Volume 99, Issue 11-12

Merrillite, Ca 18 (Ca,⃞)Mg 2 (PO 4 ) 14 -Ca 18 Na 2 Mg 2 (PO 4 ) 14 -Ca 16 REE 2 (Mg,Fe) 2 (PO 4 ) 14 occurs as a primary phosphate along with apatite, in lunar and martian rocks, and in meteorites. It is nominally anhydrous, but attempts to directly measure H in this mineral have not previously been reported. Because of the occurrence on Earth of whitlockite, Ca 18 (Mg,Fe 2+ )2(PO 4 ) 12 [HPO 4 ] 2 , and the apparent incorporation in whitlockite of a merrillite component, the lack of a whitlockite component in extraterrestrial merrillite could be taken as an indicator of low hydrogen fugacity, and this implication has been applied to lunar merrillite. On the other hand, for martian rocks, where magmatic OH or H 2 O contents were likely higher, apatite accordingly contains higher OH contents, yet coexists with anhydrous, Na-rich merrillite. With direct measurements by SIMS, McCubbin et al. (2014), which is in the July issue of American Mineralogist (p. 1347-1354), show that Shergotty merrillite is anhydrous and infer that the high T of crystallization of Shergotty precluded incorporation of a whitlockite component. The mineral pair apatitemerrillite in extraterrestrial rocks constitutes a powerful pair for recording magmatic conditions; however, as McCubbin et al. show, the implications of these minerals and their compositions must be interpreted in light of careful and complete analyses and crystal chemical constraints.

Griffithite, from Miocene volcanic and sedimentary rocks in Griffith Park, Los Angeles, is described by Treiman et al. in this issue. This is a ferrian, trioctahedral saponite with all the Fe 3+ in a distorted octahedral site. It is a close analog of the saponite identified by Mars Science Laboratory at Gale Crater. The martian saponite is likely diagenetic, rather than clastic, and thus may represent the presence of a once-habitable fluvio-lacustrine environment on Mars.

Traditional pathways to discovery in Earth sciences rely in large measure on deductive and inductive approaches, by which measurements and observations are made in the context of established principles or testable, predictive hypotheses about the natural world. Vast, but largely untapped, Earth and life science data resources offer a potentially revolutionary alternative “abductive” approach to investigate Earth’s co-evolving geosphere and biosphere. We therefore advocate a strategic, data-driven program for accelerated scientific discovery.

Fluorapatite has been synthesized through ion-exchange between NIST hydroxylapatite SRM 2910a and optical-grade fluorite. The latter end-member and additional intermediate F-OH apatite samples made through ion-exchange between the newly synthesized fluorapatite and the original hydroxylapatite have provided materials for an investigation of thermal expansion in the F-OH apatite system. Unit-cell volumes determined from X-ray powder diffraction data collected between 22 and 928 °C have been utilized to compute the coefficients of thermal expansion at a pressure of 1 bar for seven apatite samples having various F:OH ratios, as well as for natural samples of fluorapatite and hydroxylapatite. Results show that the thermal expansion coefficient for volume averages 41.0 ± 1.4 (1σ) × 10-6 deg-1 for all samples and is thus little affected by F:OH ratio. This study extends the thermodynamic characterization of this important mineral system, with potential applications for the geological, planetary, biological, medical, and materials-science communities.

Substitution of carbonate has long been recognized in both synthetic and natural biologically and geologically precipitated forms of hydroxylapatite. Although the predominant substitution mechanism in all of the calcium members of the apatite group formed below 100 °C is substitution of carbonate for phosphate (B-type), small amounts of A-type substitution of carbonate in the channel sites also occur. The present study focuses on the effect of cation size on the type of substitution of carbonate in members of the apatite group. The barium and lead members offer a larger channel site, which potentially could stabilize A-type substitution. A series of carbonated barium and lead fluorapatites were synthesized in aqueous solution and characterized by powder X-ray diffraction, and infrared and Raman spectroscopy. Carbonate content was determined by combustion analysis. Unit-cell parameters derived from X-ray diffraction showed that, as carbonate content increased, the a -axis length decreased and the c -axis increased slightly for carbonated barium fluorapatites (CBaApF), whereas the lengths of both the a - and c -axes increased for CPbApF. The co-occurrence of two sets of peaks in the n3 carbonate region of the infrared spectra of lead and barium carbonated apatites are strongly suggestive of both A- and B-type carbonate substitution. This interpretation is supported by Rietveld analysis of X-ray powder diffraction data, which confirms the presence of significant, but not dominant, A-type substituted carbonate ions. The variation in cell parameters as a function of carbonate substitution mode is discussed, and it is shown that B-type carbonate substitution need not be accompanied by a decrease in the a -axis in all apatites. The greater amount of A-type carbonate substitution in barium and lead fluorapatites compared to the their calcium homologs can be attributed to: (1) the less negative enthalpies of hydration of the barium and lead ions relative to that of calcium, and/or (2) the greater amount of space available for the relatively large carbonate ions in the channels defined by these large cations. Thus, the substitution modes can be controlled by either thermodynamic or kinetic considerations.

Four gahnite single crystals with variable colors from pale blue to green have been studied by a multi-analytical approach with the aim to evaluate existing assignments of optical absorption bands. Combined information from electron microprobe analyses, Mössbauer spectroscopy, IR-spectroscopy, single-crystal X‑ray structure refinements, and optical absorption spectroscopy confirms the conclusions of earlier studies that the absorption bands recorded in the visible spectral region up to ~540 nm (above ~18 500 cm-1) are related to electronic d-d transitions in tetrahedrally coordinated Fe 2+ . It also demonstrates that a set of absorption bands between ~550-625 nm (~16 000-18 200 cm -1 ) are caused by spin-allowed and spin-forbidden d-d transitions in tetrahedrally coordinated Co 2+ . Two absorption bands at higher wavelengths (~680 and ~800 nm, i.e., ~14 700 and ~12 500 cm -1 ) are assigned to electronic transitions in exchange coupled VIFe 3+ -IVFe 2+ pairs and a band at ~950 nm (~10 500 cm -1 ) is assigned to a spin-allowed electronic transition in VIFe 2+ . Low-Fe gahnite crystals owe their blue color to traces of cobalt at concentration levels in the order of 200 ppm and less, while the green color of gahnite crystals with higher Fe-contents is due to a combination of electronic ligand-metal transitions causing strong UV-absorption and electronic transitions in exchange coupled Fe 1 -Fe 3+ cation pairs that absorb in the red region of the visible spectrum. A detailed characterization of samples that includes cation site occupancy and iron valency data is demonstrated to be crucial for interpreting optical absorption spectra. Also electronic transitions in trace element chromophores below the detection limit of electron microprobe may participate to light absorption. All this information contribute to the comprehension of the causes of crystal color of minerals, gemstones, and ceramic pigments

