Volume 104, Issue 3

High-angle annular dark-field scanning transmission electron microscopy is a powerful Z -contrast technique able to depict the structural motifs in Pb-(Bi-Sb)-sulfosalts. Using two homologs from the kobellite homologous series, a group of “chessboard derivative structures,” represented by Bi-, and Sb-rich pairs of natural phases (the kobellite-tintinaite isotypic series and giessenite-izoklakeite homeotypic series), we visualize the slabs underpinning crystal structural modularity for the N = 2 homolog kobellite and the N = 4 homolog, in this case a Bi-rich izoklakeite [Sb/(Sb+Bi) = 0.35]. The homolog number, N , can be readily calculated as N = n 1 /6 – 1 and N = n 2 /4, where n 1 and n 2 are the numbers of atoms in the PbS- and SnS-motifs, respectively. Atom-scale imaging of thinned foils extracted in situ from samples for which compositional data are available also reveals syntactic unit-cell scale intergrowths on [001] zone axis with a kobeiiite ∥ b izoidakefa . These are as small as half-unit cells of b izokiakeite and one-unit cell a kobellite . Replacement relationships are also observed as irregular slabs of kobellite “intruding” into izoklakeite. Both banded and irregular intergrowths account for the compositional fields measured at the micrometer scale.

Chloroacetanilides and their degradation products are frequently detected in surface and subsurface water due to their relatively high water solubility and their high potential to leach and migrate through the soil and contaminate groundwater. In this study, we explored for the first time the capability of ZSM-12 zeolite for 2-ethyl-6-methyl-aniline [C 2 H 5 C 6 H 3 (CH 3 )NH 2 , labeled EMA] removal from water by combining chromatographic, thermogravimetric, and synchrotron X-ray powder diffractometric techniques. Rietveld refinement revealed the incorporation of about 4 EMA molecules per unit cell, in very good agreement with the weight loss given by TG analyses and with the saturation capacity determined by the adsorption isotherms. The formation of supramolecular complexes mediated by co-adsorbed water and their strong interaction to framework O atoms confers stability to the pollutants in the zeolite cages. This prevents adsorbed molecules from desorbing as well as the entering of other competitive molecules. The rapid kinetics combined with the good adsorption capacity makes ZSM-12 a promising material to control and minimize water pollution from acetanilide compounds as well as other agro-chemicals contaminants.

The Eckstrom-Adcock iron carbide, nominally Fe 7 C 3 , is a potential host of reduced carbon in Earth’s mantle and a candidate component of the inner core. Non-stoichiometry in Fe 7 C 3 has been observed previously, but the crystal chemistry basis for its origin and influences on the physical properties were not known. Here we report chemical and structural analyses of synthetic Fe 7 C 3 that was grown through a diffusive reaction between iron and graphite and contained 31 to 35 at% carbon. We found that more carbon-rich Fe 7 C 3 has smaller unit-cell volume, suggesting that excess carbon atoms substituted for iron atoms instead of entering the interstitial sites of closed-packed iron lattice as in FeC x steel. Carbon may be the lightest alloying element to substitute for iron. The substitution leads to a larger reduction in the unit-cell mass than the volume so that the carbon-rich end-member may be as much as 5% less dense than stoichiometric Fe 7 C 3 . If Fe 7 C 3 solidifies from Earth’s iron-rich liquid core, it is expected to have a nearly stoichiometric composition with a compositional expansion coefficient of ~1.0. However, laboratory experiments using carbon-rich synthetic Fe 7 C 3 to model the inner core may overestimate the amount of carbon that is needed to account for the core density deficit.

