Band 80, Heft 11-12

X-ray photoelectron spectroscopy (XPS) and solution chemistry were used to monitor the changes in surface composition of crocidolite fibers in a 50 mM NaCl solution at pH = 7.5 and 22 °C in the presence of several Fe chelators (citrate, EDTA, or desferrioxamine) for up to 30 d. The data show that the introduction of Fe chelators dramatically increases the rate at which Fe is released from the surface in comparison with a control group to which no chelators were added. In particular, XPS shows that Fe 3+ is more effectively removed in the presence of the chelators even though it is highly insoluble in aqueous solutions at near neutral pH. This suggests that Fe chelators can alter the dissolution mechanism of amphiboles from the process that dominates in NaCl solutions. This change in dissolution mechanism (particularly the enhanced rate of Fe release) is an important consideration for models of mineral-induced pathogenesis that rely on oxidation and re­duction processes. Efforts were made to estimate the Fe-release lifetimes of crocidolite fibers under the conditions of our experiments to guide the assessment of the biodurability of these fibers in human lung tissue. Our results suggest that crocidolite fibers may persist for several years, releasing Fe to lung fluids during this time. This estimated lifetime is longer than that previously estimated for chrysotile fibers and is consistent with the lifetimes previ­ously observed in asbestos mineral lung-burden studies.

Nonstandard trioctahedral serpentine polytypes are proving to be abundant, at least on a submicroscopic scale. These nonstandard polytypes have intermixing within the same crystal of either zero and ±b/3 interlayer shifts or -b/3 and +b/3 shifts in patterns other than -, +, -, +,neither of which occurs in the 12 standard polytypes. They may also have more complex ordered sequences along c of occupation of the two sets of octahedral cation positions not found in the standard polytypes. All possible nonstandard, interlayer shift sequences and octahedral cation sequences up to nine layers were generated by a computer program. Diffraction patterns were calculated for all sequences with periodicities encoun­tered in our studies’ of natural serpentines to date (regular-stacking sequences up to four layers and ordered octahedral sequences up to seven layers). Thirty-two theoretical, non­standard, regular-stacking polytype models are reported here. They consist of 14 regular interlayer stacking sequences (one two-layer, four three-layer, and nine four-layer) and four ordered octahedral cation sequences. The nine four-layer regular-stacking sequences can each exist with any of three possible ordered octahedral cation sequences along c (I,II,I,II, I,I,I,II, or I,I,II,II). Most of the nonstandard regular-stacking polytypes have triclinic geometry and C1 symmetry. Two have monoclinic Cc symmetry, and in one of these a and b are reversed. Assuming no a/3 interlayer shifts, the octahedral cation sequence present in any polytype up to seven layers can be identified by the periodicities and intensities of the 201 reflections. The regular interlayer stacking sequences up to four layers can be identified uniquely by the a angle and the intensities of specific 02/ reflections. Although it may not be known which of the pseudohexagonal a* or b* axes is present along with c* on a given diffraction pattern, identification is possible by assuming that true a* or b* is present and using prescribed identification rules. We report intensities suitable for both X-ray and electron diffraction analysis. We have determined that the distortional and compositional differ­ences likely to be encountered in natural serpentines are not great enough to obscure the diffraction characteristics that we recommend for identification.

