Band 88, Heft 5-6

Mullite (Mul) formation after high-T muscovite (Ms) breakdown has been studied in phyllosilicaterich bricks. At T ≥ 900 °C Ms dehydroxylation is followed by partial melting that triggers the nucleation and growth of Mul acicular crystals. An analytical electron microscopy study reveals that the Mul is a 3:2-type with a [6] (Al 1.686 Ti 0.031 Fe 0.159 Mg 0.134 ) [4] (Al 2.360 Si 1.649 )O 9.82 formula and an O atom vacancy of x = 0.18. This is consistent with X-ray diffraction results [i.e., unit-cell parameters: a = 7.553(7), b = 7.694(7), and c = 2.881(1) Å, V = 167.45 Å 3 ]. The initial stage of the process resulting in Mul growth followed the balanced reaction Ms → 0.275Mul + 0.667Melt + 0.244K 2 O + 0.01Na 2 O + 0.125H 2 O, yielding an alkali-poor peraluminous melt. H 2 O with K (and Na), which are lost along the (001) planes of dehydroxylated Ms, play a significant role as melting agents. The c-axes of the Mul crystals are oriented parallel to [010] ms or to the symmetrically equivalent <310> ms zone axis, while the (120) mul or (210) mul planes are subparallel to (001) ms (TEM results). These systematic orientations point to epitaxial Mul nucleation and growth on the remaining Ms substrate, which acts as a template for Mul heterogeneous nucleation. Randomly oriented Mul growth is also observed during the late stages of the process (i.e., melt cooling). The epitaxial nature of Mul growth after dehydroxylated Ms melting minimizes the energy requirement for nucleation. In addition, the water released after Ms breakdown and the multicomponent nature of the melt enable this high-T aluminum silicate to grow at T ~ 900 °C, almost 100 °C below the SiO 2 -Al 2 O 3 -K 2 O ternary system eutectic (after a melt with an end-member Ms composition).

The reactions of Au 3+ with green rust, stibnite, pyrite, and chalcopyrite were investigated in the laboratory and the size of the gold clusters formed was measured using extended X-ray absorption fine structure (EXAFS) spectroscopy, scanning electron microscopy (SEM), and transmission electron microscopy (TEM). The individual clusters produced were between approximately 29 Å and 77 Å in diameter and may occur individually or as composite clusters from hundreds to thousands of angstroms in size. This work shows that it is possible to form “invisible” gold through reduction of Au 3+ by green rust phases and the surfaces of sulfide minerals.

A new locality for the Ba-rich trioctahedral mica, kinoshitalite [Ba(Mg,Mn,Fe,Al) 3 Si 1.9-2 Al 2- 2.2 O 10 (OH,F) 2 ], has been found in a metamorphosed manganoan marble from Pittsylvania County, Virginia. Metamorphic grade is middle amphibolite facies, with documented P and T of 400 MPa and 575 °C. The locality is along strike not far from the well known Bald Knob, North Carolina, Mn-mineral locality, and appears to represent a silica-poorer analog of Bald Knob. The kinoshitalite occurs in a single layered hand sample containing both skarn and marble layers, and it shows significant compositional contrasts between the two lithologies. Kinoshitalite is scarce and fine-grained in manganoan marble and coexists with kutnahorite, manganoan calcite, fluorian alleghanyite, fluorian sonolite, aluminous jacobsite, and alabandite. Kinoshitalite is both more abundant and coarser-grained in skarn where it coexists with kutnahorite, tephroite, fluorian manganhumite, spessartine, jacobsite, and manganoan magnetite. A-sites in kinoshitalite in skarn are about 3/4 occupied by Ba (with the remainder mostly K), whereas in marble, Ba occupancy of A-sites exceeds 90% and most of the remainder is Ca. X Mg in octahedral sites is >0.6 and is higher in marble than in skarn, whereas X Mn is significant (>0.2) and is higher in skarn than in marble. The VI Al is significantly higher in skarn kinoshitalite, as is total Tschermak content. The total Tschermak content of these barian micas ( VI Al + Ti + Fe 3+ ) is typical of all previously reported kinoshitalites and is significantly lower than that of clintonite and biotite. The X F of kinoshitalite in marble is significantly higher than that in skarn. The petrogenesis of kinoshitalite at the Hutter Mine locality is unclear due to the lack of context for the single mine-dump sample in which the mineral was found and the absence of textural evidence for reactions. However, the two likeliest source minerals for Ba that have been found in the deposit are barite and BaCa(CO 3 ) 2 (probably barytocalcite). One hypothetical reaction to produce kinoshitalite involves decarbonation, in which BaCa(CO 3 ) 2 , rhodochrosite, Mn-garnet, and aqueous fluid react to form kinoshitalite, tephroite (or an Mn-humite), calcite, and CO 2 . A second potential reaction to form kinoshitalite involves barite, Mn-garnet, tephroite, and aqueous fluid as reactants, and kinoshitalite, alabandite, jacobsite, SiO 2 , and O 2 as products.