Small amounts of H 2 O, on the order of tens to hundreds of parts per million, can significantly influence the physical properties of mantle rocks. Determining the H 2 O contents of nominally anhydrous minerals (NAMs) is one relatively common technique that has been applied to estimate mantle H 2 O contents. However, for many mantle NAMs, the relation between H 2 O activity and H 2 O content is not well known. Furthermore, certain mantle minerals may be prone to H 2 O loss during emplacement on Earth’s surface. The goal of this study is to apply mineral equilibria to estimate values of a H 2 O in rocks that originated below the Moho. The chemical compositions of olivine + orthopyroxene + clinopyroxene + amphibole + spinel ± garnet were used to estimate values of temperature ( T ), pressure ( P ), a H 2 O , hydrogen fugacity ( f H 2 ), and oxygen fugacity ( f O 2 ) in 11 amphibole-bearing mantle xenoliths from the southwestern U.S.A. Application of amphibole dehydration equilibria yields values of a H 2 O ranging from 0.05 to 0.26 for these 11 samples and the compositions of coexisting spinel + olivine + orthopyroxene yield Δlog f O 2 (FMQ) of -1 to +0.6. For nine of the samples, values of f H 2 were estimated using amphibole dehydrogenation equilibria, and these values of f H 2 ranged from 6 to 91 bars. Values of f H 2 and f O 2 were combined, using the relation 2H 2 O = 2H 2 + O 2 , to estimate a second value of a H 2 O that ranged from 0.01 to 0.57 for these nine samples. Values of a H 2 O , estimated using these two methods on the same sample, generally agree to within 0.05. This agreement indicates that the amphibole in these samples has experienced little or no retrograde H-loss and that amphibole equilibria yields robust estimates of a H 2 O that, in these xenoliths, are generally <0.3, and are often 0.1 or less.

The chevkinite-group minerals (CGM) are dominantly monoclinic REE-Ti-Fe sorosilicates [(REE, Ca) 4 Fe 2+ (Fe 2+ ,Fe 3+ ,Ti) 2 Ti 2 (Si 2 O 7 ) 2 O 8 )], with REE 2 O 3 contents up to ~50 wt%, but members with predominant Mg, Al, Mn, Cr, Sr, or Zr in one of the cation sites are also known. Twelve members have been approved by the Commission of New Minerals, Nomenclature and Classification International Mineralogical Association (CNMNC-IMA) but more will undoubtedly be identified. Minerals of the group are known from hundreds of terrestrial localities and have also been recorded in lunar and martian rocks. The main occurrences are in igneous rocks ranging from diamond in kimberlites through mafic and intermediate lithologies to metaluminous and peralkaline felsic rocks. They also occur in metamorphic rocks, including granulites, metacarbonates, and jadeitites, and in metasomatic rocks, such as skarns and fenites, and in rare-metal deposits. Chevkinite-group minerals may form over the pressure range 50 to <10 kbar, and over a wide temperature range. Their formation appears to be relatively insensitive to P H 2 O and f O 2 . The stability of CGM vis-à-vis other REE-Ti-bearing accessories is poorly understood. They are often the major carriers of REE and actinides, and they have a high potential for fractionating the light lanthanides and Th from U. Very little systematic work has been done in determining CGM-melt partition coefficients, yet such data are critical in, inter alia , geochemical modeling. Similarly, CGM are amenable to geochronology due to their high Th abundances, commonly at the several percent level. In common with other REE-bearing accessories, CGM are prone to alteration by hydrothermal fluids. The nature and extent of the alteration are primarily determined by the composition of the fluids. Fluids poor in ligands tend to generate a Ti-enriched phase whose nature is unknown but is probably amorphous. With increasing F + CO 2 levels, complex replacement assemblages are formed, usually in more than one step. Although observational evidence of the effects of alteration and element mobility is accumulating and chemical equations can be constructed to approximate the reactions, there is still no firm geochemical basis for understanding element redistribution during these processes.

