens.

The low-temperature heat capacities of natural near end-member minerals (about 95 mol%, except tinzenite of about 34 mol% on average) of the axinite group, previously characterized in detail by means of powder and single-crystal X-ray diffraction, electron microprobe, and Mössbauer spectroscopy, were measured by heat-pulse calorimetry using the Physical Properties Measurement System (Quantum Design) at temperatures between 5(2) and 300 K. From these data, the following entropy values [in J/(mol·K)] of the natural samples at 298.15 K were derived: S 298,magnesioaxinite = 696.3 ± 1.1, S 298,ferroaxinite = 743.5 ± 3.5, S 298,manganaxinite = 737.5 ± 2.6, and S 298,tinzenite = 758.1 ± 2.8. For the end-member compositions, the corrected heat capacities at 298.15 K and standard third-law entropies of the axinites are [all in J/(mol·K)]: CP o magnesioaxinite = 827.5 ± 1.5 and S o magnesioaxinite = 693.7 ± 1.1, CP o ferroaxinite = 841.8 ± 3.3 and S o ferroaxinite = 749.6 ± 3.5, CP o manganaxinite = 849.1 ± 2.5 and S o manganaxinite = 737.8 ± 2.6, and CP o tinzenite = 841.6 ± 2.6, S o tinzenite = 754.0 ± 2.8. The standard entropies of manganaxinite and tinzenite include contributions of 1.9 and 4.3 J/(mol·K) for the range 0-5 K evaluated based on a Schottky anomaly fitted to the low-T C P values of these axinites. The lowest measured heat capacities of ferroaxinite indicate that a lambda-type C P anomaly should exist between 0 and 2 K. Its likely contribution to the standard entropy was estimated as ~5.2 J/(mol·K). A low-temperature C P anomaly below 15 K for ferroaxinite is well-explained by ferromagnetic ordering, whereas for manganaxinite by uncompensated antiferromagnetic ordering, and for tinzenite by pure antiferromagnetic ordering.

Compressional (v p ) and shear (v s ) wave velocities of plagioclase (An 51±1 ) were measured up to 900 °C at 1 GPa. The temperature derivatives of v p (∂v p /∂T) and v s (∂v p /∂T) show a discontinuous change at ~400 °C. The ∂v p /∂T is -0.9 × 10 -4 km/s/°C below 400 °C and -4.4 × 10 -4 km/s/°C above 400 °C. The ∂v p /∂T also increases from -0.7 × 10 -4 to -4.1 × 10 -4 km/s/°C. These v p and v s show reversible changes between 25 and 700 °C. In contrast, both v p and v s increase (0.08 and 0.08 km/s, respectively) at 700-800 °C, and show irreversible changes after heating to 800 and 900 °C. The X-ray powder diffraction analysis shows that the run product heated to 900 °C shows a higher lattice angle γ than the run products obtained on heating up to 700 °C, which is comparable to the lattice angle γ of high and low plagioclase, respectively. We ascribe the discontinuous change in v p , v s , ∂v p /∂T, and ∂v s /∂T to the order-disorder transition of plagioclase at high temperatures.

The existence of a disordered silica polymorph dominated by tridymite-like stacking is implied by the long-held interpretation of opaline silica as disordered intergrowths of cristobalite-like and tridymite-like domains. The currently accepted model for diagenetic transformation of amorphous silica also predicts that early formed structures will be dominated by tridymite stacking. Although the cristobalitic end-member (opal-c) and intermediate structures (opal-ct) have long been acknowledged, the tridymitic end-member has not yet been recognized. Evidence for such an end-member is found in 54 Twiggs clay samples examined by X-ray diffraction (XRD). The mineralogy of the samples is dominated by smectite and disordered silica. The silica phase is biogenic in origin and presents an XRD signature consistent with that of line-broadened tridymite, with a broad primary reflection near 4.107 Å (21.6 °2θ), additional broad peaks of lower intensity at approximately 4.328 Å (20.5 °2θ) and 2.50 Å (35.9 °2θ), and lacking diagnostic cristobalite peaks at 3.14 Å (28.5 °2θ) and 2.84 Å (31.5 °2θ). At present, the most widely used classification of disordered silica is a threefold system that recognizes an amorphous phase (opal-a), a cristobalite-like end-member (opal-c), and intermediate structures incorporating both cristobalite and tridymite domains (opal-ct) (Jones and Segnit 1971). This classification does not accommodate a dominantly tridymite-like structure as is observed in the Twiggs clay. The silica phase in these samples is more accurately described as “disordered tridymite.” A new term “opal-t” is proposed to describe this end-member phase.