In mixed-layer illite/smectite containing >80% illite, the presence of an expandable component and changes in its abundance are difficult or impossible to detect with the X-ray diffraction (XRD) method using conventional polar interstratification compounds, e.g., solvation with ethylene glycol. Although high-grade diagenetic and very low-grade metamorphic (VLGM) samples do not expand in response to ethylene glycol treatment, the illite crystallinity (IC) improves significantly with increasing diagenetic/metamorphic grade. Improvement of IC in anchizonal to epizonal samples previously was thought to be related primarily to increase in crystallite thickness and decrease in the number of stacking faults. The present study reexamines the contribution of expandable components to IC in shale and slate samples from the Gaspé Peninsula for both the <2 μm and <0.1 μm size fractions in the high-grade diagenetic and VLGM zone. As tools, we have used XRD patterns and high-resolution transmission electron microscopy (HRTEM) lattice-fringe images of samples treated with n-alkylammonium cations. These studies reveal the presence of expandable components including R1- and R3-ordered structures and “expandable illite.” Expandable 2:1 clay minerals decrease in abundance with increasing metamorphic grade. The improvement of IC with grade is due to the decrease in expandable layers and also to the decrease in the number of lattice defects and an increase in crystallite size. Stacking faults may be the main source of lattice distortion only when phyllosilicates consist of thick non-expandable layer silicates. During late diagenesis, the chlorites become ordered and are stabilized. They are the dominant phyllosilicate minerals present together with illites in the illite-chlorite facies of the anchizone. Lattice fringe images and XRD patterns document the presence of a short-range, corrensite-like R1 ordered phase and thick chlorite packets.

We present X-ray absorption near-edge structure (XANES) spectroscopy of 11 silicate and aluminosilicate minerals and two glasses at the SiK and SiL 2,3 , and OK edges. The similar nearestneighbor environments lead to similar spectral lineshapes at each edge, but the fine-structure differences allow individual and groups of structurally similar minerals to be distinguished. By combining spectra and their first energy derivative from three absorption edges, we show that every mineral studied is distinguishable with XANES. This background work, combined with X-ray PhotoElectron Emission spectroMicroscopy (X-PEEM), allows non-destructive in situ, sub-micrometer (to 35 nm) X-ray analysis of materials, including silicate inclusions, which has not been possible previously. Images and spectra from a 7 μm × 3.5 μm quartz inclusion in zircon are presented as a test of this novel technique in geology.

A single, zoned, Sri Lankan zircon exhibits a range of microstructures from crystalline to nearly amorphous that are the result of radiation damage over a dose range of 2.1-10.1·10 15 α-decay events/ mg (0.16-0.47 dpa). The zones in the crystal were examined at a variety of length scales using optical microscopy, micro-Raman spectroscopy, electron microprobe analysis, and transmission electron microscopy. Birefringence varies linearly with dose: birefringence = -4.71·10 -18 /g·D α + 4.86·10 -2 . Full width at half maximum (FWHM) measurements of the B 1g (ν 3 ) peak, as determined by micro-Raman spectroscopy, were used to estimate the extent of radiation damage in each zone. The radiation dose (calculated on the basis of U and Th concentrations and sample age) vs. damage (from the FWHM measurements of the B1g(n3) peak) relationship among zones was consistent with results for a suite of Sri Lankan single crystals, suggesting that all of these zircon crystals have undergone a similar thermal history. Based on a comparison with zircon crystals that are considered to have undergone minimal annealing (e.g., lunar zircon), the Sri Lankan zircon crystals have accumulated less damage than is expected based on their calculated dose, consistent with previous evidence for annealing. To estimate the extent of annealing in Sri Lankan samples, an equivalent damage dose (i.e., the dose required to produce the same amount of damage in an unannealed sample) was calculated for a given U and Th concentration by determining the time of damage accumulation required to create an equivalent amount of damage in an unannealed zircon. The dose vs. damage relationship in Sri Lankan zircon crystals (560 Ma) is equivalent to that of a ~375 Ma unannealed zircon, suggesting that approximately one-third of the damage has been lost due to annealing. The dose vs. damage relationship is consistent with the direct impact model of damage accumulation. Transmission electron microscopy revealed zoning on a scale finer than could be identified optically (<0.5 μm) and the presence of abundant ZrO 2 nano-particles at zone boundaries.

