Volume 89, Issue 1

The rare mineral indialite, a hexagonal high-temperature form of cordierite, has been discovered for the first time in Italy at Somma-Vesuvius volcano in unusual pyrometamorphic ejecta. The indialitebearing xenoliths are composed of several pale yellow fragments consisting of very fine-grained indialite with traces of cristobalite. On the surface of the fragments, colorless indialite and pale-blue, prismatic osumilite crystals form whitish crusts, with subordinate phlogopite and trace amounts of corundum. Gypsum and Fe-oaxides are ubiquitous. Chemical composition of the Italian indialite is characterized by high X Mg (~0.95) and significant K 2 O (~1.4 wt%). Single-crystal X-ray diffraction measurements give cell dimensions very similar to those of the indialite from the Bokaro coalfield. The diffraction pattern shows no significant deviation from the symmetry conditions of the space group P6/mcc, which is the space group generally accepted for high-cordierite. The presence of guest K (and lesser Ca) in the sixfold channel does not produce appreciable distortions in the framework of T2 tetrahedra, probably because of the weak interactions between guest and the ring O 2 atoms. Indialite-bearing rocks originated through a pyrometamorphic process at shallow depth and at maximum temperature close to ~1200 °C. Rapid crystallization and subsequent quenching favored a higher symmetry, thus “freezing” the space group P6/mcc.

Chromian clinopyroxenes with exsolution textures were found in high-pressure (HP) tremolite schist from the Osayama serpentinite melange in the Chugoku Mountains, Southwestern Japan. Chromian omphacite [jadeite (Jd) 23.3-41.4 diopside (Di) 46.7-60.7 kosmochlor (Ko) 4.5-19.6 aegirine (Ae) <7.0 ] with up to 6.6 wt% Cr 2 O 3 occurs as neoblastic crystals in a foliated tremolite-rich matrix, or as pseudomorphs after relict chromian spinel, and contains irregular-shaped thin lamellae of chromian diopside (Jd 3.8-18.2 Di 72.5- 93.6 Ko 1.1-13.9 Ae <3.4 ). Chromian omphacite contains roughly constant jadeite plus kosmochlor components at 38.2-51.6 mol%; this is equivalent to the jadeite component of ordered P2/n omphacite. Systematic analyses of coexisting chromian pyroxenes yield a clear immiscibility gap between “omphacite” and “diopside.” The compositional gap becomes much narrower with increasing Ko component; addition of only 10 mol% Ko component narrows the omphacite-diopside gap by an order of magnitude. Such an effect is similar to, but more effective than, the introduction of Fe 3+ on the omphacite-diopside immiscibility gap. Chromian pyroxenes replacing relict chromian spinel are associated with other chromian silicates including phengite, chlorite, and pumpellyite. The wide compositional gap of chromian pyroxenes and the presence of chromian pumpellyite and Si-rich chromian phengite indicate T < 300-400 °C and P > 0.8 GPa. This P-T estimate is consistent with parageneses of minerals in the host serpentinite. The variation of Cr content in chromian silicates reflects the extent of Cr ↔ Al substitution, and may be related to a chemical heterogeneity of the Cr-bearing fluid.

The 2111 pyribole chain sequence, where 2 signifies a double (amphibole-like) chain and 1 indicates a single (pyroxene-like) chain, has been reported from geological samples. Such material is commonly associated with slabs that contain multiples of three single chains, which we shall call “3n single-chain slabs” (where n = 1, 2, 3,…). We generated 2111 chain sequences and associated 3n single-chain slabs by heating natural chesterite crystals at ca. 1000 ºC in the laboratory as well as by their prolonged irradiation inside a 200 kV transmission electron microscope. Many of the synthesized 2111 sequences are intergrown with narrow regions containing isolated single- and double-chain material that contains, respectively, 3n single chains and n double chains. The single-chain triplets presumably formed as a result of the decomposition of isolated triple-chain slabs that were originally intergrown with the chesterite. The 2111 sequence and the 3n single-chain slabs can form, respectively, by decomposition of chesterite and triple-chain material during observation or perhaps during ion milling. Recognizing such artifacts is important when attempting to interpret geological history from the details of biopyribole structures and intergrowths.