Synthetic crystals of melanotekite and kentrolite were obtained at 850 °C from melt. The aluminum analogue of kentrolite Pb 2 Al 2 Si 2 O 9 was hydrothermally synthesized at 2 GPa, 650 °C together with zoisite-(Pb) and margarite-(Pb). Synthesis products were characterized by single-crystal diffraction studies and microprobe analysis. The aluminum analogue Pb 2 Al 2 Si 2 O 9 was observed in space group Pbcn with lattice parameters a = 6.8981(7) Å, b = 10.6906(15) Å, c = 9.7413(10) Å, and V = 718.37 Å 3 . Fourier mappings show no irregularities of the Pb site. Melanotekite with lattice parameters a = 6.9786(6) Å, b = 11.0170(11) Å, c = 10.0895(9) Å, and V = 775.71(17) Å 3 in space group Pbcn show a slightly deformed Pb-position in Fourier mappings. Kentrolite was observed in space group P2 1 22 1 with pseudo-symmetry to Pbcn with lattice parameters a = 7.0103(5) Å, b = 11.0729 (7) Å, c = 9.9642(7) Å, and V = 773.47(11) Å 3 . Fourier mappings of the kentrolite structure show that two different split Pb sites exist, which causes lower symmetry. The unit-cell volume of different members of the kentrolite group is a linear function of trivalent ionic radii in sixfold coordination for the elements Al, Ga, In, and also for Fe and Mn in high spin mode. The structure of Pb 2 M 2 Si 2 O 9 (M = Al 3+ , Fe 3+ , Mn 3+ ) is built on isolated M-octahedra chains parallel c, M-octahedra sharing alternately trans and skew edges. Each Si 2 O 7 -group is linked with their vertices to three octahedra chains. Their Si-O-Si bond angles depend on the size of M-octahedra and are 129.84° in Pb 2 Al 2 Si 2 O 9 , 131.08° in Pb 2 Fe 2 Si 2 O 9 , 128.34° and 130.33° in Pb 2 Mn 2 Si 2 O 9 .

A set of synthetic goethites were prepared from Fe 3+ - and Co 2+ -nitrate solutions in alkaline media with a Co/(Co + Fe) ratio (x Co ) up to 10 mol%. The structural characterization of the resultant solid phases was carried out by X-ray diffraction (XRD). XRD analyses showed that in preparations with x Co <10 mol%, Co-substituted goethite was the only crystalline phase present. Atomic and cell parameters for the samples synthesized were obtained by the Rietveld refinement of the XRD data, and showed that the unit cell in the goethite-like phase is contracted as a function of x Co . Little deviation from the Vegard rule was observed for all unit-cell parameters. Cobalt substitution produces an increase in the surface area of the goethite, as well as an increment in the dehydroxylation temperature. The acid dissolution of all Co-goethites showed an increase in dissolution rate with the Co content, and a congruent behavior was observed. The activation energy for dissolution was obtained two samples. A modified first-order Kabai equation best describes the dissolution data.

Iron substitution into sphalerite, ZnS, has been studied systematically by infrared spectroscopy. A range of natural and synthetic compositions, (Zn 1-x Fe x )S, 0 ≤ x ≤ 0.24, were examined. The IR spectrum of pure ZnS contains a single strong absorption band at 320 cm -1 . With addition of FeS, the spectra become broader and shoulders appear. For compositions ≥9 mol% FeS, a splitting of the main peak occurs, and the spectra show two absorption maxima at approximately 300 and 315 cm -1 , respectively. The observation of such extra features does not correspond to the usual behavior observed in other ternary mixed crystals, where either one-, two-, or mixed-mode behavior is observed. The spectra can be deconvoluted into up to three peaks, main Peaks A and B at around 300 and 315 cm -1 , respectively, and a shoulder at around 330 cm -1 (Peak C). The positions and area of the peaks do not change significantly with increasing Fe content. The peak at 315 cm -1 is the main absorption peak of the host ZnS structure, and the peak at 300 cm -1 is an impurity induced mode. An effective linewidth parameter Δ corr was determined by autocorrelation analysis for each spectrum, but there are no obvious trends in the values of Δ corr that can be interpreted in terms of an inhomogeneous distribution of Fe within the sphalerite structure.