Rietveld refinements using neutron powder diffraction data were used to locate H atom positions and obtain a more precise crystal structure refinement for akaganéite [Fe 3+ 7.6 Ni 2+ 0.4 O 6.35 (OH) 9.65 Cl 1.25 ·nH 2 O]. Difference Fourier maps clearly showed H atoms positions near those O atoms at the midpoints of the tunnel edges. The O-H vectors point toward the Cl sites at the center of the tunnel, and weak hydrogen bonds likely form between the framework O atoms and Cl. The Cl position is near the center of a prism defined by the eight hydroxyl H atoms. The Cl atoms fill ~2/3 of the tunnel sites, suggesting an ordering scheme in a given tunnel with every third tunnel site vacant. Such an arrangement allows the Cl anions to increase their separation distance along a tunnel by displacing away from one another toward their respective adjacent vacancies. The Fe-O octahedra in akaganéite are distorted with Fe-(O, OH) distances ranging from 1.94 to 2.13 Å and show three longer and three shorter Fe-O distances; as expected the longer distances are associated with the OH- anions. Temperature-resolved synchrotron X-ray powder diffraction data and Rietveld refinements were used to investigate changes in the akaganéite structure and its transformation into hematite as it was heated from 26 to 800 °C. Rietveld refinements revealed surprising consistency in all unit-cell parameters between room temperature and ~225 °C, resulting in nearly zero thermal expansion of the akaganéite structure over a 200 °C interval. Above ~225 °C, the unit-cell volume gradually decreased, primarily in response to decreases in c and b, and an increase in the b angle. The a parameter remained nearly constant until ~225 °C and increased thereafter. Akaganéite started to transform to hematite in the temperature range 290 to 310 °C with no evidence for maghemite as an intermediate phase.

Li-H-rich zincostaurolite locally formed at a metabauxite-marble contact along the eastern coast of Samos during early Alpine high-P, low-T metamorphism. The staurolite, which probably grew from gahnite, cookeite, kaolinite/pyrophyllite, diaspore, and minor Fe-oxide, occurs as prismatic crystals (up to several mm long) in a calcite matrix. It has an unusual composition with 7.5-12.9 wt% ZnO, 1.0-4.9 wt% FeO, 0.66-0.82 wt% Li 2 O, up to 2.6 wt% NiO and 0.59 wt% CoO, and ≤ 0.32 wt% MgO. Calculated structural formulae point to high H 2 O contents (≥4.14 H atoms per 48 oxygen atoms). During late-Alpine uplift of the Samos rocks, the staurolite was, at greenschist-facies conditions, variably replaced by bluish cobaltoan gahnite (X Zn > 0.82), white Na-Ca-Li mica, Ni-rich chlorite (X Ni = 0.42-0.59), zincohögbomite (X Zn = 0.68-0.80), diaspore, and Fe-(hydr)oxide. Detailed EMP work and determination of Li in staurolite and mica by SIMS indicates systematic Zn- Fe-Mg-Ni-Co-Li partitioning between staurolite and its decomposition products, implying local-scale chemical equilibrium. Balanced reaction equations based on mineral-chemical data indicate that the breakdown of staurolite occurred largely isochemically with only introduction of water, Na, and Ca into the reacting system. The breakdown of high-P staurolite during uplift and decompression is thought to be related to the sensitivity of its complex crystal chemistry to changes in physico-chemical conditions.