Si-bearing magnetite or silician magnetite is common in low- and high-temperature rocks. However, details about possible Fe-silicate or Si-Fe-oxide discrete phases/nano-precipitates were not available due to the limitations of conventional high-resolution TEM. Combining Z-contrast imaging and ab initio calculation using density functional theory (DFT) method, we have derived both composition and crystals structure of the discrete nano-precipitates within host magnetite. The nano-precipitates of Si-magnetite with composition of [¨0.5Fe 2+ 0.5] VI [Fe 3+ ]VISiIVO 4 or g-Fe 1.5 SiO 4 occur in silician magnetite from a banded iron formation from Western Australia. In the Si-magnetite precipitates, Si replaces Fe 3+ in tetrahedral sites of the magnetite structure and vacancies are introduced in the octahedral Fe 2+ sites. The Si-magnetite precipitates distribute along {111} of the host magnetite. Widths of the precipitates are even multiples of d 111 of magnetite, such as 2d 111 , 4d 111 , and 6d 111 . Ordering of the vacancies in the Si-magnetite will result in symmetry of P4332, which is a subgroup of Fd3m for magnetite. Stacking of Si-magnetite and magnetite (111) layers along the [111] direction also occur in magnetite. The nanoprecipitates result from exsolution of Si-magnetite from the host silician magnetite at low temperature. The occurrence of the thin nano-precipitates within the magnetite host results from the minimization of interfacial energy between the precipitate and the host magnetite. Relatively high concentrations of aqueous silica and Fe-silicate complex species in pore fluid might enhance the incorporation of Si into the silician magnetite during crystallization of the magnetite.

In this study, we report results of the application of a novel procedure for modeling cation ordering in two synthetic Mg(Al 2-y Fey 3+ )O 4 spinels (y ~ 0.39 and 0.54, samples F39 and F54, respectively). The kinetic profiles suggest a two-stage mechanism, with rapid intersite exchange of Fe 3+ with Mg followed by slow exchange of Al with Mg. The trial to apply classical approaches, based on explicit solutions of the differential equations corresponding to single-cation (Sha-Chappel model) or twocation (Müller model) exchange reactions, proved not feasible in the whole time range, thus implying a lack of information about exchange processes involving the Fe 3+ cation. Therefore we decided to adopt an alternative methodology where a multistep pathway is simulated by Gepasi modeling. The suggested set of reactions and the simultaneous solving of the related rate laws allowed us to calculate rate constants and corresponding activation energies not only for the Al/Mg cationic interchange (257 and 264 kJ/mol for the F39 and F54 samples, respectively), but also, for the first time in such kind of samples (three cations with two sites), for the Fe/Mg cationic interchange (204 and 234 kJ/mol for low- and high-Fe 3+ samples, respectively). The evaluation of the significant effect of the iron content, which actually inhibits Mg-Al exchange between the T and M sites, provided new insights useful for the construction of geothermometers based upon iron-bearing spinels.

Experiments were conducted to reproduce reaction rims of phlogopite ± diopside around olivine that have been observed within a wide range of potassic melts, including phonolite. Phlogopite is also a common secondary phase formed at the expense of olivine during metasomatic events involving K 2 Oand H 2 O-rich fluids or melts. Piston-cylinder experiments where olivine single crystals were reacted with synthetic phonolite melt at 10.7-14.7 kbar and 950-1000 °C recreate the mineralogy and textures documented in natural samples. Rim growth is parabolic with time, indicating a diffusion-controlled reaction. Fast diffusion in the melt and varying compositions across the phlogopite reaction rims suggest that diffusion through the rims, along grain boundaries is rate limiting. Reaction rates dramatically increase with temperature as well as the bulk water content of the sample charge. This is because of increasing amounts of atomically bound hydrous species along the grain boundaries that increase the rates of diffusion and thereby the rates of rim growth. Atomically bound hydrous species increase the rates of rim growth by lowering the activation energy for diffusion and by increasing the solubility of diffusing species in the grain boundary region. Transmission electron microscopy shows the presence of isolated pores and open grain boundaries. Most of these may have opened during quenching, but there is some evidence to suggest that a free fluid phase may have been locally present in experiments with high melt water contents (>8 wt%). The measured rim growth rates at different conditions are used to estimate residence times of reacting olivine crystals in natural systems.