Serpentinite from Woods Chrome Mine, Lancaster County, Pennsylvania, consists of planar serpentine, randomly interstratified serpentine-chlorite, a series of phases based on regular interstratification of serpentine and chlorite, minor chlorite, polygonal serpentine, and antigorite. The serpentine mineralogy is complex, including structures with long-range order in (1) the octahedral cation sequence and (2) the sequence of displacements between adjacent 1:1 layers. In addition to previously described (but generally less common) 2T, 2H 1 , 2H 2 , 6R 1 , and 6R 2 lizardite polytypes, the assemblage contains planar serpentines with long-range order in octahedral cation sequences but with random b/3 (and possibly a/3) displacements between adjacent layers. By comparison with calculated electron dif­fraction intensities, we identified three- (I,I,II), four- (I,I,II,II and I,I,I,II), five- (I,II,I,II,II), six- (II,I,I,I,II,II), seven- (I,II,I,II,II,II,II and I,II,I,II,I,II,II), and nine-layer octahedral se­quences. In addition, the assemblage includes rare three- and four-layer serpentines with regular stacking; in some cases the stacking is nonstandard, insofar as zero and ±b/3 displacements occur together. By comparison between electron diffraction intensities and calculated patterns, we identified a three-layer regular stacking sequence that involves 0,0,-b/3 displacements between adjacent layers (α = 98°, β = 90°, γ = 90°), a four-layer monoclinic sequence with 0,-b/3,0,+b/3 (α = 90°, β = 90°, γ = 90°; I,I,II,II), and three four-layer triclinic sequences with 0,-b/3,-b/3,-b/3 (α = 90°, β = 90°, γ = 90°), 0,0,0,-b/3 (α = 96°, β = 90°, γ = 90°), and -b/3,-b/3,-b/3,+b/3 (α = 96°, β = 90°, γ = 90°) displacements between adjacent layers. Planar layer silicates exhibit macroscopic preferred orientation. The strong lineation in the foliation defined by the silicate sheets parallels either a or b, suggesting serpentine crystals were rotated, recrystallized, or both during shear deformation. We suggest that for crystals with b parallel to the lineation, deformation induced regular layer displacements. For crystals with a parallel to the lineation, periodic displacements of OH planes may have promoted development of regular octahedral cation sequences. Two-layer (I,II) and three-layer (I,I,II) serpentines and randomly interstratified serpen­tine-chlorite contain frequent but nonperiodic planar defects perpendicular to a* and the pseudo-a* axes that are interpreted to displace polytypic sequences. These defects predate chloritization and, in some cases, appear to serve as sites for chlorite nucléation. Layer silicates are crosscut by late-stage polygonized two-layer serpentine with disordered or regular stacking. In some cases with regular stacking, enantiomorphic 6R 2 segments with a common c-axis direction parallel to their boundary are separated by sectors with 2H 2 stacking. Polygonized serpentines are nucleated on steps at the surfaces of layer silicates and may have developed at a late stage by recrystallization of curved serpentine.

The B K-edge spectra of a wide variety of minerals have been measured using the technique of parallel electron energy-loss spectroscopy (PEELS) conducted in a scanning transmission electron microscope (STEM) from sample areas of nanometer dimensions. The B K edges of the minerals exhibit electron energy-loss near-edge structure (ELNES) characteristic of B coordination. For threefold-coordinated B ( [3] B), the spectra are domi­nated by a sharp peak at ca. 194 eV because of transitions to unoccupied states of π* character, followed by a broader peak at ca. 203 eV attributed to states of σ* character. The ELNES on the B K edge (B K ELNES) of fourfold-coordinated B ( [4] B) consists of a sharp rise in intensity with a maximum at ca. 199 eV followed by several weaker structures. For [4] B, the ELNES is interpreted as transitions to states of antibonding σ* character. Minerals that possess both [3] B and [4] B exhibit an edge shape that is composed of B K edges from the respective BO 3 and BO 4 units, and we demonstrate the feasibility of quan­tification of relative site occupancies in minerals containing a mixture of B coordinations. The origins of the B K ELNES are discussed in terms of both molecular orbital (MO) and multiple scattering (MS) theory. We also present the B K-edge spectra of selected non­minerals and show how differences in edge shapes and energy onsets allow these nonmi­nerals to be readily distinguished from borates and borosilicates.

Fe 3+ -Al 3+ substitution in the octahedral site of 14 natural garnet samples was investi­gated through measurement of complete single-crystal infrared (IR) reflectance spectra across the grossular-andradite [Ca 3 (Al x Fe 1-x ) 2 Si 3 O 12 ] binary. Frequencies of all 17 IR- active fundamental modes depend nearly linearly on composition. Two peaks assigned to translations of the octahedral cations against the O framework exhibit two-mode behavior. Observation of similar frequencies for bands involving A1 and Fe 3+ and of much larger intensities for bands associated with Fe 3+ suggests that the Fe-O bond has greater strength, which compensates for the greater mass of Fe. The nearly linear behavior of the modes indicates that mixing should approach ideality for this series, as was previously observed in thermodynamic studies. The existence of ideal mixing suggests that high degrees of ordering are not present, and thus that ordering might not cause the anomalous birefrin­gence associated with this series.

A Raman spectroscopic study was performed in a diamond-anvil cell to measure the optical vibrational modes of corundum as a function of pressure up to 20 GPa. The integrals concerning the optic continuum in Kieffer’s model can be calculated as the weighted sum of each mode in the continuum. Thus, on the basis of Kieffer’s model and the experimental results of this study, thermodynamic parameters such as heat capacity, entropy, thermal expansion coefrcient, and Gruneisen parameter for corundum at various pressures and temperatures were determined. The contribution to the heat capacity of corundum approaches the classic (Dulong-Petit) limit in the interior of the Earth beneath a depth of 50 km.