A suite of pure magnesian wadsleyite (β-Mg 2 SiO 4 ) samples containing 0.005, 0.38, 1.18, and 1.66 wt% H 2 O was studied by single-crystal X-ray diffraction to determine the effects of hydration on cation ordering and crystal symmetry. Separate compressibility experiments were carried out to 9.6 GPa to determine the effects of hydration on isothermal equations of state. Crystal-structure refinements at ambient conditions show cation vacancies order onto the M3 site. The most hydrous sample (1.6 wt% H 2 O) displayed monoclinic symmetry with β = 90.090(7)°, whereas the samples with lower content were statistically orthorhombic. The density of wadsleyite decreases with increasing water content at STP according to the empirical relation, ρ = 3.470(2) - 0.046(2) CH 2 O g/cm 3 (with CH 2 O in wt% H 2 O). Bulk moduli and pressure derivatives of wadsleyite are K T0 = 173(5) GPa, K 0 ' = 4.1(15) for 0.005 wt% H 2 O; K T0 = 161(4) GPa, K 0 ' = 5.4(11) for 0.38 wt% H 2 O; K T0 = 158(4) GPa, K 0 ' = 4.2(9) for 1.18 wt% H 2 O; and K T0 = 154(4) GPa, K 0 ' = 4.9(11) for 1.66 wt% H 2 O. Variation of the bulk modulus of wadsleyite with water content is non-linear, which may be attributable to softening of the structure by ordering of vacancies onto two non-equivalent M3 sites (M3a and M3b) and an accompanying dilution of orthorhombic symmetry.

The formation of sillimanite, under metastable conditions, relative to corundum and quartz has been defined experimentally, approximately 700 to 800 MPa inside the kyanite stability field thereby allowing for the approximate location of the metastable Sil = Cnd + Qtz equilibrium to be outlined in P-T space from 600 to 800 °C. Experiments involved using a NaCl assembly with a graphite furnace in a two-piston-cylinder apparatus. The thermocouple tip was in direct contact with the flat-lying, folded Pt capsule thereby minimizing thermal gradients to <5 °C. Charges consisted of equimolar amounts of gem-quality sillimanite, corundum, and quartz, plus H 2 O as a flux, placed in a 1.3 cm long Pt capsule that was arc-welded shut and folded. During the course of the experiment, the metastability of the assemblage Sil-Cnd-Qtz implies that Sil ↔ Cnd + Qtz is, at some point, in direct competition with Cnd + Qtz → Ky and Sil → Ky. Early during the experiment it may be assumed that a steady state between dissolution and growth rates is established. However, due to the sluggish nucleation of kyanite, there is a P-T dependent induction period during which Cnd + Qtz → Sil is the controlling reaction. Once kyanite does appear, the reaction proceeds very fast to kyanite via reactions Cnd + Qtz → Ky and Sil → Ky. The kyanite surface area is probably a major factor in controlling the overall reaction rates. Under constant P and T, the system evolves from metastable sillimanite formation to sillimanite consumption, which is only dependent on the kyanite surface area. Similar competition between stable and metastable reactions could occur during contact metamorphism where metastable mineral growth is observed. The relative sluggishness of all three reactions under relatively dry conditions could help to explain the persistence of metastable corundum + quartz ± Al 2 SiO 5 assemblages in nature.

Symplectites composed of Cr-rich spinel plus one or more of the minerals orthopyroxene, clinopyroxene, pyrope garnet, Mg-rich amphibole, and phlogopite, are common in depleted peridotites recovered from kimberlites. The symplectites are unusual for being fine-grained intergrowths in a much coarser matrix of grains, and for having high concentrations of Ca and Al in an environment generally depleted in these elements. Various processes, and combinations of processes, including diffusion, discontinuous precipitation, metamorphism, and metasomatism create and alter the symplectites. Clinopyroxene + spinel symplectites are the most basic symplectites and are assumed to be precursors to all other types of symplectites. These symplectites resemble cellular intergrowths formed in alloys during cooling, by processes of diffusion and by discontinuous precipitation. Garnet in symplectites is interpreted to form by prograde mineral reactions between matrix minerals and symplectites. Symplectites containing amphibole and phlogopite are pyroxene + spinel ± garnet symplectites that have been modified by metasomatism.