Various approaches have been used to estimate when and how long liquid water was present at the surface of Mars. The olivine dissolution-lifetime application suggested by Stopar et al. (2006) and Olsen and Rimstidt (2007) is here adapted and tested at the scale of individual etch-pits using published data from an experimental system in which the volume of mineral removed and the duration of the mineralremoval episode are known. Different assumptions about the specific geometry of etch-pits on olivine result in surface-area estimates that vary by less than a factor of two from the simple hemispherical pit used in the calculations. Given that other sources of uncertainty in mass-time relationships of silicatemineral dissolution during natural weathering can be up to four orders of magnitude, the effects of differing geometric assumptions about the shapes and surface areas of the etch-pits are negligible. Using compiled experimentally determined forsterite dissolution rates and the imaged etch-pit sizes from experiments recovers the duration of the experiment that produced the imaged etch-pits to within less than a factor of two. This suggests that extensively etched olivine surfaces imply a dominance of the etch-pit walls over the bulk surface between the etch-pits during olivine corrosion. The approach adopted here recovers the timescales of experimental etch-pit production on olivine at STP and extreme undersaturation of the solution with respect to olivine in experiments where pH is known. Continued progress in understanding the fundamentals of olivine dissolution kinetics will narrow the ranges of uncertainty in mineral-lifetime estimates at Mars’ surface in support of constraining the compositions and duration of potentially habitable aqueous solutions on Mars

Ferrian saponite from the eastern Santa Monica Mountain, near Griffith Park (Los Angeles, California), was investigated as a mineralogical analog to smectites discovered on Mars by the CheMin X-ray diffraction instrument onboard the Mars Science Laboratory (MSL) rover. The martian clay minerals occur in sediment of basaltic composition and have 02l diffraction bands peaking at 4.59 Å, consistent with tri-octahedral smectites. The Griffith saponite occurs in basalts as pseudomorphs after olivine and mesostasis glass and as fillings of vesicles and cracks and has 02l diffraction bands at that same position. We obtained chemical compositions (by electron microprobe), X-ray diffraction patterns with a lab version of the CheMin instrument, Mössbauer spectra, and visible and near-IR reflectance (VNIR) spectra on several samples from that locality. The Griffith saponite is magnesian, Mg/(Mg+SFe) = 65-70%, lacks tetrahedral Fe 3+ and octahedral Al 3+ , and has Fe 3+ /SFe from 64 to 93%. Its chemical composition is consistent with a fully tri-octahedral smectite, but the abundance of Fe 3+ gives a nominal excess charge of +1 to +2 per formula unit. The excess charge is likely compensated by substitution of O2- for OH-, causing distortion of octahedral sites as inferred from Mössbauer spectra. We hypothesize that the Griffith saponite was initially deposited with all its iron as Fe 2+ and was oxidized later. X‑ray diffraction shows a sharp 001 peak at 15 Å, 00l peaks, and a 02l diffraction band at the same position (4.59 Å) and shape as those of the martian samples, indicating that the martian saponite is not fully oxidized. VNIR spectra of the Griffith saponite show distinct absorptions at 1.40, 1.90, 2.30-2.32, and 2.40 mm, arising from H 2 O and hydroxyl groups in various settings. The position of the ~2.31 mm spectral feature varies systematically with the redox state of the octahedrally coordinated Fe. This correlation may permit surface oxidation state to be inferred (in some cases) from VNIR spectra of Mars obtained from orbit, and, in any case, ferrian saponite is a viable assignment for spectral detections in the range 2.30-2.32 mm.

A global spectral survey of 4506 immature craters with diameters <1 km was carried out using near-IR data from the Kaguya Spectral Profiler to characterize the composition of the lunar megaregolith. On the basis of band minima and radiative transfer mixing models, small crater spectra fall into three groups: (1) mare basalts with strong absorptions at relatively long wavelengths indicating high ratios of high- to low-Ca pyroxene; (2) norites containing about 50% plagioclase and with pyroxene assemblages dominated by low-Ca pyroxene that occur within the South Pole-Aitken Basin (SPA), near Apollo 14 and other locations near Imbrium Basin, and three major cryptomaria deposits; and (3) noritic anorthosites occurring within the Feldspathic Highlands Terrane containing about 20 wt% pyroxene with a pyroxene assemblage containing exclusively very low-Ca pyroxene. Very few pure anorthosites are present in this survey and there are no occurrences of pyroxene-poor olivine-rich assemblages. Models of the composition of basin ejecta incorporate large amounts of mantle material and the spectral results require that that the sampled mantle is orthopyroxenite. Basin depth-diameter ratios used in the models required to match the measured composition are consistent with prior estimates for the largest basins. The composition found in the SPA and Imbrium regions are consistent with mafic impact melt breccias or basaltic impact melts of basin origin. For SPA we model this composition and find it requires an extremely low impact angle. While this is consistent with prior work on an oblique impact for the SPA event, a more robust solution invokes the production of norite in impact melt seas.

Decomposition of oxalic acid in the presence of water was examined in a hydrothermal diamondanvil cell up to 800 °C and 970-1480 MPa as a function of oxygen fugacity to assess its usefulness as a C-O-H fluid source in petrologic experiments. Fluid, vapor, and solid species were identified in situ at elevated temperature and pressure with Raman spectroscopy and optical microscopy. Under oxidizing conditions (buffered by the reaction NiO ↔ Ni+½O 2 ), oxalic acid decomposes to carbon dioxide and water. Under reducing conditions (buffered by the reaction Mo+O 2 ↔ MoO 2 ), oxalic acid decomposes to methane and hydrogen. Under unbuffered conditions, at intermediate oxygen fugacity (~0 to 1 log units below the fayalite-quartz-magnetite buffer), oxalic acid disproportionates to graphite and minor methane and carbon dioxide. The results from the Ni-NiO-buffered and Mo-MoO 2 -buffered experiments result in observed fluid species that are similar to those predicted by previous investigations. However, there are substantial differences between our results and previous studies of oxalic acid decomposition in the unbuffered experiment that was within a log unit of the fayalite-magnetite-quartz (FMQ) buffer. These include the detection of aqueous C-H species at temperatures as low as 400 °C and a solid graphite-like phase at 800 °C. These differences can be explained if we consider that aqueous H2 in our experiment reacted to form the C-H species, instead of being lost via diffusion through the H2-permeable capsules used in previous studies. Consequently, for experiments within about 1 log unit of the FMQ buffer curve, oxalic acid is likely a poor choice for a C-O-H fluid source because the formation of graphitic carbon would result in significant deviations from the expected C-O-H fluid composition and concentration (i.e., CO 2 +H 2 O). At oxygen fugacities outside a log unit of FMQ, the observed fluid species are similar to those predicted by previous investigations and the use of oxalic acid as a C-O-H fluid source is permissible from the perspective of oxygen fugacity, although other system parameters (e.g., sample geometries, capsule thickness, capsule materials, gasket materials, wall thickness) must still be considered.