Buck Creek metatroctolites preserve evidence of three stages in the geologic history of the region: a relict igneous stage; a high-pressure, granulite-facies metamorphic stage; and a hydrous alteration stage. During high-pressure granulite-facies metamorphism, adjacent olivine and plagioclase reacted to produce complex coronitic textures. Relict olivine is surrounded by radial, columnar orthopyroxene, and plagioclase is rimmed by fine-grained symplectite composed of separate grains of clinopyroxene/ spinel, tschermakitic amphibole/spinel symplectite and minor clinopyroxene/sapphirine symplectite. Examination of textural relationships in back-scattered electron images and X-ray element maps of the symplectites demonstrates that amphibole/spinel and clinopyroxene/spinel symplectites grew simultaneously at high temperature to replace plagioclase in these rocks. Clinopyroxene/sapphirine symplectite grew in the latest stages of plagioclase replacement in some samples. A successful model for the main corona-forming reaction in sample BC7, 1.00 Pl + 1.72 Ol + 0.17 H 2 O = 0.59 Di + 0.17 Amp + 0.68 Spl + 0.79 Opx, results only if amphibole is included. The model predicts relative volumes of product clinopyroxene, amphibole and spinel that are generally consistent with the proportions in imaged symplectites. Proportions of clinopyroxene- and amphibole-hosted symplectite appear to have been governed by local mass-balance of Na and Al in plagioclase and/or availability of H 2 O. The most successful model for formation of the inner clinopyroxene/sapphirine corona, 1.00 Pl + 1.47 Ol + 0.17 H 2 O = 0.59 Di + 0.17 Amp + 0.30 Spr (10 ox.) + 0.57 Opx, also includes amphibole, which is present in association with clinopyroxene/sapphirine symplectite. Sufficient H 2 O for formation of symplectic amphibole must have been present in the intergranular fluid of relict igneous troctolites or have infiltrated during metamorphism at high P-T conditions.

We have studied three synthetic Al-Fe 3+ zoisites with variable Fe 3+ -contents [X ps = 0.0, 0.035 (12), and 0.062 (23)] by means of temperature-dependent infrared powder spectroscopy. Spectra were recorded from -150 to 170 °C in steps of 20 °C in the lattice vibrational range 400 to 1200 cm -1 . At -150 °C, a total of 34 bands could be identified in all samples. With increasing temperature, the bands broaden and generally shift smoothly to lower wave numbers. The maxima of bands at ~620 and ~909 cm -1 as well as the autocorrelation analysis of the spectral region 820 to 1050 cm -1 display linear but different temperature dependencies of the spectral properties at low and high temperatures. These changes in temperature dependence of the spectral properties marks the isosymmetric displacive phase transition between zoisite I and II. The transition temperatures T tr are about 45 °C for Xps = 0.0 and = 0.035 (12), and about -35 °C for X ps = 0.062 (23) and suggest a negative correlation between T tr and X ps . A combination with literature data yields a positive slope of 9.4 ± 2.0 MPa/K and a ΔS of 0.5 ± 0.1 J/(K·mol) for the zoisite I → zoisite II transition for a composition of about Xps = 0.035.

The kinetics of the exsolution of pentlandite from the monosulfide solid solution (mss) have been investigated using a series of anneal/quench and in situ cooling neutron diffraction experiments. Five mss compositions were examined by anneal/quench techniques covering the composition range Fe 0.9 Ni 0.1 S to Fe 0.65 Ni 0.35 S and using annealing temperatures between 423 and 773 K for periods from I h to 5 months. In situ cooling experiments were performed on four mss compositions in the range Fe 0.9 Ni 0.1 S to Fe 0.7 Ni 0.3 S. The samples of these solid solutions were heated to 973 K, and then cooled to 373 K in steps of 50 K over a 24 h period. The extent of exsolution was monitored by Rietveld phase analysis using powder neutron diffraction data. The anneal/quench experiments established that initial exsolution of pentlandite from mss above 573 K is very rapid and is effectively complete within 1 h of annealing. However, the mss/pyrrhotite compositions remained Ni rich (17 at% Ni) after 5 months annealing, indicating that compositional readjustment at low-temperatures occurs over long periods. Below 573 K, exsolution is less rapid with rate constants in the range 6 × 1O -6 to 1 × 10 -5 /s and the activation energy for exsolution of pentlandite from mss Fe 0.5 Ni 0.2 S between 473 and 423 K is 5 kJ/mol. The in situ cooling experiments showed that the temperature at which exsolution commences upon cooling decreases from 873 K for Fe 0.7 Ni 0.3 S to 823 K for Fe 0.9 Ni 0.1 S and that exsolution effectively ceased on the time scale of the experiments at temperatures between 598 and 548 K. The kinetic data were analyzed using the Avrami model where y = 1 - exp(-k n t n ) and the initial rates of exsolution were found to increase with Ni content from 2 × 1O -6 Zs for Fe 0.9 Ni 0.3 S to 4 × 10 -5 /s for Fe 0.7 Ni 0.3 S. Both high Ni content and high M: S ratio served to facilitate nucleation rate, indicating that nucleation occurs at S vacancies within mss crystals rather than at grain boundaries. Values of the Avrami geometric constant n vary during exsolution upon cooling indicating three possible changes in the growth mechanism during the reaction. The roles of impurities and S fugacity on reaction rates are discussed. The rate constants for exsolution of pentlandite from mss/pyrrhotite in nature are estimated to be 4 or 5 orders of magnitude slower than those reported here, still very rapid on a geological time scale. High metal mobility persists in this system at low temperatures, even at room temperature, and the textures and compositions observed in nature are a consequence of very low-temperature (<100 0 C) equilibration of assemblages over geological time scales.