The structural evolution of a trioctahedral mica from Cava St. Antonio, Mt. Vulture, Potenza, Italy, has been studied in the temperature range 100-1023 K using in situ single-crystal X-ray diffraction techniques. The sample used shows a Ti-rich composition close to the phlogopite-annite join with the following crystal-chemical formula: (K 0.85 Na 0.11 Ba 0.03 )(Al 0.17 Mg 1.98 Mn 0.01 Fe 2+ 0.29 Fe 3+ 0.37 Ti 0.18 )(Si 2.75 Al 1.25 ) O 10.66 (F 0.13 OH 1.20 Cl 0.01 ). In the present study, the chemical and structural changes and the deprotonation process involved during heating have been characterized. Analysis of the data showed that a, b, and c cell parameters expand almost linearly up to 823 K, while the β angle remains constant. A sharp decrease in the unit-cell dimensions was observed at 873 and 1023 K. Linear trends obtained during reversal experiments from 1023 K to room temperature demonstrated the irreversibility of these events. Structure refinements of single-crystal XRD data collected at 100, 200, 298, 473, 673, 873, and 1023 K converged to 2.14 ≤ R (%) ≤ 8.47, 2.47 ≤ R w (%) ≤ 10.83. In the temperature range 100-673 K, the thermal expansion along the c direction is mainly due to interlayer thickness dilation. The tetrahedral ring approaches the ideal hexagonal shape with increasing temperature to match the expanding octahedral sheet. In the range 873-1023 K, a strong shrinking of the interlayer is associated with the shortening of the M1-O4 and M2-O4 distances and to the consequent reduction of octahedral thickness. Such structural features indicate the occurrence of Fe oxidation process, involving loss of structural H, which is responsible for a phase transition. Mössbauer spectroscopy supported this hypothesis.

A single crystal of a natural orthopyroxene with composition M 2 [Fe 2+ 0.818 Mg 0.156 Ca 0.010 Mn 0.016 ] M1 [Fe 2+ 0.081 Mg 0.767 Al 0.084 Fe 3+ 0.068 ] TA [Si] TB [Si 0.848 Al 0.152 ]O 6 and space group Pbca (sample S95) was investigated at high pressure by X-ray diffraction using a diamond-anvil cell up to 9.56 GPa. No phase transitions were detected in the pressure range investigated. The unit-cell parameters, a, b, and c, decrease non-linearly with pressure and show an axial compression anisotropy with a ratio β a :β b :β c = 1.00:1.64:1.16. The unit-cell volume decreases non-linearly as well and with a negative variation, by about 6.3% up to 9.56 GPa. The equation of state calculated using high-accuracy volume-pressure data up to 5.5 GPa gave the following coefficients: V 0 = 846.02(4) Å 3 , K T0 = 115.4(6) GPa, K' = 7.7(3). Among the Mgorthopyroxenes investigated at high pressure so far S95 shows the highest bulk modulus. Six complete intensity data were collected at 0, 0.16, 1.72, 3.95, 8.03, and 9.56 GPa. The results confirm previous conclusions regarding the compressional mechanism in orthopyroxenes. At lower pressures, compression is mostly connected to a decrease in volume of the two M-coordination octahedra, accompanied by an increased kink in the B-tetrahedral chain. At higher pressures, compression of the M sites decreases, the kink of the tetrahedral chains stops to change, and reduction in unit-cell volume is accompanied mainly by compression of tetrahedra. This change in compressional trends results in a relatively large K' parameter and a pronounced stiffening of the structure with pressure. The presence of Al in the TB tetrahedral site influences the kink of the B tetrahedral chain, which depends on the ratio of sizes of the M2 and TB coordination polyhedra. The increased stiffness of the M polyhedra, caused by the presence of Fe, is the main reason for the high bulk modulus of S95 and its resistance to shortening of the c axis. This explains the limited shortening of the c axis in S95 and the different compressional axial anisotropy with respect to other orthopyroxenes investigated under high pressure.