The systematics of Mn-Fe, Ni-Co, Ti, Cr, and V in olivine from 13 basalt suites from the Earth, Moon, and Mars were studied by electron and ion microprobe techniques. The results demonstrate that chemical signatures in olivine can be related to: (1) planetary parentage, where differences are the result of initial accretional ratios, source compositions, and oxygen fugacity; and also (2) igneous setting and process, where differences among basalt suites within a planet are a consequence of specific redox conditions in tectonic settings, differing melt compositions, and changes in element partitioning resulting from crystallization sequences and mineral modes. Manganese-Fe systematics indicate planetary parentage where the Mn/Fe ratio in olivine increases with increasing distance from the Sun (with the exception of the Moon, which can be explained). This sequence could be the result of initial Mn/Fe accretional ratios from the start of the solar system. Igneous processes such as differing melt compositions and crystallization sequences cause differences in the Mn/Fe ratios of olivine in basalt suites from the same planet. Nickel-Co and Ti systematics show that planetary signatures result from source-region differences among the three planets. For example, the lunar source regions are depleted in Ni and enriched in Ti, as compared with the Earth and Mars, and these characteristics are reflected in the olivine compositions. The differing partitioning behavior of Ni, Co, and Ti in planetary olivine suites is a result of crystallization sequences and initial melt compositions of the basalts during crystallization. Chromium concentrations in olivine result from differing oxygen fugacities and phase stabilities in the source regions of the three planets, whereas V concentrations in olivine are mostly a consequence of the different overall redox conditions on planetary bodies. Both Cr and V show igneous process signatures owing to different melt compositions, crystallization sequences, and modal mineralogy. Perhaps the most important conclusion from this study is that olivine in planetary basalts records information not only about igneous setting and process but planetary parentage as well, making the study of comparative planetary mineralogy an exciting way to gain new insights into basalt petrogenesis.

Analysis of extensive apatite fission-track experimental data reveals several features concerning the influence of crystallographic orientation on measurement biasing that had not been clear previously. At low levels of annealing, anisotropy in confined track lengths is most accurately depicted by an elliptical distribution, in comparison with two suggested alternatives. Quantification of length anisotropy by c-axis projection is simplified and improved, resulting in a model that better achieves its goal of removing anisotropy while preserving all other useful information. The variation of observational frequency of confined tracks with respect to orientation at low annealing levels reveals a large but heretofore-unreported biasing factor related to how an investigator interprets poorly etched tracks. The concept of an “under-etching bias” is introduced to encompass this effect, which can have consequences orders-of-magnitude greater than previously studied factors such as length and etch figure geometry. As annealing progresses, radical shifts in observational frequency with respect to angle make apparent another undocumented process: total annealing of substantial fractions of tracks oriented at high angles to the c axis. Incorporation of this loss into a previous model describing the link between confined track length and density improves the model such that its predictions are now close to the relationship observed in progressive annealing of a single population. Because under-etching biasing effects are intimately linked with orientation and extent of annealing, any future effort to develop a rigorous length calibration scheme to support thermal history modeling should take both of them explicitly into account.

Borocookeite, ideally Li 1+3x Al 4-x (BSi 3 )O 10 (OH,F) 8 (where “x” varies in the range 0.00-0.33 apfu), in which [4] Al is replaced by B relative to cookeite, occurs as a late-stage pocket mineral in the Sosedka and Mokhovaya pegmatite veins, Malkhan gem tourmaline deposit, Chikoy district, Chita oblast, Russia. Borocookeite proper, as well as boron-rich cookeite, is light grey with a pinkish or yellow hue and occurs as a dense, massive crypto-flaky aggregate or thin crusts and snow-like coatings on crystals of quartz, tourmaline, and feldspars from miarolitic cavities. Fragments of elbaite, danburite, and albite are included in the borocookeite mass. In some pockets, the coating is composed of borocookeite and boron-rich muscovite (or boromuscovite) which are not distinguishable visually. Chemical analysis yields (wt%): SiO 2 34.19, TiO 2 0.02, Al 2 O 3 41.77, FeO 0.06, MnO 0.07, MgO 0.04, CaO 0.08, Na 2 O 0.01, K 2 O < 0.01, Li 2 O 4.65, Rb 2 O 0.004, Cs 2 O 0.005, B 2 O 3 4.06, BeO 0.05, H 2 O + 14.17, H 2 O - 0.11, F 1.22, -O = F 0.51, total 100.00. The empirical formula calculated on the basis of 28 positive charges is: Li 1.61 Al 3.80 (Al 0.44 B 0.60 Be 0.01 Si 2.95 ) Σ4.00 O 10 [F 0.33 (OH) 7.81 ] Σ8.14 . The unitcell parameters, calculated from X-ray powder diffraction data, are a = 5.110(4), b = 8.856(3), c = 14.080(6) Å, b = 96.93°(4) these values are smaller than those for cookeite. D calc = 2.69(1) g/cm 3 . The mineral has a Mohs hardness of 3, light pinkish-grey streak, greasy luster, perfect (001) cleavage, and no parting or fracture. Optical properties (for white light): α = 1.574, β = 1.580, γ = 1.591 (all ± 0.002), 2V calc = 72°, dispersion not determined. The optical sign and the angle of the optical axis were not measured because of the small size and strong curvature of the mineral flakes. Borocookeite, as well as other boron-rich phyllosilicate minerals, crystallized from evolved residual solutions in miarolitic cavities at temperatures not less 265-240 °C. The ratio of activities of K, Li, B, F, and H 2 O in the mineral-forming fluids of isolated evolving pockets determined whether borocookeite or boromuscovite formed separately or together.