The progress of ordering of Cu + and Pb 2+ in synthetic hammarite, Cu 2 Pb 2 Bi 4 S 9 , has been monitored by the form and sharpness of satellite reflections in electron diffraction patterns and by high-resolution transmission electron microscopy (HRTEM). The initial state of the synthetic hammarite is that of the parent bismuthinite cell, with a disordered distribution of Bi and Pb on the M2 sites and Cu disordered over the available tetrahedral sites with two-thirds occupancy. Strong superlattice reflections, corresponding to a super­cell of the bismuthinite parent with a 3a period, were observed in samples annealed at 17 5 and 225 °C for 24 months or longer, and HRTEM images show crystals with domains of a well-ordered 3a supercell phase. In addition to the 3a supercell, superlattices with a 5a period were also observed in some crystal fragments, indicating chemical variation within the annealed products. HRTEM was used to monitor Cu ordering and revealed that the distribution of Cu and Pb in the cell is not that of hammarite but of a lower symmetry supercell consisting of the intergrowth of one aikinite cell (two aikinite units) with two krupkaite cells (four krupkaite units). Fully ordered natural hammarite has a 3a supercell of the bismuthite parent, but with the two aikinite units and four krupkaite units ordered into a centrosymmetric structure in Pbnm. The lattice images show steps in or­dering; cations order first on a local scale to create end-member polysome strips, then the polysome strips order to form simple supercells, and finally, with longer annealing time, the Pb and Cu are reordered into the higher symmetry hammarite cell. The electron diffraction and high-resolution lattice image data suggest that the ordering of Cu + and Pb 2+ in these minerals is coupled and that Pb 2+ does not order before Cu + . The results of these annealing experiments indicate that in nature intermediate compositions in the bis- muthinite-aikinite series consist of compositionally well-ordered polysome strips, which, depending on the initial composition and the cooling rate, may remain as disordered intergrowths (stacking disorder), form ordered intergrowths, or fully order into the super­cell minerals.

Franckeite (~FeSn 3 Pb 5 Sb 2 S 14 ) is a sulfosalt mineral with a modulated structure. Its complex composition and composite layered structure, lack of a well-defined unit cell, incommensurate character, and poor crystal quality have made its structure difficult to study by standard X-ray and electron diffraction methods. Imaging by high-angle, annular, dark-field (HAADF) scanning transmission electron microscopy provides a way of map­ping compositional distributions close to an atomic scale. By comparing the structure as viewed by standard high-resolution transmission electron microscopy with the composi­tional information provided by the HAADF technique, we were able to test and confirm the assumption that layers consisting of multiple sheets of SnS 2 and PbS are stacked along the a axis with a 1.73 nm repeat.

The occurrence of sector zoning in minerals of regional metamorphic or low-temperature intrusive origin implies that rapid growth is not required for the development of this particular form of homogeneous disequilibrium. It is shown here that sector zoning can be a natural consequence simply of slow lattice diffusion. Given anisotropic surface en­richment coupled with the low dilfusivities typical of highly charged elements in refractory accessory minerals such as zircon and titanite, sector zoning can arise even in cases of growth rjates as low as a few micrometers per million years. According to the proposed model, the development of sector zoning depends upon the competition between growth rate (V) and lattice diffusion (D i ) within the near-surface layer (l), such that above a critical value of Vl/D i (~0.5-3), sector zoning is unavoidable in crystals that exhibit selective enrichment on some growth surfaces. Known diffusivities of rare earth elements in zircon and titanite lead to the expectation of sector zoning in these minerals with respect to REEs for reasonable geologic growth rates. Even in the case of clinopyroxene, diffusion of REEs and high field-strength elements may be slow enough to contribute to the development of sector zoning in laboratory-grown crystals.