Magnesium aluminate spinel, (Mg,Fe)Al 2 O 4 , is uncommon in lunar rocks but petrologically significant. Recent near-infrared spectra of the Moon have delineated regions where spinel is the only ferromagnesian mineral; the rock is inferred to be spinel anorthosite. One hypothesis is that significant pressure is required for spinel formation; another is that spinel-bearing rocks form by low-pressure assimilation of highlands anorthosite into olivine-rich basaltic (i.e., picritic) magmas. Here, we evaluate the heat (i.e., enthalpy) required for this assimilation process. Magma compositions are the picritic Apollo 14 B green glass and an estimation of the magma parental to Mg-suite cumulate rocks. From calculated enthalpy-composition phase diagrams, assimilation of anorthite into either magma cannot produce spinel anorthosite unless the anorthite is already hotter than ~1300 °C. For cooler anorthosite, assimilation will produce olivine- and/or pyroxene-bearing rocks. Such hot anorthite could be produced by the nearby passage of large volumes of magma, but this is not obviously consistent with occurrences of spinel far from outcrops of basaltic rocks. Hot anorthosite could also be produced by global tidal flexure; that mechanism could have only been efficient early in lunar history when a solid anorthosite crust floated above an evolved magma ocean, and it is not clear how picritic magma could pass through the magma ocean to interact with anorthite in the crust. On the other hand, spinel-bearing anorthosite can form directly upon cooling of superliquidus melts of anorthite-rich composition. Such superliquidus melts can be generated by impact events; this mechanism seems likely, given the Moon’s ubiquity of impact craters, abundance of impact-metamorphosed lunar rocks, and common presence in lunar regolith of impact glasses (quenched superliquidus impact melts) of appropriate compositions. High pressure does stabilize spinel in basaltic and peridotitic systems, but available models do not permit quantitative evaluation of the effects of pressure on the enthalpy required for assimilation. Near the lunar surface, the most likely process of spinel formation is rapid crystallization of impact melts of anorthosite + picrite or peridotite compositions. The presence of spinel anorthosite on the walls and central peaks of impact craters results from rapid cooling and partial crystallization of superliquidus melts produced in the impacts, and not from uplift of deep material to the Moon’s surface.

Diffusion experiments of 2 H 2 O at 900–750 °C and water vapor pressure of 50 bar found diffusion of water in SiO 2 glass more than one order of magnitude faster than that reported previously. The fast diffusion profile of water was observed as an extended tail of the normal water diffusion profile by a line scan analysis with SIMS, and it can be fitted with a diffusion model with a constant diffusivity. The obtained fast diffusion coefficient suggests that the diffusion species responsible for the fast diffusion is not molecular hydrogen but molecular water. The diffusivity and activation energy for the fast water diffusion can be explained by the correlation between diffusivities of noble gases in silica glass and their sizes. Because noble gases diffuse through free volume in the glass structure, we conclude that molecular water can also diffuse through the free volume. The abundance of free volume in the silica glass structure estimated previously is higher than that of 2 H observed in the fast diffusion in this study, suggesting that the free volume was not fully occupied by 2 H 2 O under the present experimental condition. This implies that the contribution of the fast water diffusion to the total water transport in volcanic glass becomes larger under higher water vapor pressure conditions.

Textural and compositional microscale (10–100 μm) and nanoscale (10–100 nm) zoning in a plagioclase phenocryst from a fresh, syn-mineralization diorite porphyry (Black Mountain porphyry Cu-Au deposit, Philippines) was characterized for major and trace elements using electron microprobe, laser ablation-inductively coupled plasma-mass spectrometry, and atom probe tomography. The complex plagioclase crystal (3.0 × 5.4 mm) has a patchy andesine core (An 41–48 mol%), eroded bytownite mantle (An 71–80 mol%), and oscillatory andesine rim (An 39–51 mol%). Microscale variations with a periodic width of 50 to 200 μm were noted for most major and trace elements (Si, Ca, Al, Na, K, Fe, Mg, Ti, Sr, Ba, Pb, La, Ce, and Pr) with a ΔAn amplitude of 4–12 mol% in both the core and rim. The mantle has a distinct elemental composition, indicating the addition of hotter mafic magma to the andesitic magma. Atom probe tomography shows an absence of nanoscale variations in the andesine rim but alternating nanoscale (25–30 nm) Al-rich, Ca-rich, and Si-rich, Na-rich zones with a Ca/(Ca+Na) at% amplitude of ~10 in the bytownite mantle. The restricted variations in physiochemical parameters (H 2 O-rich, T = 865 to 895 °C, P = 5.3 to 6.2 kbar; f O 2 = NNO+0.6 to NNO+1.1 recorded by co-precipitated amphibole) suggest microscale oscillatory zoning was likely controlled by internal crystal growth mechanisms, and not by periodic variations in physiochemical conditions. However, the uniform diffusion timescale for CaAl-NaSi interdiffusion in the mantle is far shorter than the crystallization timescale of the grain from mantle to rim, suggesting nanoscale zonation in the bytownite mantle formed by exsolution after crystallization. The occurrence of micro-scale zoning in plagioclase indicates a minimum cooling rate of 0.0005 °C/yr during crystallization, assuming an initial temperature of 880 °C, the width of 50 μm, and NaSi-CaAl interdiffusion under hydrous conditions. Assuming a formation temperature of ~675 °C for the nanoscale exsolution texture as constrained by zircon crystallization temperatures, the retention of nanoscale zoning (~28 nm) requires a minimum cooling rate of 0.26 °C/yr. Given that this is significantly faster cooling than would occur in a magma chamber, this texture likely records the post-crystallization emplacement history.