The first European occurrence of slawsonite is reported from a picrite sill within Upper Ordovician strata of the Prague Basin near the village of Rovina, Czech Republic. The rare slawsonite forms an interstitial phase in association with abundant celsian and hyalophane, replacing the original calcic plagioclase (bytownite). A study of this curious natural slawsonite-celsian-hyalophane assemblage provides a valuable insight into feldspar stability and petrogenesis.Whole-rock geochemical signatures of the picrite sill and underlying doleritic basalt intrusion show conspicuous enrichment in Sr and Ba superimposed on normal basaltic multielement patterns. These two elements were most likely introduced by intergranular fluids during diffusional seafloor metasomatism (rodingitization and serpentinization) of the picrite. Strontian and barian feldspars precipitated directly from BaO-SrO-H 2 Obearing fluid, which caused decomposition of plagioclase to vuagnatite, aqueous SiO 2 and Al 2 O 3 at T ≤ 350 °C. Subsequently, vuagnatite decomposed to hydrogrossular and excess SiO2 was consumed by serpentinization of olivine. At the expense of aqueous Al 2 O 3 , serpentine reacted to chlorite closing the picrite alteration at 320-160 °C. Pressure did not exceed 0.5 GPa. The in situ EDS analyses indicate that the chemical composition of the slawsonite is Sl 91 Cn 3 An 3 Ab (core) to Sl 82 Cn 3 An 4 Ab 9 Or 2 (rim), the celsians range from Cn 96.9 An 0.3 Ab 0.2 Or 2 Sl 0.6 to Cn 76.3 An 4.7 Ab 3 Or 15.7 Sl 0.3 , and the hyalophanes vary from Cn 72.2 An- 0.8 Ab 5.1 Or 21.1 Sl 0.2 to Cn 57.3 An 0.8 Ab 3.1 Or 38.5 Sl 0.3 .

Hematite occurs in various geologic settings including igneous, metamorphic, and sedimentary rocks as well as in soils. However, it frequently occurs at low concentrations, especially in soils, where it may be <1% by weight. Because hematite has the potential to be an indicator of oxidizing and climatic conditions in soils and paleosols, it is important to understand its limit of detection. In this paper we examine the limits of detection of hematite visually and with diffuse reflectance spectrophotometry (DRS) and X‑ray diffraction (XRD). To accomplish this we used a sample set consisting of “knowns” or calibration samples. These known samples consisted of 15 different matrices of varying mineral composition into which hematite in 7 different concentrations ranging from 0.01 to 4% by weight were mixed. Including the 0% hematite, our calibration data set consisted of 120 samples. Visually, hematite can be detected at a concentration of 0.01% by weight in a light matrix and 0.5% in the darkest of our matrices. However, because of metamerism, visual techniques cannot specifically identify hematite. We find that for both DRS and XRD the limit of detection is also dependent on the matrix. For XRD the limit of detection for hematite in bulk samples is about 1%. For DRS the limit of detection depends on the data reduction technique used. The commonly used Kubelka-Munk remission function and its first and second derivatives can easily identify hematite at the 0.5% level. However, the first derivative of the percent reflectance curve can detect hematite at 0.01% by weight in a light matrix and 0.05% in a dark matrix. We suggest that the first derivative of DRS curves is the best currently available method for qualitatively detecting the mineral hematite at low concentrations found in soils, sediments, and rocks. Work described in this paper may be applied in several situations. Our study of visual limits of hematite detection should aid field geologists in assessing hematite content. Analysis of color wavelength bands may also have application in remote sensing by indicating which bands are most sensitive to hematite, reported to be an important constituent of the martian surface. Furthermore, this study could help clarify remotely sensed terrestrial albedo changes, especially the Sahara/Sahel transition where the sediments change from light, quartz-dominated to dark, hematite-dominated. Our study also points out that with laboratory-based spectra the first derivative of the reflectance curve is the most sensitive transform for processing spectral data for hematite, thereby allowing concentrations as low as 0.01% to be detected.

Multiple saturation experiments have been performed in a multicomponent system at 6.3 to 7.5 GPa and 1400-1670 °C using a split-sphere multi-anvil apparatus to constrain the conditions of kimberlite magma generation. The starting bulk compositions of samples corresponded to the average group II kimberlite (orangeite), with water contents varying from 5 to 9 wt% H 2 O and the CO 2 /(CO 2 +H 2 O) molar ratio from 0.37 to 0.24. The charges were placed inside graphite liners sealed in Pt capsules to avoid Fe loss. Oxygen fugacity ( f O2) during the experiment was buffered by the equilibrium between graphite and a hydrous carbonate-silicate melt about EMOG/D. As water in the starting kimberlite increased from 5 to 9 wt%, the temperature of its complete melting became ~100 °C lower (relative to 1670 °C), both in the 6.3 and 7.5 GPa runs. Orthopyroxene was stable just below the liquidus at all pressures and H 2 O concentrations applied in the experiments. An olivine + garnet + orthopyroxene assemblage was present at ≤100 °C below the liquidus when H 2 O was 5 wt%. At 7 and 9 wt% H 2 O, the same assemblage appeared at 100-150 and >200 °C below the liquidus, respectively. In no experiment was clinopyroxene observed as a run product. Olivine, garnet, and orthopyroxene stable in the multiply saturated melt were compositionally similar to mantle peridotite minerals found as xenoliths in kimberlites worldwide. Thus we infer that generation of group II kimberlite magma may occur by partial melting of carbonated (metasomatized) garnet harzburgite at pressures from 6.3 to 7.5 GPa, temperatures about 1500-1600 °C, and no more than 5 wt% H 2 O in the melt. Water, in the amounts required to produce this magma, may come from interaction of K-Ca-rich carbonatite melt, infiltrating from a deeper mantle source, with a peridotite protolith containing H 2 O in nominally anhydrous minerals and, possibly, also in phlogopite.