A model has been developed and applied to study the origin of compositional and phase heterogeneity in magma bodies undergoing simultaneous convection and phase change. The simulator is applied to binary-component solidification of an initially superheated and homogeneous batch of magma. The model accounts for solidified, mushy (two- or three-phase), and all-liquid regions self-consistently, including latent heat effects, perco- lative flow of melt through mush, and the variation of system enthalpy with composition, temperature, and solid fraction. Phase Equilibria and thermochemical and transport data for the system KAlSi 2 O 6 -CaMgSi 2 O 6 were utilized to address the origin of compositional zonation in model peralkaline magmatic systems. Momentum transport is accomplished by Darcy percolation in solid-dominated regions and by internal viscous stress diffusion in melt-dominated regions within which relative motion between solid and melt is not allowed. Otherwise, the mixture advects as a pseudofluid with a viscosity that depends on the local crystallinity. Energy conservation is written as a mixture-enthalpy equation with subsidiary expressions that are based on thermochemical data and phase relations that relate the mixture enthalpy to temperature, composition, and phase abundances at each location. Species conservation is written as the low-density component (KAlSi 2 O 6 ) and allows for advection and diffusion as well as the relative motion between solid and melt. Systematic simulations were performed to assess the role of thermal boundary conditions, solidification rates, and magma-body shape on the crystallization history. Examination of animations showing the spatial development of the bulk (mixture) composition (C) melt composition (C 1 ), temperature (T), solid fraction (f s ), mixture enthalpy (h), and velocity (V), reveals the unsteady and complex nature of convective solidification owing to nonlinear coupling among the momentum, energy, and species conservation equations. A consequence of the coupling includes the spontaneous development of compositional heterogeneity in terms of the modal abundances of diopside and leucite in the all-solid parts of the domain (i.e., modal mineralógical heterogeneity) as well as spatial variations in melt composition particularly within mushy regions where phase relations strongly couple compositional and thermal fields. Temporal changes in the rate of heat extraction because of bursts of crystallization and concommitant buoyancy generation are also found. Crystallization of diopside, the liquidus phase in all cases, enriches residual melt in low- density K-rich liquid. The upward flow of this material near the mush-liquid interface leads to the development of a strong vertical compositional gradient. The main effect of magma-body shape and different thermal boundary conditions is in changing the rate of solidification; in all cases compositional heterogeneities develop. The rate of formation of the compositional stratification is highest for the sill-like body because of its high cooling rate. Compositional zonation in a fully solidified body is found to be both radial and vertical. The most salient feature of this simple model is the spontaneous development of large-scale magma heterogeneity from homogeneous and slightly superheated initial states, assuming local equilibrium prevails during the course of phase change.

Igneous cumulate rocks provide an important record of planetary magmatism, but there are pitfalls in their interpretation. The rocks are composed of cumulus minerals plus assemblages that crystallized from trapped melt. Cumulus minerals may react with the trapped melt and other cumulus phases during subsolidus reactions, thus losing a direct record of their igneous history. One of the best approaches for estimating the melt com­positions parental to the cumulates is to analyze the cores of cumulus phases for elements with slow diffusion rates (e.g., REE) because these most reliably retain a record of the mineral-melt partitioning. Many of the cumulus orthopyroxene grains from the Stillwater Complex analyzed in this study have augite lamellae in their interiors but lamellae-free rims. Secondary ion mass spectrometer (SIMS) analysis shows that these orthopyroxene rims have lower Sr, Y, Zr, and Ce concentrations relative to the cores. These systematics were apparently caused by migration of augite exsolution lamellae, which preferentially incorporate these trace elements, out of the rims of the orthopyroxene grains. SIMS analyses for REE in orthopyroxene cores show lower LREE concentrations, (Ce/ Yb) n and (Dy/Yb) n than the isotope dilution data on mineral separates (Lambert and Simmons, 1987). We believe that these differences result from one or more contaminating phases (e.g., augite and plagioclase) in the orthopyroxene mineral separates. The SIMS data yield calculated parental melts similar to sills found below the Stillwater Complex, which have previously been suggested as possible parental melts (Helz, 1985). A major focus of this study was to determine whether the onset of plagioclase crystal­lization at the bronzitite zone-norite zone contact was the result of fractionation until plagioclase saturation of the magmas parental to the bronzitites, or if mixing of a second magma with higher Al activity was the cause of plagioclase crystallization. The data pre­sented in this paper, coupled with existing data, support the magma-mixing model.

H 2 O-saturated experiments with synthetic metapelite compositions (muscovite-quartz- albite and muscovite-quartz-albite-biotite ± aluminum silicate ± cordierite) performed over the temperature interval of 600-750 °C at 200 MPa (H 2 O) reveal that partial fusion commences at 625 °C along the metastable extension of the reaction muscovite + quartz + albite + H 2 O = melt + aluminum silicate. Biotite is stable over the entire temperature interval, although it reacts progressively to hercynite + melt at the high end of the tem­perature range. Muscovite that survives initial melting breaks down to corundum + or- thoclase between 700 and 725 °C. Minor corundum and aluminum silicate are present at 650 °C, whereas corundum with a large sapphire (Fe + Ti) component is present at and above 700 °C. Finally, corundum and hercynite exist with orthoclase-rich feldspar and remaining biotite at 750 °C. The normative composition of melt at the minimum is ap­proximately Ab 30 Or 15 Ms 20 Qtz 35 (in weight percent) without other minor components (e.g., cryolite). Both muscovite and its equivalent orthoclase + corundum assemblage contribute substantial excess Al to melt, bringing the value of the Al saturation index (ASI) of melt to 1.4. Li, Rb, Cs, and F are strongly enriched in melts because of the continuous reaction and reequilibration of biotite and muscovite over the temperature interval. Concentrations of the femic components (TiO 2 + FeO + MgO + MnO) are low (< 1 wt%) but rise with temperature primarily because of increasing solubilities of FeO in melt. Calculated parti­tion coefficients, D(M) Bt/gl , between biotite (Bt) and glass (gl) for the element M show that D(Li) Bt/gl (1.7-1.0), D(Ba) Bl/gl (~ 14-6), and D(F) Bt/gl (2.5-1.5) all decrease with increasing temperature, whereas partition coefficients for Sr (≈0.04), Rb (≈2.0), and Cs (≈0.4) re­main constant with temperature over a large range of concentrations. Partition coefficients for muscovite (Ms) were also determined at 650 °C; D(M) Ms/gl ≈ 0.8 (Li), 3-6 (Ba), 1.6 (Rb), 0.05 (Sr), 0.3 (Cs), and 1.8 (F). These results, together with other data for feldspar, suggest that Rb, Cs, and Ba become strongly fractionated from one another during anatexis of aluminous metasediments and the ensuing crystallization of melts. Finally, the low partition coefficients for F between micas (biotite or muscovite) and melt indicate that F-rich melts can be generated by the incipient hydrous anatexis of aluminous metasedi­ments; models that invoke remelting of a dehydrated protolith to generate such F-rich melts may not be necessary.