Here we introduce a technique for simultaneous measurement of surface normal retreat rates of specific cleavage faces by vertical scanning interferometry and the bulk dissolution rate of a mineral powder. A hydrothermal reactor is used to contain both a well-characterized powder and oriented single crystals with a masked reference surface at elevated temperatures. We show examples using both anorthite and albite reacted at temperatures between 150 and 200 °C. In the case of albite, dissolution rates of fine-grained powders are substantially enhanced compared to those prevailing on large single-crystal cleavage surfaces. Rates developed on the (010) albite cleavage surface are also substantially faster than those on the (001) face, where etch-pit development was relatively modest and surface normal retreat was not detectable within the time frame of the experiment. The reasons for this difference are not immediately clear, but may be related to anisotropy in the distribution of Al-O-Si vs. Si-O-Si bonds in the albite structure, (010) twinning expressed on the (001) surface, and possible disruption of kink propagation across the twin plane.

We present a new, high-quality electron microprobe method for analyzing F in Fe-bearing minerals and glasses down to trace levels (<1000 ppm F). This method is empirically based and corrects for the problem of overlap of the FKα peak onto the “shoulder” of the FeLα1 peak that arises when using synthetic multi-layered diffraction crystals. It also achieves high precision and accuracy while maintaining low detection limits and a small beam diameter. Analytical conditions for F using the new method and a synthetic W/Si diffraction crystal are: 10 kV accelerating voltage, 180 nA beam current, 8 μm beam diameter, and 400 second total peak count time. An iterative beam exposure method is employed to minimize potential damage to the sample due to beam heating. The method presented here can be used to address problems related to degassing of F at active volcanoes, the reservoirs of F in the mantle, and the partitioning of F between minerals and silicate liquid.

Major- and trace-element and isotopic analyses of single crystals can reveal temporal changes in the thermodynamic and mechanical behavior of magmatic systems. Additionally, zoning can be used to define crystal populations that have shared a common environment based on the correlation of zoning profiles among groups of crystals. However, as is evident in petrographic thin sections, comparison of zoning patterns is complicated by the geometric distortions resulting from crystal sections that are offcenter. Because of these distortions, identification of crystal populations becomes increasingly difficult with increasing degrees of geometric non-ideality. In addition, because of the inherent complexity of zoning profiles, often there is ambiguity in determining what level of correlation is significant, even for ideal cases. Consequently, significance levels must be determined for each correlation technique and set of natural profiles. We evaluate the effectiveness of standard correlation, wavelet-based correlation (WBC), and profile-normalization techniques designed to counter the effects of non-ideality using Monte Carlo experiments correlating synthetic profiles. Results from the experiments show that adaptive profile normalization provides more significant correlations from fewer profiles than other techniques indicating that it is at least partially effective in counteracting the effects of non-ideality.

The formation mechanisms of tremolite in a time series of hydrothermal experiments at 850 °C and 6.1 kbar were studied using transmission electron microscopy (TEM), analytical TEM, and Xray diffraction. The starting materials were unseeded mixtures of reagent-grade oxides. After the first 8 min of reaction, the synthesis products were diopside, enstatite, clinoenstatite, quartz, and minor C-centered and primitive monoclinic amphibole (13 wt% total). The compositions of the firstformed amphiboles vary considerably, from essentially pure cummingtonite (Ca/Mg = 0) to nearly ideal tremolite (Ca/Mg = 0.4). Three groupings of amphiboles with successively higher Ca/Mg ratios were identified, respectively 0.1, 0.22, and 0.35. Over a period of 97 h, the proportions of Mg-rich amphibole gradually decrease, and the average Ca/Mg ratios of all amphiboles steadily increase from about 0.22 to 0.35. Other changes include: (1) a decrease in the aspect ratio of amphibole crystals with increasing time; (2) a decrease in the number of non-amphibole chain multiplicity faults (CMFs), from early A′(2) values as low as 0.81 to a final value of 0.94; and (3) a distinct change in the rate of amphibole growth after 1 h of synthesis. These observations suggest that two mechanisms of amphibole formation are dominant: (1) formation of pyroxene modules with subsequent hydration and arrangement into the double-chain amphibole structure and (2) direct nucleation and growth of amphibole from hydrothermal solution. A third mechanism involving formation of talc-like modules, followed by their breakdown into double-chain modules, incorporation of Ca, and ordering into the amphibole structure, may also occur but probably only at the lamellar or unit-cell scale, because these syntheses were performed outside the stability field of talc.