Chlorite is a ubiquitous product of metamorphism, alteration of magmatic rocks and hydrothermal processes owing to its large stability field and wide compositional range. Its composition is governed by several substitutions and has been used as a geothermometer, on the basis of empirical, semi-empirical, and thermodynamic models. As in some other phyllosilicates of petrological interest, the oxidation state of iron in chlorite may differ from the usually assumed divalent state. However, the crystal chemistry of trivalent iron in chlorite remains poorly known, and the thermodynamic properties of ferric chlorite are missing from databases used for petrological modeling. As part of an attempt to fill this gap, we present results from in situ, micrometer-scale measurements of the oxidation state of iron in various chlorite-bearing samples. X-ray absorption near-edge spectroscopy (XANES) was combined with electron probe microanalysis (EPMA) on the same crystals. Results show iron oxidation states varying from ferrous to ferric; iron is in octahedral coordination in all ferromagnesian chlorites but to ~25% tetrahedral in the lithian chlorite cookeite (1.0 wt% Fe 2 O 3(total) ). Absolute amounts of ferric iron cover an unprecedented range (0 to ~30 wt% Fe 2 O 3 ). For highly magnesian, ferric chlorite, Fe concentrations are low and can be accounted for by Al = Fe 3+ substitution. In Fe-rich samples, Fe 3+ may exceed 2 atoms per formula unit (pfu, 18 oxygen basis). When structural formulas are normalized to 28 charges corresponding to the standard O 10 (OH) 8 anionic basis, these measurements define the exchange vector of a di-trioctahedral-type substitution: 3 VI (Mg, Fe 2+ ) = VI □ + 2 VI Fe 3+ , as described in earlier studies. However, structural formulas calculated on the basis of the oxygen contents actually measured by EPMA show that this trend is an artifact, due to the neglect of variations in the number of protons in the structure. Our measurements indicate increasing hydrogen deficiency with increasing Fe 3+ content, up to ~ 2 H+ pfu in the Fe 3+ -rich chlorite samples, corresponding to a net exchange vector of the type R 2+ + H + = Fe 3+ . These results do not support substitutions toward di-trioctahedral ferric end-members, and highlight the need for considering substitution toward an “oxychlorite” (i.e., H-deficient) ferric component, close to tri-trioctahedral, with an O 12 (OH) 6 anionic basis, even in green, pristine-looking chlorite. The effects of iron oxidation and H deficiency on chlorite geothermometers were explored. They are deterring if H deficiency is ignored but, given the sensitivity of most thermometers to octahedral vacancy, the assumption Fe total = Fe 2+ is still safer than using high measured Fe 3+ contents and the standard 28 charge basis, which artificially increases vacancies. In such ferric chlorites, EPMA measurement of oxygen allows a fair estimate of H content if Fe 3+ /Fe 2+ is known; it should be more systematically implemented. For the same reasons, literature data reporting Fe 3+ -rich chlorite with vacancy content along the possibly artificial di-trioctahedral-type substitution should be verified. With the help of constraints from thermodynamic models, charge balance, crystal symmetry, and proton loss, a new cation site distribution is proposed for di-tri- to tri-trioctahedral chlorites in the Fe 2+ -Fe 3+ -Mg-Al-Si-O-H system, allowing a more realistic thermodynamic handling of their solid solutions.