Barrier heights (BHs) for hydrolysis and H 2 O exchange reactions at M-O-Si (M = Ni 2+ , Mg 2+ , and Ca 2+ ) linkages on olivine (M 2 SiO 4 ) mineral surfaces were determined via DFT calculations. BHs for hydrolysis of protonated Ni-O-Si, Mg-O-Si, and Ca-O-Si sites are 76, 54, and 27 kJ/mol, respectively, and are 69 and 24 kJ/mol for H 2 O exchange reactions of protonated Mg-O-Si and Ca-O-Si sites, respectively. Rate constants were calculated via classical transition state theory (TST) using these BHs. For protonated Ni-O-Si, Mg-O-Si, and Ca-O-Si sites, these are 7.2 × 10-1, 4.7 × 104, and 1.5 × 109 s-1 [pseudo-first-order where (H 2 O) is assumed to be constant], respectively, and for H 2 O exchange at protonated Mg-O-Si and Ca-O-Si sites are 2.6 × 101 and 3.7 × 109 s-1 [pseudo-first-order where (H 2 O) is assumed to be constant], respectively. Approach of an H 2 O molecule from the second hydration sphere toward a protonated Ni-O-Si site leads to breakage of the Ni-O bond and subsequent release of Ni 2+ to solution. For protonated Mg-O-Si sites, however, H 2 O exchange does not lead to rupture of the Mg-O bond and would not be a step toward dissolution of the mineral. Potential energy surface (PES) scans of H 2 O exchange indicated formation of a hepta-coordinated Ca 2+ , so neither H 2 O exchange nor hydrolysis of the Ca-O-Si linkage occurred in this case. Calculated rate constants are consistent with experimental data for end-member composition olivine minerals where observed rates of dissolution increase in the order Ni 2+ < Mg 2+ < Ca 2+ .

The Miller Range (MIL) 03346 nakhlite contains ~20% mesostasis, which contains skeletal titanomagnetite. The titanomagnetite contains trellis-type {111} lamellae of ilmenite similar to those found in terrestrial titanomagnetites that have experienced subsolidus redox reactions during cooling of their host rocks. We have characterized the MIL 03346 titanomagnetite-ilmenite intergrowths by a combination of focused ion beam (FIB), energy-dispersive spectroscopy (EDX), and high-resolution transmission electron microscopy (TEM). The resulting structural and chemical analyses have been combined with temperature and f O 2 data from previous studies of MIL 03346, as well as previous work on two-oxide thermobarometry of nakhlites. Our calculations show that as MIL 03346 and other nakhlites cooled below 800 °C, they recorded increasingly reducing conditions, such that the lowest temperatures calculated correspond to f O 2 conditions as low as 4 log f O 2 units below the FMQ buffer. However, the MIL 03346 lamellae must have formed by oxidation and thus record a very late stage lowtemperature (<350 °C) oxidation event. When considered together with previous studies of MIL 03346 and nakhlites in general, the overall cooling history could be explained by early oxidation followed by intermediate stage reduction caused by S 2 loss by degassing, followed by late loss of C 1 by degassing.

Lanthanides “Ln” (rare earths) are critical elements found in natural minerals such as calcium phosphate apatites, in sedimentary and igneous settings as well as in skeletal diagenesis. From a medical point of view, nanoparticles of lanthanide-doped apatites can be produced for conferring luminescence properties of interest in cancer cells detection. However, the impact of the substitution of Ln for Ca on the stability and solubility of related apatite phases is still essentially unknown. To investigate the thermochemical effects of such lanthanide substitution for calcium in apatite, we prepared and analyzed four series of apatites with up to 10% lanthanide substitution for calcium. After thorough physicochemical characterization via complementary techniques (XRD, FTIR, TG/DSC, and IPCAES), high-temperature oxide melt solution calorimetry in molten sodium molybdate at 973 K was performed to determine their enthalpies of formation from constituent oxides and from the elements, at 298 K. Our results indicate that although enthalpies of formation are strongly exothermic in all cases, Ln-doping has a destabilizing effect, which increases with dopant concentration and with the size of the incorporated Ln 3+ ion. After estimating standard entropies, Gibbs free energies of formation and equilibrium constants for Ca 2+ /Ln 3+ exchange reactions in apatite were then evaluated, for the first time allowing access to quantitative thermodynamic data that may be used in various fields for stability calculations or partitioning estimates between fluids and solids.

Non-equilibrium molecular dynamics (NEMD) simulations are used to compute the phonon thermal conductivity (k) for liquids and glasses of composition Mg 2 SiO 4 , CaMgSi 2 O 6 , and NaAlSi 3 O 8 at 2000-4500 K and 0-30 GPa based on classical potentials. These compositions span the range of melt polymerization states in natural systems at ambient pressure. The NEMD results compare well with available laboratory measurements on molten NaAlSi 3 O 8 and CaMgSi 2 O 6 at 1 bar. Thermal conductivities decrease with increasing temperature (T), increase with increasing pressure (P), and at low pressure increase slightly as the mean coordination number of Si and Al around oxygen increases, in the sequence Mg 2 SiO 4 , CaMgSi 2 O 6 , and NaAlSi 3 O 8 . At 3500 K, the thermal conductivity of CaMgSi 2 O 6 at 0, 10, 20, and 30 GPa is 1.1, 2.1, 2.5, and 3 W/mK, respectively. At ambient pressure (0.2 ± 0.15 GPa), k = 1.2 and 0.5 W/mK at 2500 and 4500 K, respectively, for CaMgSi 2 O 6 . For NaAlSi 3 O 8 composition, k varies from 1.7 to 2.7 W/mK at 3050 K for pressures of 6 and 30 GPa, respectively. Mg 2 SiO 4 liquid at ambient pressure (0.07 ± 0.16 GPa) is found to have thermal conductivities of 1.36 and 0.7 W/mK at 2500 and 4500 K, respectively. Tables giving computed k values for all compositions are included for state points studied. The trade-off between T and P implies that the phonon thermal conductivity of silicate liquids at mantle depths increases substantially (factor of 2-3) along isentropes.

