Richard L. Hervig

Trace element partitioning between amphibole, phlogopite, and basanite melt

Abstract We have measured amphibole-melt and phlogopite-melt partition coefficients ( D ) for 22 trace elements in experimentally crystallized natural basanites with the ion microprobe. The synthesized phases display an exceptional degree of homogeneity for both major and trace elements, as demonstrated by the ratio of the standard deviation to the mean counting statistics uncertainty of the measurements. In pargasitic hornblende, actinides are highly incompatible ( D = 0.001), LILE and HFSE are mildly incompatible ( D = 0.04 – 0.2and0.1 – 0.2, respectively), and REE partition coefficients vary from 0.05 to 0.6, with a maximum near Ho. Except for the LILE ( D = 0.1 – 3.7), phlogopite partition coefficients are generally lower, especially the REE ( D ≈ 0.01). The partitioning results are consistent with a model in which the variation in partition coefficient with ionic radius results from the crystal lattice strain induced by the size misfit of the substituting trace element. This result predicts a decrease in Youngs Modulus ( E ) with increasing size of the cation sites in the crystal lattice, and E derived for the largest site in both amphibole and phlogopite agree well with experimentally determined bulk mineral values. The ability to model partitioning with an elastic strain model provides an important link between trace element partitioning and the macroscopic properties of minerals. Relative to an anhydrous peridotite, partial melting of an amphibole or phlogopite bearing peridotite will result in no Th-U fractionation, slight LILE depletions, and, aside from Ti, no significant HFSE depletions. Thus, barring the addition of any slab components besides H 2 O, partial melting of hydrated peridotite is not a plausible explanation for any of the geochemical features commonly associated with subduction zone magmas.

The pressure and temperature dependence of carbon dioxide solubility in tholeiitic basalt melts

The solubility of carbon dioxide in tholeiitic melt (1921 Kilauea basalt ) was determined under experimental conditions of 1 kbar, 1200°C; 10 and 15 kbar and 1300–1600°C. We examined the solubility at pressure and temperature conditions intermediate to those reported in previous studies, and, in particular, we addressed the effect of temperature on carbon dioxide solubility. Two different carbon sources were used in the experiments, silver oxalate and a mixture of carbonate minerals, to examine the effects of dissolved silver on carbon dioxide solubility. Three analytical methods were employed to measure accurately and precisely the dissolved carbon in the run products: ( 1 ) Fourier transform micro-infrared spectroscopy, ( 2 ) secondary ion mass spectrometry, and ( 3 ) bulk carbon analysis with a Perkin Elmer Elemental Analyzer. The first two methods are micro-beam techniques which allowed for assessment of sample homogeneity. Consistent with previous solubility studies, infrared analyses showed that carbon is dissolved in basaltic melt in the form of carbonate. However, our experimental results differ from the previous solubility study in that we demonstrate carbon dioxide solubility is temperature independent. At 1 kbar and 1200°C, carbon dioxide solubility is 543 ppm; at 10 kbar and 1300, 1400, and 1500°C, carbon dioxide solubility is approximately 0.77 ± .07 wt%; and at 15 kbar and 1400, 1450, 1500, 1550, and 1600°C, the solubility is approximately 1.21 ± .13 wt%. Dissolved silver does not appear to affect the solubility. These results invalidate previous models for carbon dioxide solubility. We have developed a new model which describes the pressure and temperature dependence of carbon dioxide solubility for tholeiitic basalts. Regression of the solubility data for the reaction CO2vapor + O2−melt = CO32−melt gives a heat of solution (ΔH0 at 1 kbar and 1473 K) of 5.20 ± 4.30 kJ/mol and the change in partial molar volume ΔV0[CO32−melt− O2−melt of 23.14 ± 1.03 cm3/mol. Application of this model suggests that fluid-saturated partial melting of the MORB source region cannot be supported.

High-resolution 23Na, 27Al and 29Si NMR spectroscopy of framework Aluminosilicate glasses

The results of high-resolution 23Na, 27Al and 29Si NMR spectroscopy of aluminosilicate glasses with MO/Al2O3 or L2OAl2O3 = 1 (M = +2, L = +1 cations) shows that these glasses have a fully polymerized tetrahedral framework structure with only a defect level of non-bridging oxygens. The chemical shifts of the peak maxima for all three nuclides become less shielded (less negative or more positive) with decreasing Si/ (Si + Al). For 29Si and 27Al, this variation parallels the variation of framework crystalline silicates. The 23Na chemicAl shifts for the glasses becomes less shielded (less negative) with decreasing Na/(Na + K), opposite to the trend for crystalline alkali feldspars. Few data exist for the 23Na chemical shifts of crystalline samples, and the structural causes of variation in 23Na chemical shifts are not well understood. The 27Al and 29Si chemical shifts of the glasses do not vary significantly with different large (modifier) cations. The 29Si chemical shifts provide estimates of average Si-O-T (T = Si, Al)bond angles and Si-O bond distances and the 27AL and 29Si chemical shifts and peak breadths are consistent with a decrease in the tetrahedralring order (number of tetrahedra per ring) with decreasing Si/(Si + Al). The data presented here for fully polymerized glasses form a base from which the data for aluminosilicate glasses containing both fully polymerized sites and less polymerized sites can be interpreted.