Calcite (CaCO 3 ) is widely considered an important carbon carrier in the Earth’s interior. Laboratory measurements of the velocities and elastic properties of calcite are important for understanding the deep carbon cycle. The sound velocities of calcite were determined up to 10.3 GPa at ambient temperature by Brillouin scattering spectroscopy. Dramatic decreases in the velocity of compressional wave ( V p ) and shear wave ( V s ) and abrupt increases in the V p anisotropy (Ap) and maximum V s anisotropy (As max ) were detected across the phase transition from CaCO 3 -I to CaCO 3 -II. Dramatic increases in the V p and V s and an abrupt decrease in Ap were observed across the phase transition from CaCO 3 -II to CaCO 3 -III. The phase transition from CaCO 3 -I to CaCO 3 -II may potentially explain the Gutenberg discontinuity at 51 km in the Izu-Bonin region. The V p and V s values of calcite were low. Our new results combined with literature data suggest that the low velocities of CaCO 3 could potentially explain the low-velocity zone occurring in northeastern (NE) Japan.

Compositional data, comprising electron probe microanalysis and laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) trace-element data, are presented for common (Cu)-Pb-Sb sulfosalts (bournonite, jamesonite, tetrahedrite, and boulangerite), subordinate semseyite, heteromorphite, robinsonite, and (Cu)-Pb-Bi-Sb sulfosalts, as well as for accompanying base metal sulfides (BMS) in auriferous gold veins from the Gutaishan Au-Sb deposit, southern China. The objectives of the study were to identify whether these sulfosalts represent overlooked hosts for precious metals and other trace elements of petrogenetic or economic interest, establish partitioning trends among coexisting sulfosalt species and between sulfosalts and BMS, and to seek evidence for a genetic link between the abundance of (Cu)-Pb-Sb sulfosalts and the high-fineness of native gold in the deposit. All (Cu)-Pb-Sb sulfosalts analyzed were found to be remarkably poor hosts for gold and thus do not contribute to the overall mineralogical balance for gold. Trace yet measurable concentrations of Au are, however, noted in the (Cu)-Pb-Bi-Sb sulfosalts, in agreement with published data indicating that (Cu)-Pb-Bi-Sb sulfosalts may be minor Au-hosts in some ore systems. Silver is preferentially partitioned into tetrahedrite at the expense of other sulfosalt phases, and tetrahedrite is thus the major host for Ag in the Gutaishan deposit. LA-ICP-MS trace element mapping allows partitioning relationships among different sulfosalt and BMS phases to be determined for several trace elements. Jamesonite concentrates Fe, Zn, Bi, Cd, Ag, Ni, and In over coexisting bournonite, yet boulangerite is the better host for As, Ag, Sn, Se, and Te than jamesonite. Cd and Co are typically enriched in sphalerite relative to any sulfosalt, and when present, pyrite is always enriched in Au and Co relative to all other phases. A high Au/Ag ratio in the ore-forming fluid, the presence of abundant tetrahedrite that has sequestered silver during mineral precipitation, and a lack of evidence for cooling-driven precipitation may be significant reasons for the formation of high-fineness gold throughout the deposit. Two generations of native gold are documented whereby the first is coarse-grained, Ag- and Bi-bearing, and is associated with the main (Cu)-Pb-Sb sulfosalts (bournonite, jamesonite, tetrahedrite, and boulangerite). The second generation is fine-grained and has the highest fineness. Increase in the complexity of sulfosalt assemblages, re-distribution of Ag within coarse native gold and dissolution-reprecipitation reactions among the sulfosalt-gold association increase the gold fineness. The present study shows that linking petrographic aspects at the micrometer-scale with minor/trace element distributions in complex sulfide-sulfosalt assemblages can track a complex history of Au deposition and enrichment.