The high-pressure phase relation of MnSiO 3 was examined up to 85 GPa and 2600 K using a laserheated diamond-anvil cell combined with synchrotron radiation. MnSiO 3 garnet decomposes into a mixture of MnO with a rock-salt structure (B1) + SiO 2 stishovite at pressures higher than ~20 GPa and temperatures higher than ~1200 K. However, MnO (B1) + SiO 2 stishovite further transforms to a perovskite structure with increasing pressure. The phase boundary between these structures is positive in the pressure-temperature diagram. The triple point of garnet, MnO + SiO 2 and perovskite in the pressure-temperature diagram is ~20 GPa and 1200 K. MnSiO 3 perovskite is orthorhombic, and consistent with space group Pbnm, both at high pressure and high temperature and at high pressure and room temperature, but becomes amorphous during decompression. The refined cell parameters of MnSiO 3 perovskite at 85 GPa and 2600 K are a = 4.616(2) Å, b = 4.653(2) Å, c = 6.574(3) Å, and V = 141.2(2) Å 3 . The a/b ratio increases (approaches 1) with pressure and temperature, while the √2a/c ratio remains nearly constant (<1). This indicates that the orthorhombic distortion decreases and the structure tends toward a tetragonal perovskite with increasing pressure and temperature.

Accurate crystal-chemical analysis of complex minerals such as tourmalines belonging to the elbaite-schorl-dravite series was obtained by combining Mössbauer spectroscopy (MS) and structural information. Well-defined relationships were established between the two approaches, leading to a close match of results obtained and a strong link between observed parameters. Although MS information is a powerful tool for quantifying the amount of Fe 2+ and Fe 3+ , it is not always useful in determining their site distribution. In particular, both for Fe 3+ and for (Fe 2+ -Fe 3+ ) interactions structural information is still essential. Fe 3+ MS doublets were identified and assigned to iron in Y and Z sites on the basis of structural information. In a few spectra, Fe 3+ doublets with very low parameters (δ ~ 0.2 mm/s) were observed and, in contrast with the T-site assignment of previous works, were assigned to Fe 3+ in octahedral coordination. Electron delocalization between Fe 2+ and Fe 3+ was observed and related to three different interactions (Y-Y, Y-Z, and Z-Z). Notably, MS hyperfine parameters of Fe 2+ were self-consistent and particularly reliable in determining Fe 2+ site partitioning. Fe 2+ at Y was modeled by three doublets (ΔE Q = 2.45, 2.19, and 1.72 mm/s). The sum of their absorption areas perfectly matches the Y Fe 2+ populations derived from structural data (r 2 = 0.97). The fourth doublet observed (ΔE Q = 1.38 mm/s) is consistent with Fe 2+ at Z, which is an octahedron smaller and less distorted than Y (λ Z = 1.014, λ Y = 1.024). The absorption area of this doublet is highly correlated with the amount of Z Fe 2+ obtained from site-occupancy refinement (r 2 = 0.95). For Y Fe 3+ a link between the quadrupole splitting parameter ΔE Q and variations in the chemical/ structural environment surrounding Fe nucleus was observed. The ΔE Q of Y Fe 3+ increases with ZO 6 volume (r 2 = 0.84) and is linked to the variation of electrical field gradient generated by the Z R 2+ → Z Al substitution. Since the Z skeleton completely surrounds Y islands, ΔE Q of Y Fe 3+ shows much more susceptibility to inductive effects from the second rather than the first coordination sphere.