The equilibrium P-T curve of the reaction MgCO 3 + SiO 2 = MgSiO 3 + CO 2 (1) magnesite quartz enstatite was determined by reversed reactions in well-calibrated, solid-pressure medium (NaCl) piston-cylinder apparatus. External hematite + magnetite + H 2 O f O₂ buffers were used in all experiments. A reversed point at 715 °C of the reaction MgCO 3 = MgO + CO 2 (2) magnesite periclase was obtained in an internally heated Ar pressure vessel, using a coarse-grained synthetic periclase to inhibit quenching back reaction, along with synthetic magnesite and silver oxalate. At 715 °C our pressure for Reaction 1 is 10.4 ±0.2 kbar, and for Reaction 2 it is 590 ± 30 bars. From these two reactions we derive ΔH 0 f,298 = 33.41 ± 0.83 kJ for enstatite from the oxides. Using an extrapolated heat-capacity equation for magnesite, we obtain ΔH 0 f,298 = -116.68 kJ (from the oxides) and -1111.68 kJ from the elements for magnesite, somewhat less negative than the values of either the Holland and Powell (1990) or the Berman (1988) data sets. A calculated 2 kbar temperature for the magnesite + quartz reaction is 492 °C, about 25 °C lower than the data set predictions but in agreement with a prediction based on natural occurrences of magnesite, quartz, and enstatite in metacarbonates from the Alps by Trommsdorif and Connolly (1990).

The behavior of Ca and U in the system CaO-UO 2 -La 2 O 3 -H 3 PO 4 has been experimen­tally investigated at 780 °C and 200 MPa and at f O₂ conditions fixed by the Ni + NiO buffer. The results obtained indicate that a complete solid solution was formed between LaPO 4 and (Ca 0.5 U 0.5 )PO 4 . The coupled substitution of two atoms of La 3+ by U 4+ + Ca 2+ is represented by the equation LaPO 4 + xU 4+ + xCa 2+ = (La 1-2x U x Ca x )PO 4 + 2xLa 3+ with the mole fraction x varying from 0 to 0.5. The cell parameters of the solid solution (La 1-2x U x Ca x )PO 4 decrease linearly with decreasing amounts of La 3+ . This is a result of the smaller ionic radii of U 4+ and Ca 2+ (1.05 and 1.18 Å, respectively) compared with that of La 3+ (1.216 Å). The end-member (Cao. 5 U 05 )PO 4 was hydrothermally synthesized for the first time in the system CaO-UO 2 -H 3 PO 4 under similar experimental conditions. The data obtained show that there is no crystal-chemical limitation to U substitution in monazite crystals at temperature and pressure conditions corresponding to those of granitic magma crystallization, provided that Ca is available.