Enthalpies of solution of the NH 4 + -bearing minerals buddingtonite (NH 4 AlSi 3 O 8 feldspar) and tobelite [NH 4 Al 2 (AlSi 3 O 10 )(OH) 2 mica] have been measured in 20.1 wt% hydrofluoric acid (HF) at 50 °C. Utilizing additional HF solution calorimetric data from this laboratory for sanidine, albite, muscovite, paragonite, halite, sylvite, salammoniac, gibbsite, and water, along with enthalpies of formation from Robie and Hemingway (1995) for the same phases, the heats of formation at 298.15 K of buddingtonite and tobelite have been determined to be -3883.9 and -5881.3 kJ/mol, respectively. Entropies of ~247 and ~319 J/(mol·K) for these minerals (298.15 K) have been approximated from various ion-exchange equilibria involving feldspars, micas, and chlorides. Collectively, these data result in 298.15 K Gibbs free energies of formation from the elements of -3579 and -5422 kJ/mol, respectively, for buddingtonite and tobelite, values that differ by 40 to 50 kJ/mol from those given previously by Mäder et al. (1996). The calculated phase equilibria that result from the data of this investigation are reasonable relative to the conditions required to synthesize both buddingtonite and tobelite (Harlov et al. 2001b, 2001c).

The thermal dehydration process of the natural Li zeolite bikitaite has been studied in situ by synchrotron radiation powder diffraction. The temperature-resolved experiments were performed using a translating imaging plate system. Rietveld refinements were carried out on 42 powder patterns in the temperature range from room temperature to 800 °C. Bikitaite is stable at least up to 800 °C, the temperature at which the phase transition to γ-spodumene begins. The dehydration process begins at about 200 °C, affecting the two water sites in a similar way, and is complete at 468 °C. Such a process induces only very minor structural distortions in the framework, due to the anti-rotation of the internal T-O-T hinges. In this sense, bikitaite can be defined as a flexible but non-collapsible framework, and it is the zeolite that undergoes the lowest heating-induced distortion among those studied up to now. The high thermal and structural stability suggest that anhydrous bikitaite could be used as a solid porous matrix for embedding nanosized materials in its one-dimensional channels.

High-temperature behavior and the process of thermal dehydration in the natural zeolite bikitaite have been studied by ab initio molecular dynamics simulations, and favorably compared with the X-ray powder diffraction data presented in Part I of this work (Ferro et al. 2003). The microscopic dynamical behavior of the extraframework species (water molecules and Li cations) has been characterized as a function of temperature. Two regimes have been detected, and the transition is characterized by the breaking of the one-dimensional water chain typical of bikitaite at room temperature. The elementary steps for the diffusion of water inside the bikitaite channels have been studied by means of a rare-events-sampling technique (Bluemoon Ensemble). The activation free-energy for a site-to-site water jump has been calculated and a mechanism for the dehydration process is proposed.

The Cañada pegmatite (Salamanca, Spain), a Li-P-(Sn-Nb ± Ta)-bearing granitic pegmatite, is intruded partly into a gabbro and partly into a leucogranite. Three phosphate associations have been distinguished, based on paragenesis, texture, and chemistry: (1) ferrisicklerite-magniotriplite-johnsomervilleite association (in the border zone), in which Mg-bearing phosphates and coexisting Fe-Mg silicates are abundant; (2) ferrisicklerite-graftonite association (in the transition zone) characterized by higher values of Fe/(Fe + Mg) than in the border zone; and (3) triphylite-sarcopside association (in an inner zone), which also includes minor montebrasite, ferrocolumbite, and cassiterite, typical of an evolved pegmatite facies. As a result of the decrease of Mg, the Fe/(Fe + Mg) ratios for phosphates, biotite, and tourmaline increase from the border to the inner association (e.g., for ferrisicklerite and graftonite, from 0.67 and 0.85 in the border to 0.94 and 0.98 in the inner association, respectively). This difference is particularly evident for biotite and tourmaline; for example, the Fe/(Fe + Mg) ratios for tourmaline range from 0.59 in the border to 0.86 in the inner zone. These variations seem to reflect contamination of marginal zones of the pegmatite by some type of reaction with the host gabbro. Thus, an evolutionary trend involving inward crystallization from the margins and contamination of fluids from wallrocks into pegmatite-forming melt may be a plausible genetic model. The occurrence of phosphates along with Fe-Mg silicates would indicate that the melt contained on the order of 1.3-2.4 wt% P 2 O 5 , based on experimental silicate-phosphate equilibria.

A model is presented to calculate crystal size distributions (CSDs) for coarsened mineral aggregates. The equations consider the second-phase particles inside coarsening aggregates. This approach is different from other published kinetic growth equations. The proposed model includes grain size limits due to the presence of second-phase particles. These limits control the behavior of growing grains. The physical basis for this model is taken from a neighborhood-coarsening model. The model is able to compute coarsening of a given set of crystal sizes, to simulate the evolution of the CSD, and to describe the influence of the amount and size of the second-phase particles. The model is limited to rocks that can be described as matrix with second phases. For very low second-phase concentrations (volume fraction <0.01), the model gives results similar to Lifshitz-Slyozov-Wagner models (LSW). In the case where the second-phase content is extremely high (volume fraction >0.5), the model would not allow coarsening or the system can be no longer described as matrix and second-phase particles. Depending on the size and amount of second phases, the CSD develops similar to LSW at high growth rates, but intermediate growth rates produce CSDs that are unknown in other closedsystem coarsening models. To test the model, natural data from a contact-metamorphic calcite marble with different mica contents have been compared with simulated CSDs. The measured and simulated CSDs can be well described by the proposed model for variable amounts of second-phase particles. The two-phase model is applicable to impure carbonates, mica-bearing quartzites, or impure dunites. The proposed model has interesting applications to experimental data, where porosity may influence grain coarsening.