The energetics of the Saint Claire dickite sample, a highly pure and low defect material, have been studied by high-temperature oxide-melt solution calorimetry. Data were corrected for adsorbed water, the only significant impurity to the ideal formula Al 2 Si 2 O 5 (OH) 4 . Enthalpies of dickite dissolution into molten lead borate at 702 °C and of dickite formation from the oxides and from the elements at 25 °C and 1 bar are 356.3 ± 3.3 kJ/mol, -31.1 ± 4.2 kJ/mol, and -4099.8 ± 4.8 kJ/mol, respectively. Using entropy data from the literature, the standard Gibbs free energy of dickite formation from the elements at 25 °C is -3777.5 ± 4.8 kJ/mol. This value is 7 to 25 kJ/mol more endothermic than literature values obtained for other dickites by calculations, solubility measurements, or calorimetry. Since the Saint Claire dickite is almost pure, the applied corrections are very small and these results are likely to be the most reliable of the available data. The Gibbs free energy of formation of the Saint Claire dickite at 25 °C and 1 bar is 12 to 25 kJ/mol less negative than data obtained for kaolinite using the same and different techniques. This trend is unlikely to reverse at higher temperature and/or pressure in the range of kaolin mineral occurrences. Thus, the new data further corroborate our earlier findings that dickite is thermodynamically metastable relative to kaolinite. This is in agreement with results obtained by mineral synthesis and with most of the observations of kaolinite and dickite occurrences in natural environments. However, further investigations are required to fully explain the transformation of kaolinite to dickite, which appears to occur during diagenesis.

The heat capacities (C P ) of goethite (goe, α-FeOOH), lepidocrocite (lep, γ-FeOOH), and maghemite (mag, γ-Fe 2 O 3 ) were measured from below liquid helium temperature up to their decomposition temperatures by a combination of adiabatic, semi-adiabatic, and differential scanning calorimetry. All three phases were synthetic, with <160 ppm of Al. Chlorine content in goe (32 ppm) and lep (202 ppm) is too low to affect the calorimetric results of this study. Phase purity was verified by Rietveld refinement of the powder X-ray diffraction (XRD) patterns; we determined lattice parameters, atomic positions, crystallite size, and microstrain for all three samples. The Brunauer-Emmet-Teller (BET) surface area is 21 (goe), 23 (lep), and 18 (mag) m 2 /g. No amorphous impurity was found in the goethite sample by extraction of the oxalate soluble fraction. The excess water, determined from weight loss after firing at 1200 K overnight, is 0.083 ± 0.010 (goe), 0.087 ± 0.005 (lep), 0.042 ± 0.003 (mag) moles of water per mole of FeOOH or Fe 2 O 3 . The entropy at standard temperature and pressure (STP) was calculated from subambient C P data and corrected for the excess water content using a Debye-Einstein representation of the C P of hexagonal ice. The entropy at STP is 59.7 ± 0.2 (goe), 65.1 ± 0.2 (lep), and 93.0 ± 0.2 (mag) J/(K·mol). The XRD pattern of maghemite lacks superstructure peaks, and complete disorder of the vacancies leads to configurational entropy S conf = 2.0 J/K·mol. Because very weak superstructure peaks can be overlooked, or the vacancies may be short-range ordered, this calculated S conf represents only an upper limit. The heat capacity above 273 K was fitted to a Maier-Kelley polynomial C P [J/(K·mol), T in K] = a + bT + cT -2 . The C P polynomial coefficients are a = 1.246, b = 0.2332, c = 313900 (goe, valid in temperature range 273-375 K), a = 59.76, b = 0.06052, c = -772900 (lep, 273-390 K), and a = 106.8, b = 0.06509, c = -1886000 (mag, 273-760 K).