Preliminary phase relations for the bulk composition Na 2 O·CaO·5SiO 2 were experi­mentally determined at 8-16 GPa and 950-2300 °C, using a split-sphere anvil apparatus (USSA-2000). The dominant phase at pressures between 8 and 14 GPa was a new com­pound with the ideal formula Na 2 CaSi 6 O 14 ; microprobe analyses suggest vacancies that cannot be resolved by X-ray data, resulting in a composition better expressed as Na 1.8 Ca 1.1 Si 6 O I4 . Single crystals suitable for structure determination were synthesized at 14 GPa and 1900 °C. Two independent sets of X-ray diffraction data were collected on different crystals from the same batch: one with a conventional diffractometer and the other using imaging plates and a synchrotron X-ray source. The compound crystallizes in space group P321 with a = 7.903(2) and c = 4.595(1) Å. Its structure consists of layers in (001); one is centered at z = 0 and contains the (Ca,Na) and [6] Si sites. The other, centered at z = ½, is a sheet of comer-linked [4] Si tetrahedra composed of open-branched crankshaft vierer chains linked by single tetrahedra. Although the SiO 6 octahedra are close to regular, with Si-O = 1.789(2) Å, the two symmetry-independent SiO 4 tetrahedra are distorted as a result of underbonding between O atoms bonding outside the sheet, with Si 1-O = 1.557(2) and 1.638(2) Å, and Si2-O = 1.619(1) and 1.648(2) Å. Half-normal probability plots show the positional parameters derived from the two data sets to be statistically equivalent.

The structure of lawsonite, CaAl 2 (Si 2 O 7 )(OH) 2 · H 2 O, was refined from X-ray single­crystal data at six temperatures between 110 and 500 K. Two reversible phase transitions were found to occur: (1) Below 273(5) K the space group changes from Cmcm to Pmcn, accompanied by the loss of the --m mirror plane because of rotation of both H 2 O and OH groups. This transition was monitored by the appearance of C-centering forbidden reflec­tions (h + k = 2n + 1) and by nonlinear changes of cell constants and optical birefringence values vs. temperature. (2) Below 155(5) K the space group symmetry is further reduced to P2 1 cn, accompanied by the loss of the m-- mirror plane because of unidirectional shifts of the x coordinates of the H atoms of the H 2 O and OH groups, thus resulting in a polar structure. The driving force of both transitions is apparently the development of cooperative hydrogen bonds. At elevated temperatures the OH and H 2 O groups are constrained by strong librations to occupy sites with high point symmetry. At low temperatures and reduced thermal vibration the H atoms are allowed to shift towards neighboring O atoms, building up strong hydrogen bonds. Consequently, the H 2 O and OH groups are shifted off their highly symmetric positions in an ordered fashion, thus lowering the space group symmetry in two steps. The present structural data allow reinterpretation of previously published, low-temperature, infrared-spectroscopic and thermodynamic measurements on lawsonite.

The high-pressure phase relationships in a H 2 O-saturated synthetic CaO-Al 2 O 3 -SiO 2 - H 2 O (CASH) system were studied by multi-anvil experiments. The most extreme pressure condition under which pure lawsonite [CaAl 2 Si 2 O 7 (OH) 2 H 2 O] exists is 120 kbar at 960 °C. This maximum stability is located at the intersection of the lawsonite breakdown reaction with the topaz-OH + stishovite = phase “egg” reaction. At lower pressures and higher temperatures lawsonite decomposes to grossular + topaz-OH + stishovite + H 2 O, whereas at lower pressures and lower temperatures lawsonite first reacts to grossular + phase egg + topaz-OH + H 2 O and at still lower temperatures to grossular + phase egg + diaspore + H 2 O. The two latter reactions have positive dP/dT slopes, with lawsonite on the low-pressure side, and thus delimit the occurrence of lawsonite toward higher pres­sures. The occurrence of topaz-OH (10.7 wt% H 2 O) is limited through a reaction to phase egg (7.5 wt% H 2 O) + diaspore; the phase boundary extends from 110 kbar and 720 °C to 130 kbar and 920 °C. Phase egg is inferred to have a composition of AlSiO 3 (0H) and a monoclinic unit cell similar to that proposed by Eggleton et al. (1978). The high-pressure breakdown of lawsonite in CASH does not result in an anhydrous assemblage. Lawsonite is known to occur experimentally in basaltic and andesitic com­positions to at least 77 kbar; however, it is unknown whether topaz-OH and phase egg appear in natural multicomponent systems, in particular those saturated in kyanite and H 2 O.

To explore the transformation behavior of metastable olivine at large overpressures equivalent to 500-670 km depth, a reconnaissance study using Mg 2 GeO 4 olivine at pres­sures up to 16 GPa was conducted. Optical microscopy and transmission electron mi­croscopy of the specimens were used to observe the behavior of both the reconstructive and martensitic-like transformation mechanisms. The experiments revealed that the ki­netics of the thermally activated reconstructive transformation are insensitive to pressure; comparison with previous growth rate measurements made at 1-2 GPa lead to an estimate of the activation volume for growjh of V* = 0 ± 2 cm 3 /mol. In contrast, the martensitic- like mechanism becomes more important at high pressure, producing optically visible features in Mg 2 GeO 4 olivine by 9 GPa. Similar conditions would be experienced by natural olivine that has metastably persisted in cold subducting slabs to the bottom of the tran­sition zone, these observations suggest that the conversion of natural olivine to its spinel structure by the martensitic-like mechanism could become significant in that environment. This conversion would provide a uniform final cut-off depth for earthquakes caused by transformational faulting independent of slab thermal structure.