Vapor-undersaturated experiments with Macusani glass+H2O at 200 MPa, and the internal differentiation of granitic pegmatites

Vapor-undersaturated fractional crystallization experiments with Macusani glass (macusanite), a peraluminous rhyolite obsidian, at 200 MPa yield mineralogical fabrics and zonation, and melt fractionation trends that closely resemble those found in zoned granitic pegmatites and other granitoids of comparable composition (typically peraluminous, Li-Be-Ta-rich deposits). The zonation from the edge of charges inward is characterized by: (1) fine-grained sodic feldspar-quartz border zones; (2) a fringe of very coarse-grained graphic quartz-feldspar intergrowths that flair radially toward melt and terminate with nearly monophase K-feldspar; (3) cores of very coarse-grained, nearly monominerallic quartz or virgilite (LiAlSi5O12)±mica; and (4) late-stage, fine-grained albite+mica intergrowths that are deposited from alkaline, Na-rich interstitial melt at vapor saturation. Similar experimental products have been observed in compositionally simpler, less evolved systems. Liquid lines of descent from initially H2O-undersaturated runs are marked by a decrease in SiO2, and increases in Na/K, B, P, F, H2O, and a variety of trace lithophile cations. These trends are believed to be governed by three factors: (1) disequilibrium growth of feldspars (±quartz) via metastable supersaturation; (2) fractionation of melt toward SiO2-depleted, Na-rich compositions due to increases in B, P, and F; and (3) changes in nucleation and growth rates, mostly as a function of the H2O content of melt (Xwm). In contrast, experiments that are cooled below the liquidus from the field of melt+aqueous vapor (London et al. 1988) fail to replicate pegmatitic characteristics in most respects. On the basis of these and other experiments, we suggest that the formation of pegmatite fabrics stems primarily from fractional crystallization in volatile-rich melts, and that enrichments in normally trace lithophile elements result from melt differentiation trends toward increasingly alkaline, silica-depleted compositions. Although vapor saturation at near-solidus and subsolidus conditions may promote extensive recrystallization, an aqueous vapor phase does not appear to be necessary for the generation of most of the salient characteristics of pegmatites.

BORON ISOTOPIC COMPOSITION OF SUBDUCTION-ZONE METAMORPHIC ROCKS

Abstract Many arc lavas contain material derived from subducted oceanic crust and sediments, but it remains unresolved whether this distinctive geochemical signature is transferred from the subducting slab by aqueous fluids, silicate melts, or both. Boron isotopic measurements have the potential to distinguish between slab transfer mechanisms because 11 B fractionates preferentially into aqueous fluids whereas little fractionation may occur during partial melting. Previous studies have shown that δ 11 B values of island arc lavas (−6 to +7) overlap the range of δ 11 B values for altered oceanic crust (−5 to +25) and pelagic sediments and turbidites (−7 to +11). Secondary ion mass spectrometry (SIMS) analyses of minerals in subduction-zone metamorphic rocks yield δ 11 B =−11 to −3 suggesting that slab dehydration reactions significantly lower the δ 11 B values of subducted oceanic crust and sediments. In order to explain the higher δ 11 B values reported for arc lavas as compared to subduction-zone metamorphic rocks, the B-bearing component derived from the metamorphosed slab must be enriched in 11 B relative to the slab, favoring an aqueous fluid as the slab transfer mechanism.

Experimental evidence for a hydrous transition zone in the early Earth's mantle

Abstract Partition coefficients of H2O between β and γ phases of olivine stoichiometry and coexisting ultra mafic melt have been estimated to be > 0.1 ± 0.04 (1σ) and 0.04, respectively; based on experiments at 15–16.5 GPa, 1300–1500°C in a hydrous KLB-1 peridotite system. The high H2O contents of β (1.5–3 wt%) and γ phases (0.7 wt%) would form a reservoir for H2O after cooling and crystallization of a hydrous magma ocean. Subsequent upwelling of this hydrous mantle would release H2O at the β phase-olivine boundary near 400 km depth, inducing partial melting of the peridotite to produce hydrous ultramafic magma. Most subducting hydrous minerals dehydrate at pressures shallower than 6.5 GPa if the down-dragged hydrous peridotite follows a P-T path hotter than 900°C at 8 GPa and cannot re-hydrate the transition zone. Therefore, the above proposed partial melting would gradually deplete the H2O reservoir, which is consistent with the decrease in the activity of ultramafic magmatism and the apparent degree of melting of komatiites from the Archean to the Mesozoic.