Aluminum hydroxide ions in the e-Keggin structure provide geochemical models for how structure affects reactivity, and consequently, how aqueous ions evolve to bulk precipitates. Here we report a systematic comparison of heterometal substitution into the MAl12 e-Keggin structure, where M = Ga III , Al III , or Ge IV . We use direct solution calorimetric techniques to compare the energetics of these substituted structures and complement these measurements with density functional theory (DFT) calculations to further examine this structure as a host to alternative heterometals. The measured enthalpy of solution (DHsoln) at 28 °C in 5 N HCl for the selenate salts of GaAl 12 7 + and AlAl 12 7+ , was measured as -869.71 ± 5.18 and -958.04 ± 2.79 kJ/mol, respectively. The enthalpies of formation from the elements, DH°f,el, for the selenate salts of GaAl 12 7+ and AlAl 12 7+ , are -23 075.02 ± 61.68 and -23 334.18 ± 60.38 kJ/mol, respectively, supplanting previous values. We compare structural relationships to both experimental and calculated energies to identify the driving forces that control these substitutions and stability, and establish that tetrahedral M-O bond lengths are closely related to the strain and stability of the structure. We show that substitution depends on the size and valence of the heterometal through energetics, and we extend our thermodynamic and structural relationships to other not yet synthesized MAl 12 clusters (M = Si IV , Fe III , Be II , Mg II , or Zn II ).

Plagioclase feldspars comprise a large portion of the Earth’s crust and are very anisotropic, making accurate knowledge of their elastic properties important for understanding the crust’s anisotropic seismic signature. However, except for albite, existing elastic constants for plagioclase feldspars are derived from measurements that cannot resolve the triclinic symmetry. We calculate elastic constants for plagioclase end-members albite NaAlSi 3 O 8 and anorthite CaAl 2 Si 2 O 8 and intermediate andesine/ labradorite NaCaAl 3 Si 5 O 16 using density functional theory to compare with and improve existing elastic constants and to study trends in elasticity with changing composition. We obtain elastic constants similar to measured elastic constants and find that anisotropy decreases with anorthite content.

The occurrence of a trioctahedral analog of illite, the dioctahedral interlayer-deficient K-mica, has long been debated. Due to the inherent difficulties of determining structure and chemical composition of the extremely fine-grained material, earlier descriptions based on separated material are equivocal. Here we describe low-temperature (diagenetic) formation of fluorophlogopite, which is interlayerdeficient and therefore analogous to illite, using high-resolution in situ methods (transmission electron microscopy, TEM, with preparation by focused ion beam milling, combined with wavelength-dispersive analysis by field-emission gun electron microprobe). The average composition is K 0.5 Mg 2.8 V 0.01 Fe 0.005 [Si 3.15 Al 0.85 O 10 (OH) 0.65 F 1.35 ], including minor amounts of NH 4 for charge compensation as determined by electron energy loss spectroscopy. The K-deficient Mg-mica occurs in layer packages of ~10 layers, and no indications for interlayering with other sheet silicate layers such as chlorite or vermiculite could be identified with TEM. X-ray powder diffraction patterns of separated material confirm the absence of smectite components. The mineral was identified in phosphorites from the lowermost Cambrian Tal Group, Mussoori Syncline, Lesser Himalayas, India. The rocks are alternating phosphatic mudstones and phosphatic dolostones, at times interbedded with phosphate-poor carbonate layers, which are rich in organic matter. Sedimentary fluorophlogopite occurs in both rock types and in two textural associations; one in vesicles filled with amorphic organic matter, the other as reaction rims around illite, which contains up to 5 wt% V 2 O 3 in its rims. Textural arguments favor an early diagenetic formation of both, V-bearing illite and fluorophlogopite, closely associated with organic matter and linked to dolomitization. The high-F content stabilizes phlogopite to low temperatures. Our findings confirm that the stability field of fluorophlogopite extends from magmatic to metamorphic and sedimentary conditions.

Sepiolite is widely used in various fields due to its unique colloidal-rheological and physicochemical properties. The first experimental thermochemical study of natural sepiolite Mg 8 Si 12 O 30 (OH) 4 (H 2 O) 4 ·nH 2 O from Akkermanovskoe field (Southern Ural, Russia) was performed utilizing the high-temperature heat-flux Tian-Calvet microcalorimeter. X-ray powder diffraction, thermal analysis, and FTIR spectroscopy methods were used to characterize sepiolite. Processes of dehydration, dehydroxylation, and various water types’ removal enthalpies were studied using thermochemical methods. The values of Ddehydr.H 0 (298.15 K) of adsorbed, zeolitic, and bound water calculated per 1 mol of released H 2 O, were as follows: 15 ± 4, 28 ± 8, and 39 ± 15 kJ/mol, respectively. The enthalpy of dehydroxylation of sepiolite was found as 145 ± 14 kJ/(mol H 2 O). Obtained data point at different binding strengths of water in the structure of sepiolite. The enthalpies of formation from the elements ΔH0f (298.15 K) were derived by melt solution calorimetry for sepiolite with various content of different water types: -18 773 ± 28 kJ/mol for Mg 8 Si 12 O 30 (OH) 4 (H 2 O) 4 ·4H 2 O and -16 426 ± 21 kJ/mol for Mg 8 Si 12 O 30 (OH) 4 (H 2 O) 4

Dense hydrous magnesium silicates (DHMSs) may play an important role in water transport during planetary accretion and as water reservoirs in the Earth’s deep mantle. We show that the dynamic decomposition products of antigorite, Mg 3 Si 2 O 5 (OH) 4 , can be interpreted as containing the newly discovered, dense hydrous silicate, phase H (MgSiO 4 H 2 ). The Hugoniot for phase H was calculated based on the Hugoniots for its constituent oxides and equation of state data derived from first-principles calculations. The measured antigorite Hugoniot, previously suggested to decompose into high-pressure phases without generating fluid H 2 O, was compared with those derived from calculations involving phase H. Sound velocity data were also compared to confirm that the dynamic breakdown product of antigorite at pressures above ~40 GPa is most likely phase H plus MgO without formation of fluid H 2 O.