Si oct -O-Si tet -O Infrared reflectance, thin-film absorbance, and polarized Raman spectra were acquired from -100 to -4000 cm -1 from polycrystalline garnets along the pyrope (Py = Mg 3 Al 2 Si 3 O 12 )-majorite [Mj = Mg3(MgSi)Si 3 O 12 ] join. We measured Py 100 , Mj 39 , Mj 45 , Mj 80 , and natural single-crystal, nearly endmember pyrope. Consistency is obtained between measured absorption spectra and absorption coefficients calculated through Kramers-Kronig analysis of reflectivity (R) data, if R-values are scaled to account for losses through scattering. The widths of the lattice modes double from each end-member toward the middle of the binary, whereas frequencies are nearly independent of composition. Strong polarization of Raman peaks of all our samples except pyrope shows that the space group is reduced from cubic, even for the Al-rich majorites. The largest number of IR or Raman peaks is observed for Mj 80 . The number of peaks for Al-rich majorites lies between those seen for cubic pyrope and I4 1 /a majorite. The behavior of symmetry-breaking modes supports the I4 1 /acd space group, which was inferred from X-ray data for Mj contents near 30 to 45% (Nakatsuka et al. 1999a). Also, the internal mode types that are expected for isolated SiO 4 tetrahedra in any garnet-like structure do not fully describe the patterns of peaks seen for some intermediate compositions along the Py-Mj join. The appearance of Si oct -O-Si tet bending modes in the available spectral data for Mj contents near 30-40 and 80-90%, but not near 45-50%, is attributed to two-mode behavior. Specifically, disruption of the structural chain -Si oct -O-Si tet -O-Si oct -O-Si tet - in Si-rich majorites by units such as alters the electrostatic balance so that the internal modes are not entirely described by motions of isolated SiO 4 tetrahedra. Short chains are formed at both high and low Al contents, promoting coupling of Si octahedral and tetrahedral vibrations. Near Mj 50 , each tetrahedron is linked (on average) to four octahedra containing two Al-, one Si-, and one Mg, and the Si oct -O-Si tet bending modes disappear. Order-disorder in majorite garnets is characterized not only by increased peak widths, but also by the presence of modes due to the bridging O atoms, whereas splittings and the appearance of new peaks indicate the space group for non-cubic garnets.

The modulated crystal structure of an antigorite polysome with m = 17 was refined by single crystal X-ray diffraction in the Pm space group, using highly ordered single crystals from Val Malenco, Italy. The chemical composition is (Mg 2.673 Fe 2+ 0.098 Fe 3+ 0.015 Al 0.035 Cr 0.007 Ni 0.003 Mn 0.002 ) Σ=0 . 823 (Si 1.997 Al 0.003 ) Σ=2O5 (OH) 3.639 . Lattice parameters [a = 43.505(6), b = 9.251(1), c = 7.263(1) Å, β = 91.32(1)°] were determined using a single-crystal diffractometer equipped with an area detector. The structure was refined using 9242 independent reflections, obtaining a final R4σ factor of 0.0577. A continuous, wavy octahedral sheet is linked to a tetrahedral sheet with tetrahedral apices alternatively pointing +c and -c. This sheet is located on the concave side of the octahedral-sheet wave. The octahedral sheet shows normal thickness for a serpentine of this composition, and does not have any internal offset. The tetrahedral sheet inverts its polarity through six- and eight-membered tetrahedral rings (6- and 8-reversals). Between reversals, 6-membered rings are distorted toward ditrigonal configuration, with tetrahedral rotation, α values, ranging along the wave from 4 to 13.6°. The two half-waves have curvature radii of 99.4 and 110.9 Å. Variable interlayer O-O distances occur, indicating the absence of homogeneous, continuous hydrogen bonding. The bond geometry, very similar to that of lizardite, suggests common crystal chemical and geochemical properties. The larger stability field of antigorite compared to lizardite is interpreted to arise from the occurrence of three-dimensionally connected chemical bonds.