One of the most important characteristics of the apatite family of minerals is its ability to undergo substitution by ions such as sulfate and carbonate. The substitution of carbonate is particularly important because of the presence of carbonate in bone mineral and the recent suggestion that most of the substituted carbonate resides in the apatite channels (A-type substitution) rather than in place of phosphate (B-type substitution). The possibility of additional channel species or environments in carbonated Ca hydroxylapatite has been pursued via its Sr homolog, which is known to have a larger channel volume and a greater unit-cell a-axial length than its unsubstituted parent. Strontium hydroxyl-, chlor-, and fluorapatites, containing incorporated 13 C-carbonate up to 7 wt%, were synthesized by aqueous precipitation reactions in the presence of Na, K, and ammonium counter cations. 13 C-labeled carbonate was used to facilitate observation of the solid-state 13 C MAS NMR spectra. Band-fitting of the IR carbonate asymmetric stretch ν 3 region required the use of eight bands, arranged as four doublets, that are interpreted as representing three channel environments (A-type substitution) and one B-type substitution. The carbonate ν 3 region of low Na, carbonated Sr hydroxylapatite contained the high-frequency limbs of the A-type doublets in the area between 1520 cm –1 and the B-type region (1400–1360 cm –1 ), whereas in the high Na analog the A-type doublets produced a broad high-frequency tail into the B-type region. Similar behavior was exhibited by the chlor- and fluorapatite analogs. Heating samples to 600 °C resulted in the loss of carbonate and conversion to A-type carbonate demonstrating the stability of A-type carbonate at higher temperatures. Analysis of the populations of A-, A′-, and A″-, and B-sites for the hydroxyl-, chlor-, and fluorapatites, prepared under both low Na and high Na conditions, revealed that high Na/carbonate ratios produce a larger amount of channel substitution, contrary to observations for Ca homologs. It is speculated that multiple A-environments also exist for Ca hydroxylapatite prepared by aqueous precipitation, which is consistent with Fleet’s analysis of apatite prepared at high temperature and high pressure.

We present the first discovery of a Zn-rich mineral on the pristine surface of orange pyroclastic beads from Apollo sample 74220. This Zn-rich mineral is wide occurring, trigonal or hexagonal in shape, with a normalized composition of ~59 wt% Zn, ~26 wt% O (calculated), ~6 wt% S, ~5 wt% Na, and ~4 wt% Cl. The crystal morphology, homogeneity, and chemistry of individual grains are most consistent with gordaite, a zinc chlorohydroxosulfate mineral, showing an empirical formula of Na 1.02 Zn 3.98 [(SO 4 ) 0.84 (OH) 0.30 ](OH) 6 [Cl 0.50 (OH) 0.50 ]·nH 2 O, albeit the exact amounts of OH and H 2 O are uncertain. The pristine 74220 sample used in this study was only directly exposed to air for a cumulative period of 18 days before our study. The same Zn-rich crystals, examined 12 to 15 months apart, show no visible physical and chemical changes. Thus, this zinc-rich mineral likely formed through rapid alteration (oxidation and hydration) by terrestrial air of the original vapor-deposited Zn, Cl, S, and Na-bearing solids. The composition of the zinc-rich mineral indicates that the vapor condensates consist of metallic Zn and metallic Na with either ZnS or native S, and either ZnCl 2 or NaCl. This is the first direct evidence that metallic Zn and Na are key components in the vapor condensates of lunar volcanic gas, which implies lunar volcanic gas may be under higher pressure than previously thought, and the gas composition may be different than previously inferred. Our study is also relevant to the collection, handling, curation, and sample preparation of returned samples from other planetary bodies.

This paper presents a new X-ray absorption spectroscopy (XAS) method for making two-dimensional maps of Fe 3+ in-situ in polished glass samples, which opens the door to study redox changes associated with magmatic processes such as crystallization, assimilation, ascent, and eruption. Multivariate analysis (MVA) allows selection of specific channels in a spectrum to inform predictions of spectral characteristics. Here, the sparse model of the least absolute shrinkage and selection operator (Lasso) is used to select key channels in XAS channels that can be used to predict accurate in-situ Fe 3+ analyses of silicate glasses. By tuning the model to use only six channels, analytical time is decreased enough to allow mapping of Fe 3+ variations in samples by making gridded point analyses at the scale of the XAS beam (1–2 μm). Maps of Fe 3+ concentration can then be constructed using freely available, open source software (http://cars.uchicago.edu/xraylarch/). This result shows the enormous potential of using MVA to select indicative spectral regions for predicting variables of interest across a wide variety of spectroscopic applications. Redox gradients in lunar picritic glass beads first observed with point analyses are confirmed through this XAS mapping and suggest degassing processes during ascent and eruption are responsible for the range of Fe 3+ values measured in these samples.