Rietveld refinements using synchrotron powder X-ray diffraction data were used to study the crystal structure and dehydration behavior of pure monoclinic palygorskite samples from Korea and Alaska. The 300 and 100 K palygorskite structures in air compare well with previous models but provide additional details about zeolitic H 2 O sites and reveal that the Al atoms are ordered into the inner M2 octahedral sites and the Mg cations into the M3 sites at the edges of the tunnels. Real-time, temperature-resolved synchrotron powder X-ray diffraction data and Rietveld refinements were used to investigate the monoclinic palygorskite structure from 300 to 1400 K (in air). Rietveld refinements showed that most of the zeolitic H 2 O is lost by ~425 K, accompanied by a decrease in the unit-cell volume of 1.3%, primarily owing to a decrease in the a unit-cell parameter and an increase in the β angle. The structurally bound H 2 O is lost in two stages, at temperature intervals of 475-540 and 580-725 K. Above ~825 K in air a portion of the Korean sample transformed to a folded structure; the Alaskan sample folded at ~575 K under vacuum. A structure model was refined for the folded structure. At ~1015 K for the sample heated in air, β-quartz diffraction peaks appeared and increased in intensity as heating continued to the maximum temperature. Cristobalite formed above ~1050 K, along with a small amount of clinoenstatite, and both phases persisted to the maximum temperature studied.

Apatite-hosting maskelynite and alkali maskelynite within regions interstitial to cumulus olivine differ compositionally from that found within melt inclusions in the Chassigny martian meteorite. Feldspar glass compositions within the interstitial regions evolve along a high-temperature crystallization path. Within the melt inclusions, feldspar glass shows evolution to low temperatures, extending into the subsolidus regime. Coupled with these differences in maskelynite compositions are differences in volatile abundance in apatite included within the maskelynite. Apatite found within the large olivine-hosted polyphase melt inclusions is uniformly fluorapatite, whereas that interstitial to cumulus olivine is chlor-fluorapatite. We propose that the differences in maskelynite and apatite compositions within the melt inclusion and interstitial regions arose primarily from different crystallization conditions. Melt-inclusion maskelynite and apatite are consistent with nearly closed-system buildup of magmatic volatiles during crystallization within the melt inclusions, exsolution of a Cl-bearing fluid phase, and retention of this fluid phase within the melt inclusions into the hydrothermal regime. For the interstitial regions, however, the higher solidus temperatures of the interstitial melts, the early crystallization of feldspar with significant ternary component, and the Cl-rich nature of the apatite, all suggest open-system fluid migration through the cumulus pile, ingress of Cl-rich, H 2 O-poor brine from a hotter, less evolved portion of the magma plumbing system, and interaction of this Cl-rich fluid with melt prior to the crystallization of feldspar. Such processes of fluid migration through a cumulus pile have been suggested on Earth in layered intrusions, and apatite is the primary recorder of this process.

The mineral eskolaite (Cr 2 O 3 ) has been discovered in association with natural diamond from the Udachnaya pipe in Yakutia, where it is intergrown with an octahedral diamond, mostly as an inclusion in the diamond, but also exposed at its surface. A detailed study was performed on fragments extracted from the outer surface of the diamond, using single-crystal X-ray diffraction (XRD), high-resolution electron microscopy (HRTEM), analytical electron microscopy (AEM), including line-scan and elemental-mapping, EMP, and SIMS. These applied techniques confirmed the nature of the eskolaite with 86.8 wt% Cr 2 O 3 and notable impurities of TiO 2 (3.99 wt%), Al 2 O 3 (2.00 wt%), Fe 2 O 3 (5.83 wt%), and MgO (1.11 wt%). Trace elements, including V (4900 ppm), Mn (129 ppm), Zr (56 ppm), and Nb (32 ppm) were also detected. The entire range of REE is just at or below the limits of detection. A small picrochromite inclusion (X Mg 81.2; Y Cr 94.7) was detected in the eskolaite; its chemistry is typical of chromite diamond inclusions. It also contains minute inclusions of perovskite, corundum, and an unidentified Ti-phase. Nano-sized cavities in picrochromite were determined to consist of carbonate and quench products, including Si, Mg, Ca, P, K, and Cl. This may represent relics of the diamond-forming metasomatic fluids. The eskolaite, containing a picrochromite inclusion, was formed at high pressure within the diamond stability field from C-O-H-bearing fluids containing Ca, K, Cl, P, and possibly even peridotitic (U-type) oxides and silicates.