A new general model that takes into account the pressure and redox state effect is presented to calculate melt viscosities of natural Fe-bearing melts. This new model is applicable to melts that span a wide range of temperatures (from 733 to 1873 K), pressures (0.001-15 kbar), H 2 O content (from 0 to 12.3 wt%), and compositions (from ultramafic, mafic to silicic melts). The accuracy of the model is calculated to be ±0.23 log units of viscosity, which is within or close to experimental uncertainty. The transport properties, including glass transition temperature and melt fragility, can also be calculated from this model. A spreadsheet to calculate the viscosity is provided in an Electronic Supplement.

We report the replacement of chalcopyrite by bornite under hydrothermal conditions in solutions containing Cu(I) and hydrosulfide over the temperature range 200-320 °C at autogenous pressures. Chalcopyrite was replaced by bornite under all studied conditions. The reaction proceeds via an interface coupled dissolution-reprecipitation (ICDR) mechanism and via additional overgrowth of bornite from the bulk solution. Initially, the reaction is fast and results in a bornite rim of homogeneous thickness. Reaction rates then slow down, probably reflecting healing of the porosity, and the reaction proceeds predominantly along twin boundaries of the chalcopyrite. The composition of the bornite product is generally Cu-rich, corresponding to the bornite-digenite (Cu 5 FeS 4 -Cu 9 S 5 ; Bn-Dg) solid solution (bdss). The Cu and Fe contents were controlled principally by temperature, with solution pH having only a small effect. The percentage of Cu in bdss decreased and the percentage of Fe increased with increasing reaction temperature: at 200 °C a composition of Bn 47 Dg 53 was obtained; at 300 °C the composition was Bn 90 Dg 10 and at 320 °C it was near-stoichiometric bornite. The influence of temperature rather than solution chemistry on the composition of bdss, as well as the homogeneity of the bornite product grown both via replacement of chalcopyrite and from the bulk solution as overgrowth, are interpreted to reflect buffering of the bornite activity in bdss via solids (e.g., reaction chalcopyrite + 2 chalcocite = bornite). Only the end-member compositions of the bdss are found in nature, indicating that the products obtained are metastable, and illustrating the importance of reaction mechanism for controlling the chemistry of the mineral product. The unique features of the chalcopyrite to bornite reaction investigated here are related to interaction between a solution controlled ICDR reaction with solid-state diffusion processes driving porosity healing.

Porosity plays a key role in the formation and alteration of sulfide ore minerals, yet our knowledge of the nature and formation of the residual pores is very limited. Herein, we report the application of ultra-small-angle neutron scattering and small-angle neutron scattering (USANS/SANS) to assess the porosity in five natural sulfide minerals (violarite, marcasite, pyrite, chalcopyrite, and bornite) possibly formed by hydrothermal mineral replacement reactions and two synthetic sulfide minerals (violarite and marcasite) prepared experimentally by mimicking natural hydrothermal conditions. USANS/ SANS data showed very different pore size distributions for these minerals. Natural violarite and marcasite tend to possess less pores in the small size range (<100 nm) compared with their synthetic counterparts. This phenomenon is consistent with a higher degree of pore healing or diagenetic compaction experienced by the natural violarite and marcasite. Surprisingly, nanometer-sized (<20 nm) pores were revealed for a natural pyrite cube from La Rioga, Spain, and the sample has a pore volume fraction of ~7.7%. Both chalcopyrite and bornite from the massive sulfide assemblage of the Olympic Dam deposit in Roxby Downs, South Australia, were found to be porous with a similar pore volume fraction (~15%), but chalcopyrite tends to have a higher proportion of nanometer-size pores centered at ~4 nm while bornite tends to have a broader pore size distribution. The specific surface area is generally low for these minerals ranging from 0.94 to 6.28 m2/g, and the surfaces are generally rough as surface fractal behavior was observed for all these minerals. This investigation has demonstrated that USANS/SANS is a very useful tool for analyzing porosity in ore minerals. We believe that with this quantified porosity information a deeper understanding of the complex fluid flow behavior within the porous minerals can be expected.

We report the synthesis of h-magnetite, ideally h-Fe 3 O 4 with considerable amounts of substitutional cations (Cr, Mg, Al, Si) and quenchable to ambient conditions. Two types of experiments were performed at 18 GPa and 1800 °C in a multi-anvil press. In one, we used an oxide mixture with a majoritic stoichiometry Mg 1.8 Fe 1.2 (Al 1.4 Cr 0.2 Si 0.2 Mg 0.2 )Si 3 O 12 , with Si and Mg in excess as starting material (MA-367, MA-380). In the second type of experiment (MA-376), we started from an oxide mixture on the composition of the Fe-oxide phase obtained in MA-367. The Fe-oxide phases of both experiments were investigated by electron microprobe and transmission electron microscopy including electron diffraction tomography. Our investigations show that the Fe-oxide phases crystallize in the structure-type of h-magnetite. However, electron diffraction data show that keeping the cell setting from literature, this phase crystallizes in space group Amam and not in space group Bbmm as previously proposed. In the experiment MA-367, the Fe-oxide phase are mutually intergrown with majorite, the major phase of the run products. The formula for h-magnetite in this run was calculated as Fe1(Fe2+ 0.75Mg0.26)Fe2(Fe3+ 0.70 Cr 0.15 Al 0.11 Si 0.04 ) 2 O 4 . In the experiment on the bulk composition of the Fe-oxide, the main phase was h-magnetite with composition Fe 1 (Fe 2+ 1.02 )Fe 2 (Fe 3+ 0.65 Cr 0.19 Al 0.13 Si 0.03 ) 2 O 4 and traces of nearly pure endmember wadsleyite and stishovite. Our results indicate that the substitution of 20 to 30% of Fe (0.7 to 0.9 atoms per formula unit) by smaller cations favored the preservation of the high-pressure form to ambient conditions. We prove that the h-magnetite-type oxide is also stable in chemical systems more complex than Fe-O. Based on our results, which were obtained at 18 GPa and 1800 °C in a system (MA-367) that is closely related to Fe-enriched oceanic lithospheric material, we suggest that a Fe 3 O 4 -rich phase may be present in environments connected to deeply subducted slabs and possibly associated with deep carbonatitic melting. Our observations show that Cr strongly partitions in the oxide phase such that the coexisting silicates are depleted in Cr compared to Fe 3 O 4 -free assemblages. This may significantly affect the chemical signature of melts produced in the deep mantle.