The enthalpy of formation from the elements at 298.15 K (ΔH 0 f ) of lepidocrocite (γ-FeOOH) and maghemite (γ-Fe 2 O 3 ) has been measured by acid-solution calorimetry as -549.4 ± 1.4 and -808.1 ± 2.0 kJ/mol, respectively. The ΔH 0 f of goethite (α-FeOOH) was measured by high-temperature transposed temperature drop and acid-solution calorimetry as -559.5 ± 1.1 and -560.7 ± 1.2 kJ/mol, respectively. Mathematical programming analysis (MAP) was used to generate an internally consistent data set for goethite and hematite, using the thermodynamic data presented in this study for goethite, and additional thermodynamic data for hematite and synthesis experiments of Baneyeva and Bendeliani (1973) (BB) and Voigt and Will (1981) (VW). Using BB brackets, the thermodynamic values for goethite were refined to ΔH 0 f = -561.9 kJ/mol and entropy at standard pressure and temperature (S 0 ) = 59.2 J/K·mol; using VW brackets, we arrived at ΔH 0 f = -561.4 kJ/mol and S 0 = 59.5 J/(K·mol). However, MAP failed to include the magnetic transition in goethite, and the derived data should be used with caution. \ Combined with the entropies for the studied phases, the Gibbs free energies of formation from the elements at 298.15 K are -489.8 ± 1.2, -480.1 ± 1.4, and -727.9 ± 2.0 kJ/mol, for goethite, lepidocrocite, and maghemite, respectively. Only hematite (α-Fe 2 O 3 ) and goethite have a stability field in the Fe 2 O 3 -H 2 O system at low to moderate pressures; maghemite and lepidocrocite are metastable at all pressures and temperatures. Goethite is 1.0 ± 1.4 kJ/mol metastable in ΔG with respect to hematite and liquid water, and 2.0 ± 1.4 kJ/mol metastable with respect to hematite and water vapor at 298 K and 50% relative humidity.

Temperature-dependent cation order-disorder has been studied in many 2+ - 3+ oxide spinels but 4+ - 2+ spinels have been found to be either completely normal or completely inverse when examined at room temperature. Here we report the temperature dependence of the cation distribution in the 4-2 spinel synthetic qandilite (Mg 2 TiO 4 ) from in situ time-of-flight neutron powder diffraction experiments to 1416 °C. At room temperature, Mg 2 TiO 4 is confirmed to have completely inverse cation distribution, with Ti atoms occupying half the octahedrally coordinated cation sites. Cation disordering becomes observable above about 900 °C, with 4% of the Ti occupying the tetrahedral site by 1416 °C. The rate of reordering on cooling is fast, such that high-temperature disorder is not preserved on cooling to room temperature. The thermodynamics of the change in cation distribution with temperature can be described by an enthalpy of Mg-Ti disorder of -46.1 ± 0.4 kJ/mol.

Reservoir sandstones from well 30/16-7 in the Upper Jurassic Humber Group, UK North Sea, contain plagioclase feldspars with a range of chemical compositions, microtextures, and origins. The chemical composition of feldspar grains in samples from three different depths in the core ranges from ~0 to 18 mol% An. Two of the samples have a well-defined gap in the compositional range between ~An 4 and ~An 9 . This gap reflects the absence of plagioclase feldspars with those compositions in the ultimate source area of the Humber Group sandstones. Examination of Anbearing feldspars on either side of the gap by TEM shows that they have lamellar and/or tweed exsolution microtextures that are characteristic of plagioclase feldspars having bulk compositions within the peristerite miscibility gap. This is the first time that peristerites have been found in a sedimentary rock, and they suggest a low-grade metamorphic provenance for the plagioclase population. Compositional gaps have been commonly observed in Ab-rich plagioclase feldspars from metamorphic rocks, although the precise reasons why feldspars do not form within the gap remain poorly understood. In addition to the An-bearing peristerites, 30/16-7 contains a varied suite of Anpoor feldspars. The microtexture of some of these grains is consistent with a low-grade metamorphic provenance, whereas others are inferred to be diagenetic in origin and to have formed by albitization of detrital alkali or plagioclase feldspar. Stratigraphic differences in the presence and size of the compositional gap within the Humber Group suggests that plagioclase feldspars may be a powerful new tool for high-resolution stratigraphic correlation and may also be very sensitive indicators of changes in provenance.