Lakebogaite, ideally CaNaFe 2 3+ H(UO 2 ) 2 (PO 4 ) 4 (OH) 2 (H 2 O) 8 , is a new Ca-Na-Fe uranyl phosphate mineral from a quarry in Upper Devonian granite near Lake Boga, northern Victoria, Australia. It is associated with Na-analogue of meurigite (IMA 2007-024), torbernite, and saléeite on a matrix of microcline, albite, smoky quartz, and muscovite. Lakebogaite occurs as bright lemon-yellow transparent prismatic crystals up to 0.4 mm across. The crystals have a vitreous luster and a pale yellow streak. Mohs hardness is about 3. The fracture is even to conchoidal. In transmitted light, the mineral is pale yellow with very weak pleochroism: X = yellow, Y = grayish yellow, Z = grayish yellow: dispersion r > v, strong. Lakebogaite crystals are biaxial (+), with slightly variable refractive indices within the ranges: n α = 1.650(2)-1.652(2), n β = 1.660(4)-1.664(3), n γ = 1.681(3)-1.686(2), measured using white light, and with 2V meas = 80-85° and 2V calc = 70-74°. Orientation: Y = b; crystals are elongated along [010], resulting in straight extinction. The empirical chemical formula (mean of nine electron microprobe analyses) calculated on the basis of 30 anions is (Ca 1.00 Na 0.80 Sr 0.10 ) Σ1.90 (Fe 3+ 1.85 Al 0.30 ) Σ2.15 (UO 2 ) 1.80 (PO 4 ) 4.07 (OH,H 2 O) 10.12 . Lakebogaite is monoclinic, space group Cc, a = 19.6441(5), b = 7.0958(2), c = 18.7029(5) Å, β = 115.692(1)°, V = 2349.3(7) Å 3 , Z = 4. The seven strongest reflections in the powder X-ray diffraction pattern are [dobs in Å (I) (hkl)]: 6.60 (100) (110), 3.16 (40) (514̅, 604̅), 4.07 (20) (404̅), 3.80 (20) (314̅), 3.56 (20) (020, 312), 3.31 (20) (114, 220), 2.797 (20) (006). The crystal structure was solved from single-crystal X-ray diffraction data and refined to R 1 = 0.038 on the basis of 5222 unique reflections with F > 4σF. It comprises pairs of edge-shared UO 7 pentagonal bipyramids that are inter-linked via corner-sharing with PO 4 tetrahedra, to form chains parallel to the c-axis. Each UO 7 polyhedron also shares one of its edges with another PO 4 tetrahedron. The (UO 2 ) 2 (PO 4 ) 4 chains are cross-linked via corner-sharing between the PO 4 tetrahedra and Fe 3+ O 4 (OH) 2 octahedra. The octahedra join together by corner-sharing via OH anions to form chains parallel to b. The Na + and Ca 2+ cations, and 4 water molecules occupy eight-sided channels along [010]. The remaining water molecules occupy large ten-sided channels directed along [001] and intersecting the [010] channels. The mineral is named for the nearest township.

Hydroxyl-bästnasite-(Ce), ideally CeCO 3 (OH), had been regarded isostructural with bästnasite-(Ce), CeCO 3 F, the dominant member of the bästnasite family that produces ~70% of the world’s supply of rare-earth elements. Using single-crystal X-ray diffraction and Raman spectroscopy, our structural analysis on hydroxyl-bästnasite-(Ce) shows that the previous assumption is incorrect. The crystal structure of hydroxyl-bästnasite-(Ce) possesses P6̅ symmetry with unit-cell parameters a = 12.4112(2), c = 9.8511(3) Å, and V = 1314.2(1) Å 3 , in contrast to the space group P6̅2c and a ≈ 7.10, c ≈ 9.76 Å, and V ≈ 430 Å 3 for bästnasite-(Ce). Moreover, there are 6, 3, and 5 symmetrically-distinct CO 3 groups, Ce cations, and (OH/F) ions, respectively, in hydroxyl-bästnasite-(Ce), but 1, 1, and 2 in bästnasite-(Ce). The two structures, nevertheless, are similarly characterized by the layers of CO 3 groups alternating with the Ce-(OH/F) layers along the c direction. The Raman spectrum of hydroxyl-bästnasite-(Ce) is dominated by three strong bands at 1080, 1087, and 1098 cm -1 in the CO 3 symmetrical stretching region, along with at least four bands in the OH stretching region. Our study further suggests that natural hydroxyl-bästnasite-(Nd) is most likely isotypic with hydroxyl-bästnasite-(Ce), rather than with bästnasite-(Ce), as previously proposed.