Garnet is the second most abundant mineral phase in the upper mantle and transition zone settings. The crystal structure of garnet is quite flexible and hence it is able to accommodate various cations, including large incompatible cation such as sodium. We used first-principles simulation based on density functional theory and two widely used approximations-local density approximation (LDA) and generalized gradient approximation (GGA)-to explore the crystal structure, equation of state, and elasticity of sodium bearing majorite garnet with Na 2 MgSi 5 O 12 stoichiometry at pressures relevant to the upper mantle and transition zone. We find that the pressure-volume results based on LDA can be explained by a Birch Murnaghan finite strain equation of state with V0 = 1447.6 (± 0.1) Å 3 , K0 = 177.4 (± 0.4) GPa, and K′0 = 3.93 (± 0.02). The results based on GGA can be explained by a Birch Murnaghan finite strain equation of state with V0 = 1525.8 (± 0.2) Å 3 , K0 = 160.2 (± 0.4) GPa, and K′0 = 3.96 (± 0.02). The full elastic moduli tensor for Na-majorite with tetragonal symmetry exhibits slight deviation from the cubic symmetry with C 11 < C 33 , C 12 ~ C 13 , and C 44 ~ C 66 . The magnitude of the tetragonal strain also captures the slight deviation from the cubic symmetry. At pressures corresponding to the upper mantle and mantle transition zone, the compressional wave velocity, v P , and shear wave velocity, v S , for the Na-majorite garnet are fast compared to a wide variety of garnets such as pyrope, grossular, almandine, and majorite garnet. Although single-crystal anisotropy of Na-majorite is greater than pyrope, it is still low compared to the major mantle phases.

The results of a combined electron probe microanalysis, single-crystal X-ray diffraction, and Fourier transform infrared study of a crystal of armstrongite from Khan Bogdo deposit (Gobi, Mongolia) are reported. Major element analysis provided (wt%): CaO 9.2 (1), ZrO 2 20.9(2), and SiO 2 62.5(2). Significant concentrations of REE (0.45 wt%) were also detected. From single-crystal structural refinement, armstrongite resulted monoclinic [space group C2/m, a = 14.0178(7), b = 14.1289(6), c = 7.8366(3) Å, b = 109.436(3)°, V = 1463.6(1) Å 3 , Z = 4] and twinned with two individuals rotated around a twin twofold axis parallel to [100]. The analyzed crystal was refined up to R = 3.3% (Rw = 2.9%). The structural refinement showed that the investigated armstrongite has only two water groups per formula unit consistent with the infrared analysis. Indeed, the occurrence in the infrared spectrum of the armstrongite (here reported for the first time) of two bending vibration bands at about 1640 and 1610 cm -1 testifies to the presence of two water groups environments. The results of this integrated approach converged to the following empirical formula (based on Si = 6 atoms per formula unit): (Ca 0.96 Ce 0.01 Yb 0.01 )Zr 0.99 Si 6 O 14.97 · 2.02 H 2 O. Finally, the studied mineral shows a framework density (FD = 21.86) lying in the range of zeolites and microporous heterosilicates with tetrahedral-octahedral frameworks. The determined crystal chemical features are relevant for the possible employment of this mineral or of its synthetic analogs for technological applications.

Steinhardtite is a new mineral from the Khatyrka meteorite; it is a new allotropic form of aluminum. It occurs as rare crystals up to ~10 μm across in meteoritic fragments that contain evidence of a heterogeneous distribution of pressures and temperatures during impact shock, in which some portions of the meteorite reached at least 5 GPa and 1200 °C. The meteorite fragments contain the high-pressure phases ahrensite, coesite, stishovite, and an unnamed spinelloid with composition Fe 3-x Si x O 4 (x ≈ 0.4). Other minerals include trevorite, Ni-Al-Mg-Fe spinels, magnetite, diopside, forsterite, clinoenstatite, nepheline, pentlandite, Cu-bearing troilite, icosahedrite, khatyrkite, cupalite, taenite, and Al-bearing taenite. Given the exceedingly small grain size of steinhardtite, it was not possible to determine most of the physical properties for the mineral. A mean of 9 electron microprobe analyses (obtained from two different fragments) gave the formula Al0.38Ni0.32Fe0.30, on the basis of 1 atom. A combined TEM and single-crystal X‑ray diffraction study revealed steinhardtite to be cubic, space group Im3m, with a = 3.0214(8) Å, and V = 27.58(2) Å 3 , Z = 2. In the crystal structure [R1 = 0.0254], the three elements are disordered at the origin of the unit cell in a body-centered-cubic packing (α-Fe structure type). The five strongest powder-diffraction lines [d in Å (I/I0) (hkl)] are: 2.1355 (100) (110); 1.5100 (15) (200); 1.2329 (25) (211); 0.9550 (10) (310); 0.8071 (30) (321). The new mineral has been approved by the IMA-NMNC Commission (2014-036) and named in honor of Paul J. Steinhardt, Professor at the Department of Physics of Princeton University, for his extraordinary and enthusiastic dedication to the study of the mineralogy of the Khatyrka meteorite, a unique CV3 carbonaceous chondrite containing the first natural quasicrystalline phase icosahedrite. The recovery of the polymorph of Al described here that contains essential amounts of Ni and Fe suggests that Al could be a contributing candidate for the anomalously low density of the Earth’s presumed Fe-Ni core.