The structures of four mercurian tetrahedrite samples, [IV] [(Cu + )4(Hg,Zn,Fe) 2 ] [III] (Cu + )6 [III] (Sb,As) 4 [IV] S 12 [VI] S, with Hg-contents ranging from 0.32 to 1.15 apfu were refined by least-squares in space group I4̅3m to agreement factors R < 0.027. The α cell dimension varied from 10.3310(2) to 10.4033(2) Å; both the α cell dimension and the size of the [IV] M1 site vary linearly with the Hg content. The magnitude of the observed variation in the α cell dimension partially validates the empirical relationship developed by previous investigators. The degree of distortion and amount of rotation of the [IV] M1 tetrahedron as a function of the M1-S1 bond length is comparable to that predicted by the empirical relationships of previous investigators when the As content is taken into account. The spinner blade length remains virtually unchanged because the Hg substitution takes place at M1 and the As:Sb ratio varies only slightly. The structural trends as a function of composition in these mercurian tetrahedrites closely follows those previously predicted by modeling using ionic radii and

In previous papers, we investigated via Monte Carlo simulation the order-disorder behavior of an individual octahedral phyllosilicate sheet, with respect to two-species systems Al/Fe, Al/Mg, and Fe/Mg, and some three-species systems Al/Fe/Mg that were relevant to clay compositions found in nature. We have extended the work on Al/Fe/Mg systems to include a wide range of different octahedral compositions that can represent different natural and synthetic clay minerals, by means of Monte Carlo simulations based only on atomistic models. In many cases, phase transitions do not occur, in that long-range order is not attained, but most systems exhibit short-range order at low temperatures. The ordering of the octahedral cations is highly dependent on the cation composition.

We have investigated by Monte Carlo simulation the Al/Si ordering behavior of the tetrahedral phyllosilicate sheet, with a variety of compositions from Al 1 Si 1 to Al 1 Si 7 , using atomic interaction parameters determined for the tetrahedral sheet in muscovite. Three different ordering schemes operate, depending on composition, with relatively Al-poor systems ordering in a muscovite-like (Al:Si = 1:3) pattern and relatively Al-rich compositions ordering in an “ABABAB” or margarite-like (Al:Si = 1:1) pattern, where ABABAB indicates the arrangement of atoms around a hexagonal ring of tetrahedral cation sites. The pattern corresponding to Al:Si = 1:2 occurs in intermediate compositions, but always in conjunction with another ordering pattern, except for one composition close to Al:Si = 1:2. Simulations of the same composition but with different ordering schemes can show different behavior, and this is evidence for metastability fields. The transition temperature for order-disorder T c is strongly dependent on composition, and the dilution effect can be observed at low Al concentrations, with a critical concentration x c between 0.12 and 0.15.

The compressibility of zircon and reidite (ZrSiO 4 ) was investigated by powder X-ray diffraction (XRD) up to 20 GPa using a synchrotron radiation source. Reliable equations of state for zircon and reidite were obtained using a diamond anvil cell (DAC) coupled with yttrium-aluminum-garnet (YAG) laser-heating. Samples were heated with the laser at each pressure increment to relax deviatoric stress. XRD measurements were carried out at 300 K using an angle-dispersive technique. Data refinements yield third-order Birch-Murnaghan equations of state parameters: the bulk modulus K 0 = 205(±8) GPa when the first pressure derivative of the bulk modulus, K 0 ', is fixed at 4 for zircon, K 0 = 392(±9) when K 0 ' is fixed at 4 for reidite. The volume reduction of the phase transition from zircon to reidite is about 10%. The tentative phase boundary between zircon and reidite is located at about 10 GPa.

A single-crystal high pressure X-ray investigation was performed up to P = 6.5 GPa on a synthetic clinopyroxene of composition Ca 0.15 (1)Mg 1.85 (1)Si 2.00 (1)O 6 [Di 15 En 85 , unit-cell parameters at room pressure: a = 9.6525(6)Å, b = 8.8461(2)Å, c = 5.2036(5)Å, β = 108.370(5)°, V = 421.68(4)Å 3 ]. A first order P21/c-C2/c displacive phase transition was found at P = 5.1 GPa; the transition was revealed by the disappearance of the b reflections (h + k = odd) and by sharp changes in the unit-cell parameters. Reversals through the transformation show that, if present, hysteresis is smaller than 0.1 GPa. The volume variation has been described by a third-order Birch-Murnaghan equation of state with V 0 = 421.68(8) Å 3 , K T0 = 102(2) GPa, and K′ = 8(1) for the low-symmetry phase (P2 1 /c) and with V 0 = 411.06(3) Å 3 and K T0 = 108(2) GPa for the high-symmetry phase (C2/c), with K′ fixed to the value obtained for the low-symmetry form. The axial compressibility shows the following scheme: β b > β a ≅ β c > β asinβ for both phases. In comparison with pure clinoenstatite, Di 15 En 85 shows a similar step in unit-cell parameters at the transition, the disappearence of hysteresis and a decrease of transition pressure and of bulk modulus. Full intensity data sets were collected at room pressure, 2.6 and 4.5 GPa for the P2 1 /c phase and at P = 6.2 GPa for the C2/c phase. A slight increase of the intensity of h + k odd reflections and of the difference in the A and B chain kinking angles were observed. A comparison of the structural behavior of the P2 1 /c phase at high temperature and high pressure shows opposite behavior for M2-O bond lengths and O3-O3-O3 kinking angle.

