Solubility and speciation of nitrogen in silicate melts have been investigated between 1400 and 1700 degrees C and at pressures ranging from 10 to 30 kbar for six different binary alkali and alkaline-earth silicate liquids and a Ca-Mg-alumino silicate. Experiments were performed in a piston-cylinder apparatus. The nitrogen source is silver azide, which breaks down to Ag and molecular N(2) below 300 degrees C. At high pressure and temperature, the nitrogen content may be as high as 0.7 wt% depending on the melt composition, pressure, and temperature. It increases with T, P and the polymerization state of the liquid. Characterization by Raman spectroscopy and (15)N solid state MAS NMR indicates that nitrogen is not only physically dissolved as N(2) within the melt structure like noble gases, but a fraction of nitrogen interacts strongly with the silicate network. The most likely nitrogen-bearing species that can account for Raman and NMR results is nitrosyl group. Solubility data follow an apparent Henry's law behavior and are in good agreement with previous studies when the nitrosyl content is low. On the other hand, a significant departure from a Henry's law behavior is observed for highly depolymerized melts, which contain more nitrosyl than polymerized melts. Possible solubility mechanisms are also discussed. Finally, a multi-variant empirical relation is given to predict the relative content of nitrosyl and molecular nitrogen as a function of P, T, and melt composition and structure. This complex speciation of nitrogen in melts under high pressure may have significant implication concerning crystal-melt partitioning of nitrogen as well as for potential elemental and isotopic fractionation of nitrogen in the deep Earth. (c) 2006 Elsevier Inc. All rights reserved.

Olivine/melt partitioning of Sigma Fe, Fe2+, Mg2+, Ca2+, Mn2+, Co2+, and Ni2+ has been determined in the systems CaO-MgO-FeO-Fe2O3-SiO2 (FD) and CaO-MgO-FeO-Fe2O3-Al2O3-SiO2 (FDA3) as a function of oxygen fugacity (f(O2)) at 0.1 MPa pressure. Total iron oxide content of the starting materials was similar to 20 wt%. The f(O2) was to used to control the Fe3+/Sigma Fe (Sigma Fe: total iron) of the melts. The Fe3+/Sigma Fe and structural roles of Fe2+ and Fe3+ were determined with Fe-57 resonant absorption Mossbauer spectroscopy. Changes in melt polymerization, NBO/T, as a function of f(O2) was estimated from the M6ssbauer data and existing melt structure information. It varies by similar to 100% in melts coexisting with olivine in the FDA3 system and by about 300% in the FD system in the Fe3+/Sigma Fe range of the experiments (0.805-0.092). The partition coefficients (D-i(ol-melt) = wt% in olivine/wt% in melt) are systematic functions of f(O2) and, therefore, NBO/T of the melt. There is a D-i(ol-melt)-minimum in the FDA3 system at NBO/T-values corresponding to intermediate Fe3+/Sigma Fe (0.34-0.44). In the Al-free system, FD, where the NBO/T values of melts range between similar to 1 and similar to 2.9, the partition coefficients are positively correlated with NBO/T (decreasing Fe3+/Sigma Fe). These relationships are explained by consideration of solution behavior in the melts governed by Q''-unit distribution and structural changes of the divalent cations in the melts (coordination number, complexing with Fe3+, and distortion of the polyhedra). (c) 2006 Elsevier Inc. All rights reserved.

Fractionation of Fe isotopes between a silicate melt and metallic alloys has been quantified experimentally at 1500 degrees C. The effects of oxygen fugacity and run duration have been investigated to distinguish between kinetic and equilibrium fractionations. A new experimental, setup is presented, in which metallic Fe is produced by reduction of oxidized iron from a silicate melt due to a change of redox conditions. This metallic Fe is physically removed from the silicate and sequestered as a (Pt,Fe) alloy. Bulk analyses of the silicate and metallic fractions using multi-collector ICP-MS methods are coupled with in situ analyses using an ion microprobe.

Solubility and solution mechanisms of H2O in depolymerized melts in the system Na2O-Al2O3-SiO2 were deduced from spectroscopic data of glasses quenched from melts at 1100 degrees C at 0.8-2.0 GPa. Data were obtained along a join with fixed nominal NBO/T = 0.5 of the anhydrous materials [Na2Si4O9-Na-2(NaAl)(4)O-9] with Al/(Al+Si) = 0.00-0.25. The H2O solubility was fitted to the expression, X-H20 = 0.20 + 0.0020f(H2O) - 0.7X(Al) + 0.9(X-Al)(2), where X-H2O is the mole fraction of H2O (calculated with O = 1), f(H2O) the fugacity of H2O, and X-Al = Al/(Al+Si). Partial molar volume of H2O in the melts, V-H2O(melt), calculated from the H2O-solulbility data assuming ideal mixing of melt-H2O solutions, is 12.5 cm(3)/mol for Al-free melts and decreases linearly to 8.9 cm(3)/mol for melts with Al/(Al+Si) similar to 0.25. However, if recent suggestion that V-H2O(melt) is composition-independent is applied to constrain activity-composition relations of the hydrous melts, the activity coefficient of H2O, gamma(melt)(H2O), increases with Al/(Al+Si).