Goethite (α-FeOOH) is abundant at the Earth’s surface and has the unusual capacity to adsorb and fix ions from migrating solutions. Understanding the mechanisms by which foreign elements are incorporated into natural goethite has implications for environmental and mining problems. X-ray absorption spectroscopy (XAS) was used to obtain structural information on the local environment around Ni in natural Ni-containing goethite (1.8-4.1 mol% Ni) from Vermelho lateritic deposit of Serra dos Carajás (Brazil) and in synthetic analogues. The data were collected at the LNLS XAS beam line at the Ni and Fe K-edges, at room temperature, and at the Ni K-edge at 8 K. Nickel was found in essentially the same environment in all natural and synthetic samples, with negligible thermal disorder. The coordination polyhedron is a tetragonal dipyramid of oxygen atoms showing that Ni preserves its usual local symmetry. This finding is compatible with a model in which substitution of Ni for Fe is accompanied by a proton capture resulting in NiO 2 (OH) 4 octahedra. The polyhedral linkages are similar to that of pure a-FeOOH, consisting of four shared edges at about the same metal-metal distances, as in the pure mineral. The third and longest metal-metal distance is about 6% larger than the expected corner-sharing distance in the α-FeOOH structure, showing that incorporation of Ni locally distorts and opens the structure.

About 1000 zircon grains from a diamondiferous quartzofeldspathic rock of the Saxonian Erzgebirge were investigated for inclusions using optical microscopy and confocal laser-Raman spectroscopy. Cathodoluminescence imagery was applied to characterize the growth zone of zircon where the inclusions occurred. The most abundant inclusion minerals are microdiamonds. Coesite was not detected. However, garnet and jadeite occur as rare inclusions in zircon cores where diamonds are lacking. Jadeite was detected for the first time in quartzofeldspathic rocks from the crystalline complex of the Erzgebirge. The compositions of the pristine garnets in the zircons are similar to those of core areas of millimeter-sized garnets but the original garnet composition of the early metamorphic stage is only preserved in zircon. Intracrystalline diffusion at temperatures as high as 1000 °C resulted, for instance, in higher Ti concentrations in garnet cores compared with garnet enclosed in zircon. Rutile, quartz, and the compositions of jadeite and garnet inclusions in zircon and of phengite inclusions in cores of large garnets were applied for geothermobarometry. The results, related to an early metamorphic stage, are 650 °C and 18 kbar, which represent conditions at the base of a thickened continental crust before deeper subduction.

This first X-ray Photoemission Spectroscopic (XPS) study of gersdorffite (NiAsS) reveals a strong, asymmetric, singlet peak in the Ni 2p spectrum. It represents Ni of the bulk phase (fully coordinated), bonded to As-S dimers. The high binding-energy tail on the Ni 2p spectrum includes a core-levelshifted Ni surface species and a Doniach-Sunjic high-energy tail characteristic of conductors. A third contribution, likely a hydroxylated Ni species (Ni-OH), is also observed. The XPS characteristics of gersdorffite suggest it is a compound conductor. Identical secondary surface products form on gersdorffite during reaction with both air and aerated, distilled water. Reaction rates in aerated water are more than tenfold greater than in air with arsenic being the most reactive element toward these oxidizing fluids and sulfur being the least reactive. Arsenate and arsenite are produced either in acidic form or as Ni-salts. The S 2p spectrum does not change appreciably with exposure to the oxidized fluids, although minor quantities of surface species of intermediate oxidation state were detected. The Ni 2p signal of gersdorffite becomes weaker relative to the As 3d signal, and the O 1s signal strength increases, as reaction with air proceeds. Changes to the strengths of these signals probably results from accumulation of H 2 O and oxidized As species in the secondary overlayer developed on bulk gersdorffite. The S 2p signal strength decreases with respect to the As 3d signal during reaction with air, probably because S is “buried” beneath the hydrated, As-enriched secondary overlayer (overlayer deficient in Ni and S). In contrast to air-reacted surfaces, the Ni 2p signal of water-reacted surfaces remains strong relative to the intensity of the As 3d signal. This may be explained by the high solubility of arsenites and arsenates, relative to Ni(OH) 2 , present at the oxidized gersdorffite surface.