We have determined the room-temperature compressibility of pure, synthetic zircon (ZrSiO 4 ). Unit-cell volumes of a powdered sample were determined in situ as a function of pressure up to 27 GPa in a diamond anvil cell (DAC), by using angle-dispersive synchrotron X-ray diffraction (XRD) techniques. Unit-cell volumes were fitted to a Birch-Murnaghan equation of state, resulting in a roomtemperature bulk modulus for the zircon structure, K T0 = 199 ± 1 GPa, and ambient pressure unit-cell volume V 0 = 260.89 ± 0.03 Å 3 , when (∂K T0 /∂P) T = K' T0 is fixed at 4. This bulk modulus is over 12% lower than that suggested by earlier measurements using impure, natural zircon sample. In addition, we observed the start of the transformation of zircon to reidite (scheelite-structured ZrSiO 4 ) at a pressure of 19.7 GPa, over 3 GPa lower than previously determined for natural (impure) zircon. Together with compressibility measurements of a trace-element-doped zircon, these observations suggest that impurities affect the phase transition kinetics and compressibility of zircon, and by analogy, perhaps of other silicate minerals.

Unit-cell volumes of four single-phase intermediate halite-sylvite solid solutions have been measured to pressures and temperatures of ~28 kbar and ~700 °C. Equation-of-state fitting of the data yields thermal expansion and compressibility as a function of composition across the chloride series. The variation of the product α 0 ·K 0 is linear (ideal) in composition between the accepted values for halite and sylvite. Taken separately, the individual values of α 0 and K 0 are not linear in composition. α 0 shows a maximum near the consolute composition (X NaCl = 0.64) that exceeds the value for either end-member. There is a corresponding minimum in K 0 . The fact that the α 0 ·K 0 product is variable (and incidentally so well behaved as to be linear across the composition series) reinforces the significance of the complementary maxima and minima in α 0 and K 0 (significantly, near the consolute composition). These extrema in α 0 and K 0 provide an example of intermediate properties that do not follow simply from values for the end-members. Cell volumes across this series show small, well-behaved positive excesses, consistent with K-Na substitution causing defects through lattice mismatches. Barrett and Wallace (1954) showed maximum defect concentrations in the consolute region. Defect-riddled, weakened structures in the consolute region are more easily compressed or more easily thermally expanded, providing an explanation for our observed α 0 and K 0 variations. These compliant, loosened lattices should resist diffusive transfer less than non-defective crystals and, hence, might be expected to show higher diffusivities. Tracer diffusion rates are predicted to peak across the consolute region as exchange diffusion rates drop to zero.

Grossular (Grs 99 Adr 01 ) crystals were synthesized in hydrothermal experiments by the reaction wollastonite + calcite + anorthite = grossular + CO 2 (400 MPa; 695-750 °C; X CO₂ = 0.1; solid:fluid = 2:1) using natural minerals as starting materials. Growth spirals are visible due to silica decorations. The crystals are bounded by {110} and {211}. If crystallized at T ≤ 730 °C, they are birefringent and show sector twinning; grown at T ≥ 730 °C, they are optically isotropic. The crystal growth rates in the experiments ranged from about 1 to >100 μm/d. The edges of the growth terraces moved by 5-200 nm/s. The morphology of the growth steps is interpreted against the background of supersaturation with respect to grossular in the fluid. The highest supersaturation occurs during the initial stage of the reaction (~1 h), ultimately at the wollastonite surfaces. In this initial stage, grossular crystallizes with only {111} faces present covered by round growth spirals. When a steady state with respect to dissolution and precipitation rates is reached (after ~24 h), all grossular crystals show {110} and {211} with polygonal spirals on {110} and parallel growth steps on {211}. All straight growth-step edges parallel [11-1]. When the solid phases assume equilibrium with the fluid (after calcite in calcite-deficient experiments is used up completely), round growth spirals appear on {211}. The spirals are elongated parallel to [11-1] and compete against parallel steps in that same direction. The surface morphology of the crystals synthesized under steady state conditions mimicks that of natural grandite crystals, and their growth rates equal the estimated growth rates of grandite crystals from natural hydrothermal systems.

A suite of 18 zircon gemstones from placers in the Highland/Southwestern Complex, Sri Lanka, were subjected to a comprehensive study of their radiation damages and ages. The investigation included X-ray diffraction, Raman and PL spectroscopy, electron microprobe, PIXE and HRTEM analysis, as well as (U-Th)/He and SHRIMP U-Th-Pb age determinations. Zircon samples described in this study are virtually homogeneous. They cover the range from slightly metamict to nearly amorphous. Generally concordant U-Th-Pb ages averaging 555 ± 11 Ma were obtained. Late Ordovician zircon (U-Th)/He ages scattering around 443 ± 9 Ma correspond reasonably well with previously determined biotite Rb-Sr ages for rocks from the HSWC. Slightly to moderately metamict zircon has retained the radiogenic He whereas only strongly radiation-damaged zircon (calculated total fluences exceeding ~3.5 × 10 18 α-events/g) has experienced significant He loss. When compared to unannealed zircon from other localities, Sri Lanka zircon is about half as metamict as would correspond to complete damage accumulation over a ~555 m.y. lasting self-irradiation period, suggesting significant annealing of the structural radiation damage. Insufficient consideration of this has often resulted in significant underestimation of radiation effects in zircon. We suggest to estimate “effective α-doses” for Sri Lanka zircon by multiplying total α-fluences, which were calculated using the zircon U-Th-Pb age, by a correction factor of 0.55. This conversion may be applied to literature data as well, because all gem-zircon samples from Sri Lanka (this work and previous studies) seem to reveal the same general trends of property changes depending on the radiation damage. The use of “effective α-doses” for Sri Lanka zircon contributes to more reliable quantitative estimates of radiation effects and makes possible direct comparison between natural and synthetic radiation-damaged zircon.