For H 2 O densities > 1.0 g/cm 3 , a determination of the ice melting temperature provides the density information required to calculate the P-T path that the sample in a hydro- thermal diamond-anvil cell follows when the sample is heated isochorically. The principal difficulty is the identification of the polymorph because of metastable behavior of ices in the H 2 O system. Usually, an accurate identification of the liquidus ice phase can be made without analytical instrumentation and requires only careful observations.

Two types of chromite deposits occur in the Hongguleleng ophiolite in Xinjiang, north­west China. One is iocated in the mantle sequence, the other occurs in the transition zone between the mantle sequence and layered cumulates. Abundant primary silicate inclusions such as phlogopite, pargasite, clinopyroxene, orthopyroxene, and olivine are found in the segregated chromite, but silicate inclusions occur only rarely in accessory chromite of the ultramafic rocks from the transition zone and cumulates. These silicate phases are consid­ered to have been entrapped as discrete and rare composite inclusions during magmatic precipitation of chromite rather than formed by postmagmatic entrapment. Phlogopite is the most abundant mineral found as inclusions in chromite in the Hong­guleleng ophiolite. There are two types of substitutions for K in the phlogopite inclusions: (1) Na substitutes for K, and phlogopite shows a continuous range from almost pure sodium phlogopite to phlogopite; (2) Ca partially substitutes for K., resulting in the for­mation of Ca-bearing phlogopite. It is proposed that the alkalic aqueous liquid (melt) responsible for the formation of the phlogopite inclusions was derived from the mixture of the K-rich aqueous liquid related to the subduction of the oceanic slab and the Na-rich aqueous liquid from the primary magma of the ophiolite. Two types of phlogopite hydrates, hydrate I and possibly a new hydrate (hydrate H), occur as inclusions in the chromite and result from later hydrothermal processes. The fractures from brittle deformation provided passage for meteoric water to enter and react with the phlogopite inclusions. Compared with those in the transition zone, the inclusions of phlogopite and phlogopite hydrates in the mantle sequence are characterized by (1) smaller grain size and greater abundance, (2) undulatory extinction, (3) higher Si, Cr, Ni, and Ca, and (4) lower Ti and Al. These differences are possibly due to (1) P, T, and composition of chromite-precipi­tating magma, (2) subsolidus reequilibration with the host chromite, and (3) postmagmatic deformation and hydrothermal processes.

The Zr-bearing minerals baddeleyite (ZrO 3 ) and zircon (ZrSiO 4 ) occur within plagioclase-rich (61-95% plagioclase) cumulates of the Laramie anorthosite complex (LAC), southeastern Wyoming. In each of the examined samples, zircon is present as relatively coarse (1-2 mm) interstitial grains, and baddeleyite occurs as small (0.05 mm) inclusions within cumulus plagioclase. Zircon crystallized between cumulus plagioclase crystals near solidus temperatures from highly fractionated, Zr-saturated liquids. The resultant shape of zircon was controlled by the form of the remaining pore space. The origin of baddeleyite in the anorthositic rocks of the LAC is less well constrained. It may have crystallized early from the anorthositic parental magmas at relatively low silica activities; however, this would require baddeleyite saturation at extremely low Zr concentrations in the parental magmas (≪ 100 ppm). Baddeleyite and zircon U-Pb ages reveal that several petrologically distinct intrusions were emplaced and crystallized in the LAC over a relatively restricted 1-3 m.y. interval at ca. 1434 Ma. The 2O7 Pb/ 206 Pb ages obtained for the baddeleyite and zircon in each sample are identical within error (±1-3 m.y.), and U concentrations are uniformly low (<240 ppm), supporting a genetically related origin for the minerals. Two anorthositic layered cumulates and a crosscutting, oxide-rich troctolite from the Poe Mountain anorthosite have crystallization ages that are identical within error: 1434.4 ± 0.6, 1434.5 ± 0.6, and 1434.1 ±0.7 Ma, respectively. The only earlier period of anorthositic magmatism that can be identified from the Poe Mountain anorthosite is represented by a leucogabbroic xenolith (1436.2 ± 0.6 Ma), which settled onto the floor of the magma chamber that produced the layered cumulates. A sample composed almost entirely of zoned, iridescent, plagioclase megacrysts from the Chugwater anorthosite yields a baddeleyite age of 1435.4 ±0.5 Ma, intermediate between the ages of the xenolith and the layered anorthositic rocks. All the ca. 1.4-1.5 Ga anorthosites in North America, including the LAC, are located near or on Paleoproterozoic boundaries between Archean cratons and accreted Proterozoic island arc terranes. These preexisting crustal structures appear to play a major role in the origin and ascent of anorthositic magmas. The evidence for anorthositic magmatism at 1.43 Ga in the LAC suggests that there may be a strong genetic link between these high- temperature mafic magmas and the regional production of anorogenic granites in the western and southwestern U.S., many of which have similar crystallization ages in the interval of 1.43-1.44 Ga.