Experiments on crystal/liquid partitioning of Ru, Rh and Pd for magnetite and hematite solid solutions crystallized from silicate melt☆

Experiments to characterize crystal/melt partitioning of Ru, Rh and Pd between Fe-oxides (magnetite and hematite solid solutions) and silicate melt are reported for oxygen fugacities imposed by CO2 decomposition. Oxides were equilibrated near 1275°C at 1 atm with silica-saturated melts in the compositional system “FeO”CaOAl2O3SiO2±MgO±Cr2O3±TiO2. Ru and Rh are strongly compatible while Pd is slightly incompatible. Partition coefficients (elemental weight ratios) for Ru are variable but range from 100 to > 4000 depending, in a poorly defined way, on the oxide crystal composition. We report enhanced compatibility for Ru compared to previous work for spinels in an Fe-free system. Rh compatibility is also enhanced compared to the Fe-free system, but to a lesser degree. However, Rh partition coefficients (grand average is 250±120) are more uniform than those measured for Ru. Pd is slightly incompatible in magnetite and hematite (grand average is 0.7 ± 0.3) in contrast to the Fe-free system where Pd is highly incompatible. The differences in compatibility among Ru, Rh and Pd suggest that spinels could play a role in determining platinum-group element (PGE) fractionation trends, in particular, for rocks crystallizing at high oxygen fugacity where it is most likely that the dissolved PGEs are present as oxidized species. We also propose a reaction mechanism to explain the common observation of platinum-group minerals (PGMs) included within spinel phases. Finally, our data suggest that Ru, at least, should be concentrated in spinels crystallizing in Earths atmosphere, either within micrometeroids, or from atmospheric crystallization of impact ejecta.

Boron isotope geochemistry during diagenesis; Part I, Experimental determination of fractionation during illitization of smectite

Experiments were performed to measure the isotopic fractionation of boron between illite/smectite (I/S) clay minerals and water as a function of temperature (300° and 350°C) and degree of illitization. Corresponding changes in the oxygen isotopes were monitored as an indication of the approach to equilibrium. The kinetics of the B-isotope exchange follows the mineralogical restructuring of smectite as it recrystallizes to illite. An initial decline in δ11BI/S occurs when the I/S is randomly ordered (RO). The δ11BI/S values reach a plateau during R1 ordering of the I/S, representing a metastable condition. The greatest change in δ11BI/S is observed during long-range (R3) ordering of the I/S when neoformation occurs. Values of δ11BI/S measured on the equilibrium reaction products were used to construct a B-isotope fractionation curve. There is a linear correlation among data from these experiments and 1100°C basaltic melt-fluid fractionation experiments (Hervig and Moore, 2000) that can be extrapolated to include adsorption experiments at 25°C (Palmer et al., 1987). Unlike other stable isotopic systems (e.g., oxygen) there is no mineral-specific fractionation of B-isotopes, but rather a coordination dependence of the fractionation. Under diagenetic conditions B is predominantly in trigonal coordination in fluids but substitutes in tetrahedral sites of silicates. The preference of 10B for tetrahedral bonds is the major fractionating factor of B in silicates.

Isotopic and elemental partitioning of boron between hydrous fluid and silicate melt

Abstract The fractionation of B and its isotopes between aqueous fluid and silicate melt has been studied from 550 to 1100 °C and 100-500 MPa. Fluid-melt partition coefficients are <1 for basaltic melt and >1 for rhyolite melt. This shows that B is not always strongly extracted from melts into hydrous fluids. Boron isotopic fractionation is large compared with the carbon and oxygen stable isotopic systems (especially at high T) and is most simply explained by differences in coordination (trigonal vs. tetrahedral) among coexisting phases. Combined with earlier measurements on illite-water (300- 350 °C), B isotopic fractionation defines a temperature-dependent trend from 300 to 1100 °C. Because of the large magnitude and apparent low sensitivity to bulk composition, B isotopic fractionation can be readily applied to studies of diagenesis, hydrothermal alteration of planetary bodies, subduction- zone processing and arc magma generation, and magma chamber evolution.

Los Angeles: The Most Differentiated Basaltic Martian Meteorite

Los Angeles is a new martian meteorite that expands the compositional range of basaltic shergottites. Compared to Shergotty, Zagami, QUE94201, and EET79001-B, Los Angeles is more differentiated, with higher concentrations of incompatible elements (e.g., La) and a higher abundance of late-stage phases such as phosphates and K-rich feldspathic glass. The pyroxene crystallization trend starts at compositions more ferroan than in other martian basalts. Trace elements indicate a greater similarity to Shergotty and Zagami than to QUE94201 or EET79001-B, but the Mg/Fe ratio is low even compared to postulated parent melts of Shergotty and Zagami. Pyroxene in Los Angeles has 0.7–4-µm-thick exsolution lamellae, ∼10 times thicker than those in Shergotty and Zagami. Opaque oxide compositions suggest a low equilibration temperature at an oxygen fugacity near the fayalite-magnetite-quartz buffer. Los Angeles cooled more slowly than Shergotty and Zagami. Slow cooling, coupled with the ferroan bulk composition, produced abundant fine-grained intergrowths of fayalite, hedenbergite, and silica, by the breakdown of pyroxferroite. Shock effects in Los Angeles include maskelynitized plagioclase, pyroxene with mosaic extinction, and rare fault zones. One such fault ruptured a previously decomposed zone of pyroxferroite. Although highly differentiated, the bulk composition of Los Angeles is not close to the low-Ca/Si composition of the globally wind-stirred soil of Mars.