Shirozulite is a new Mn-dominant trioctahedral mica from the Taguchi mine, Aichi Prefecture, Japan. The mineral occurs in tephroite-rhodochrosite ores in contact with a Ba-bearing, K-feldspar vein. Shirozulite formed during regional low-P/T metamorphism and, thereafter, suffered thermal metamorphism from a local granodiorite. Grains of shirozulite are up to 0.5 mm across and have a typical micaceous habit. Color: dark reddish brown. Cleavage: (001), perfect. Optical properties: biaxial negative, 2V x = very small. Strongly pleochroic: X = pale yellow, Y = Z = pale brown, absorption X < Y ≈ Z. Refractive indices: n α = 1.592(2), n β ≈ n γ = 1.635(2). The structural formula is (K 0.90 Ba 0.09 ) (Mn 2+ 1.53Mg 0.94 Fe 2+ 0.20 Ti 0.04 Al 0.29 ) (Si 2.54 Al 1.46 ) O 10 [(OH) 1.97 F 0.03 ], and the end-member composition is KMn 2+ 3 AlSi 3 O 10 (OH) 2 . Density: obs. = 3.20(3) g/cm 3 by pycnometer, calc. = 3.14(2) g/cm 3 . Shirozulite is monoclinic, C2/m, 1M polytype, a = 5.3791(7), b = 9.319(1), c = 10.2918(9) Å, β = 100.186(9)°, V = 507.8(1) Å 3 . The six strongest lines in the powder X-ray diffraction pattern are as follows: d (Å), l (%), (hkl): 10.16, 100, (001); 2.654, 96, (1̅31); 3.386, 51, (003); 1.556, 48, (3̅̅13); 2.467, 46, (1̅32); 2.202, 36, (1̅33). The crystal structure has been refined to an R value of 4.1% based on 663 observed reflections collected with MoKα X-radiation from a single crystal. The mean bond lengths, tetrahedral rotation, and octahedral flattening angles are as follows: <T-O> = 1.668, <M1-O> = 2.118, <M2-O> = 2.103, <K-O>(inner) = 2.995, and <K-O>(outer) = 3.376 Å, α = 8.36°, ψ M1 = 58.5°, ψ M2 = 58.2°. The apparent element distribution coefficient analyses among coexisting manganese or manganoan silicate minerals indicate that the trioctahedral mica structure cannot contain larger amounts of Mn 2+ relative to Mg and Fe 2+ than in olivine, pyroxenoid, and garnet.

Single-crystal Brillouin scattering measurements on natural orthoenstatite [OEN] to 1350 °C at 1 atm show significant softening of the elastic moduli C 33 and C 55 ahead of a phase transition. To our knowledge, these are the first observations of acoustic mode-softening in orthoenstatite at high temperature and room pressure and could have important implications for Earth's mantle. The phase transition is rapid and shows some hysteresis in the observed transition temperature, T tr . Experiments performed on increasing and decreasing temperature bracket the transition temperature between 1090(10) °C ≤ T tr ≤ 1175(10) °C, and pronounced acoustic mode-softening is evident at temperatures above 900 °C. Backscattering measurements to T = 1350 °C show no evidence for additional transitions. OEN was recovered at room temperature. Our results are interpreted in terms of elastic softening ahead of a displacive phase transition. Before the displacive transition can occur, however, the elastic softening appears to trigger the observed reconstructive transition to the more-stable protoenstatite (or high clinoenstatite) structure. We suggest that the displacive phase transition would lead to a previously unreported pyroxene structure with Cmca symmetry.

The morphological and chemical changes occurring during the thermal decomposition of weddelite, CaC 2 O 4 ·2H 2 O, have been followed in real time in a heating stage attached to an Environmental Scanning Electron Microscope operating at a pressure of 2 Torr, with a heating rate of 10 °C/min and an equilibration time of approximately 10 min. The dehydration step around 120 °C and the loss of CO around 425 °C do not involve changes in morphology, but changes in the composition were observed. The final reaction of CaCO 3 to CaO while evolving CO 2 around 600 °C involved the formation of chains of very small oxide particles pseudomorphic to the original oxalate crystals. The change in chemical composition could only be observed after cooling the sample to 350 °C because of the effects of thermal radiation.