Peralkaline and peraluminous glasses close to the metaluminous join in the systems Na2O-Al2O3-SiO2 (NAS), CaO-Al2O3-SiO2 (CAS), and MgO-Al2O3-SiO2 (MAS) have been examined with Raman spectroscopy. At least three different SiO2 contents in each of the systems (NAS, CAS, and MAS) have been studied. Each series of glasses spans the metaluminous join at constant silica content. The spectra of glasses in all three systems show changes consistent with a continuous decrease in abundance of depolymerized species and an increase in fully polymerized species as compositions change from peralkaline to peraluminous. There is no evidence for maxima or minima of these abundances across the metaluminous joins for any of the studied series. These observations confirm previous suggestions that non-bridging 0 atoms are a general feature of "fully polymerized" glasses, and that a population of Al exists in the melt structure that is not associated with a charge-balancing cation, even in peralkaline compositions.

Olivine/melt partitioning of the transition metal cations, Fe2+, Mn2+, Co2+, and Ni2+, together with Mg2+ and Ca2+, has been examined experimentally as a function of melt composition at ambient pressure. Melt structure was inferred from bulk-chemical composition, existing structural data, and Fe-57 resonant absorption Mossbauer spectroscopy. Under isothermal conditions, K-D(i-Mg)(olivine/melt) = (C-i/C-Mg)(olivine)(C-i/ C-Mg)(melt) , is an exponential function of melt NBO/T for i = Ca2+, Mn2+, Co2+, and Ni2+. For i = Fe2+, the relationship is parabolic with maximum K-D(Fe2+-Mg)(olivine/melt)-values at NBO/T near 1. At constant melt NBO/T, K-D(i-Mg)(olivine/melt) increases systematically with decreasing cation radius, an effect that is more pronounced the more polymerized the melt. The K-D(i-Mg)(olivine/melt) is also a positive and linear function of Na/(Na + Ca) of Al-free melts. This latter effect results from changes in Q(n)-species abundance governed by Na/(Na + Ca) of the melts. The enthalpy of the exchange equilibrium, i(olivine) + Mg-melt = i(melt) + Mg-olivine, derived from the temperature-dependence of K-D(i-Mg)(olivine/melt), is also a positive function of the ionic radius of the, i-mg) cation. The relationship of enthalpy to melt polymerization also depends on cation radius. The K-D(Fe2+-Mg)(olivine/melt) does not, however, follow this trend possibly because the bond distance, d(Fe2+-O), in the melts depends on melt composition.

We report a comprehensive imaging study including confocal microRaman spectroscopy, scanning electron microscopy (SEM), and 3-D extended focal imaging light microscopy of carbonate globules throughout a depth profile of the Martian meteorite Allan Hills (ALH) 84001 and similar objects in mantle peridotite xenoliths from the Bockfjorden volcanic complex (BVC), Svalbard. Carbonate and iron oxide zoning in ALH 84001 is similar to that seen in BVC globules. Hematite appears to be present in all ALH 84001 carbonate-bearing assemblages except within a magnesite outer rim found in some globules. Macromolecular carbon (MMC) was found in intimate association with magnetite in both ALH 84001 and BVC carbonates. The MMC synthesis mechanism appears similar to established reactions within the Fe-C-O system. By inference to a terrestrial analogue of mantle origin (BVC), these results appear to represent the first measurements of the products of an abiotic MMC synthesis mechanism in Martian samples. Furthermore, the ubiquitous but heterogeneous distribution of hematite throughout carbonate globules in ALH 84001 may be partly responsible for some of the wide range in measured oxygen isotopes reported in previous studies. Using BVC carbonates as a suitable analogue, we postulate that a low temperature hydrothermal model of ALH 84001 globule formation is most likely, although alteration (decarbonation) of a subset of globules possibly occurred during a later impact event.

The influence on olivine/melt transition metal (Mn, Co, Ni) partitioning of Al3+ double left right arrow Si4+ substitution in the tetrahedral network of silicate melt structure has been examined at ambient pressure in the 1450-1550 degrees C temperature range. Experiments were conducted in the systems NaAlSiO4-Mg2SiO4-SiO2 and CaAl2Si2O8-Mg2SiO4-SiO2 with about 1 wt% each of MnO, CoO, and NiO added. These compositions were used to evaluate how, in silicate melts, Al3+ double left right arrow Si4+ substitution and ionization potential of charge-balancing cations affect activity-composition relations in silicate melts and mineral/melt partitioning.