Near-infrared spectra of hydrous silicate melts are reported up to 10 kbar and 1000 °C. Measurements were performed using an externally heated diamond-anvil cell that allows precise control of pressure and temperature. In hydrous glasses, changes in H 2 O speciation occur below the glass-transformation temperature; however, the standard enthalpy of the speciation reaction H 2 O + O = 2OH is very different for the melt and glass phases. For a peraluminous sodium aluminosilicate glass containing 8.1 wt% total H 2 O, the temper­ature dependence of the H z O speciation equilibrium can be described by two equations: In K = -2.65 - 1.91 × 10 2 K T ₋1 (glass phase, ΔH = 1.6 kJ/mol) and In K = 3.04 - 3.64 × 10 3 K T ₋1 (melt phase, ΔH = 30.3 kJ/mol). The intersection of these two equilib­rium curves defines a glass-transformation temperature of 335 °C. The near-infrared spec­trum of a peraluminous sodium silicate melt with 30-40 wt% H 2 O at 10 kbar and 1000 °C shows the presence of both OH and H 2 O; however, exceptionally strong speciation changes occur upon quenching. From visual observation, it appears that at 10 kbar and 1000 °C there is complete miscibility between this melt and water. Therefore, these data constitute the first infrared spectra of a phase beyond the critical point in a system of silicate melt and H 2 O.

Basanitic glasses with known dissolved C concentrations have been analyzed using in­frared spectroscopy to calibrate the molar absorptivity of carbonate. C is dissolved as carbonate complexes in the glass resulting in absorption bands at 1525 and 1425 cm ₋1 . Molar absorptivities of 283 ± 8 L/(mol·cm) were determined for both 1525 and 1425 cm ₋1 absorption bands. An integrated molar absorptivity of 60000 ± 1700 L/(mol·cm 2 ) was determined using the integrated area under the doublet. These values are about 20- 25% lower than those determined for tholeiitic and leucititic glasses and are intermediate between values characteristic of Na- and Ca-rich silicate glasses. Carbonate molar absorp­tivities for a range of basaltic glass compositions correlate with the molar ratio of Na/(Na + Ca).

Five magma types occur as sills in the footwall of the Stillwater Complex, all of them coeval with the Stillwater in age. Two of the magma types have compositions that suggest they are similar to the magmas from which the cumulates of the Stillwater’s Basal and Ultramafic series crystallized. Melting experiments performed on samples of these two magma types and on a 50-50 mix of the two types show that the crystallization sequence inferred from the Ultramafic series cumulates (olivine > orthopyroxene > plagioclase > augite) is matched only in the mixture of the two sill types at 1.5-3 kbar. Thus, the observed stratigraphy of the Basal and Ultramafic series may result from mixing of two distinct magmas rather than fractionation of a single magma. The permissible pressure range of 1.5-3 kbar implies that the depth of the Stillwater magma chamber was similar to that of subvolcanic magma chambers such as that of Kilauea Volcano, which raises the possibility that the Stillwater magma body may also have been a subvolcanic reservoir.

Experiments conducted on a natural kaersutitic amphibole, combined with statistical analysis of high-quality chemical analyses reported in the literature, document that both Fe 3+ and Ti in this group of Ti-rich hornblendes are accommodated by the oxy substitu­tions Fe 3+ O 2- Fe 2+ -1 (OH) ₋ -1 , and Ti 4+ O 2- Ri n+ -1 (OH) ₋ -1 .The valence of the cation R n+ -1 can be variable, provided that the sum of all substitutions maintains charge balance within the amphibole crystal structure. If the OH lost as a result of the two oxy substitutions is accounted for, the mean amphibole O3 site occupancies (i.e., OH ₋ + F ₋ + Cl ₋ + O 2- ) of both the experimentally treated and naturally occurring sample populations sum to the theoretical value of 2.00, within the statistical uncertainty. The unusually low H contents reported for kaersutite from Martian meteorites are con­cluded to result from their unusually high Ti and total Fe contents relative to terrestrial kaersutite.