A universal absorption coefficient was determined for quantitative analysis of OH and H 2 O in feldspars using infrared spectroscopy. 1 H MAS (magic-angle spinning) NMR spectroscopy was used to determine the H concentration in three alkali feldspars, and for the first time, eight plagioclase samples. To accurately measure structural H concentration in samples with low H (<1000 ppm H 2 O) it was necessary to eliminate the signal due to adsorbed water in the powdered NMR sample. The pegmatitic and metamorphic albite samples are transparent, but contain variable (40-280 ppm H 2 O) concentrations of microscopic to sub-microscopic fluid inclusions. The pegmatitic albites also have sharp bands in the mid-IR similar to the OH bands found in quartz. The other plagioclase samples used in the IR calibration have broad anisotropic bands around 3200 cm -1 in the mid-IR and weak combination stretch-bend bands near 4550 cm -1 in the near-IR, indicative of structural OH. The OH vector in plagioclase is preferentially aligned parallel to the crystallographic a axis. The concentration of structural OH in the plagioclase samples ranges from 210-510 ppm H 2 O by weight. The microcline samples contain structural H 2 O molecules (1000-1400 ppm H 2 O) and the sanidine sample contains structural OH (170 ppm H 2 O). An approximately linear trend is produced when the total integrated mid-IR absorbance is plotted vs. the concentration of structural H determined from NMR (OH and H 2 O) for plagioclase and alkali feldspars. The integral absorption coefficient for the total mid-IR peak area is 15.3 ± 0.7 ppm -1 ·cm -2 [107000 ± 5000 L/(mol H 2 O·cm 2 )] for natural feldspar samples that contain structural OH or H 2 O. Measurements of band areas of unpolarized IR spectra on (001) cleavage fragments provide an estimate of H concentration for alkali feldspars, but this method does not work for most plagioclase samples.

Large diamonds (10-200 carats) from Udachnaya, Mir, and Aikhal kimberlite pipes of Yakutia contain variable mineral inclusions typical of diamonds of smaller size. Among these inclusions, extracted from outer portions of these diamonds, the majority are garnets of harzburgitic paragenesis, with a small number of inclusions belonging to lherzolitic and eclogitic groups, in addition to chromite diamond inclusions (DIs). The most striking feature of the garnet DIs is the sinusoidal rareearth elements (REE) patterns observed for all harzburgitic, and some lherzolitic garnets, whereas eclogitic garnets form normal LREE-depleted patterns. These sinusoidal REE patterns are indicative of a complex petrogenesis for the garnets, involving partial-melting, followed by metasomatic enrichment, particularly of the LREEs/MREEs. The metasomatic agents were probably not those of carbonatitic fluids, but rather C-O-H-N-S fluids, as indicated by the negative anomalies of Ba and Sr in the garnets. T he general axiom among investigators that most, if not all, DIs are syngenetic with their diamonds is seriously questioned. The petrogenetic history of the harzburgitic garnet DIs involved complex processing prior to their diamond encapsulation. That is, the garnet DIs retained a signature of their earlier, pre-diamond encapsulation history. This result is taken as proof that basically all harzburgitic garnet DIs are non-syngenetic (i.e., protogenetic) with their host diamonds.

A new force field for modeling calcium carbonate has been derived that corrects deficiencies of previous models. The model correctly reproduces the structure of the gas phase species, as predicted from ab initio calculations, as well as the bulk structure and properties of calcite and aragonite. With this new model, a (2 × 1) reconstruction is predicted to occur for the dominant (101̄4) surface of calcite, involving rotation of half of the surface carbonate anions. This reconstruction matches the results of low energy electron diffraction measured in vacuo and provides the first independent verification of this observation, as well as yielding the atomic detail of the nature of the reconstruction. While there is only a small exothermic energy associated with the formation of the supercell, the presence of an imaginary phonon mode at (½0) in the two-dimensional Brillouin zone for a single surface cell verifies the existence of the reconstruction.

A natural α-quartz disc with flat surfaces perpendicular to (100) was sheared at 5 GPa in order to confirm that the sheared state at high pressure is not always the same as that obtained at high hydrostatic pressure. A transmission X-ray diffraction pattern taken from the sample that was sheared at high pressure revealed five broad halo rings; the pattern was taken about three months after the highpressure experiment. The ratios of the interplanar spacing corresponding to the densest radius of the smallest ring to those of the other ones are 0.610, 0.514, 0.459, and 0.399, respectively. This set of rings does not appear from α-quartz. These are approximately equal to those obtained from solids with space group Fd3m. A Raman spectroscopic study also showed the structural change due to shearing at high pressure. The spectrum appears neither from α-quartz nor from coesite. These facts indicate, therefore, that the crystalline α-quartz sample transformed to the amorphous state based on a structure consistent with Fd3m space group.