Tracer diffusion of some alkali,alkaline-earth and transition element ions in a basaltic and an andesitic melt,and the implications concerning melt structure

The diffusion properties of Na, Cs, Sr, Ba, Co, Mn, Fe and Sc ions in a basaltic and an andesitic melt have been determined experimentally using the radiotracer residual-activity method, and narrow platinum capillaries, over the temperature range 1,300–1,400° C. Diffusion of all cations follows an Arrhenius relationship; the values of the activation energies range from 24 kcal mol^–1 for Na to 67 kcal mol^–1 for Co in the andesitic melt, and from 39 kcal mol^–1 for Na to 65 kcal mol^–1 for Cs in the basaltic melt. Relative diffusivities in the basaltic melt, but not in the andesitic melt, correlate with assumed ionic radii values. Each cation, except Na⁺, diffuses faster in the basaltic melt than in the andesitic melt over the studied temperature range. Sodium shows similar diffusivity in the two melts.Compensation diagrams incorporating new and some previously-published data indicate that Cs probably diffuses by different mechanisms in different silicate glass and melt systems. Iron has a relatively high activation energy which is consistent with its part occupancy of tetrahedral co-ordination polyhedra.     

Oxygen diffusion in basalt and andesite melts: experimental results and discussion of chemical versus tracer diffusion

Chemical diffusion coefficients for oxygen in melts of Columbia River basalt (Ice Harbor Dam flow) and Mt. Hood andesite have been determined at 1 atm. The diffusion model is that of sorption or desorption of oxygen into a sphere of uniform initial concentration from a constant and semi-infinite atmosphere. The experimental design utilizes a thermogravimetric balance to monitor the rate of weight change arising from the response of the sample redox state to an imposed fO₂. Oxygen diffusion coefficients are approximately an order-ofmagnitude greater for basaltic melt than for andesitic melt. At 1260° C, the oxygen diffusion coefficients are: D=1.65×10^–6cm²/s and D=1.43×10^–7cm²/s for the basalt and andesite melts, respectively. The high oxygen diffusivity in basaltic melt correlates with a high ratio of nonbridging oxygen/tetrahedrally coordinated cations, low melt viscosity, and high contents of network-modifying cations. The dependence of the oxygen diffusion coefficient on temperature is: D=36.4exp(–51,600±3200/RT)cm²/s for the basalt and D=52.5exp(–60,060±4900/RT)cm²/s for the andesite (R in cal/deg-mol; T in Kelvin). Diffusion coefficients are independent of the direction of oxygen diffusion (equilibrium can be approached from extremely oxidizing or reducing conditions) and thus, melt redox state. Characteristic diffusion distances for oxygen at 1260° C vary from 10^-2 to 10² m over the time interval of 1 to 10⁶ years. A compensation diagram shows two distinct trends for oxygen chemical diffusion and oxygen tracer diffusion. These different linear relationships are interpreted as supporting distinct oxygen transport mechanisms. Because oxygen chemical diffusivities are generally greater than tracer diffusivities and their Arrhenius activation energies are less, transport mechanisms involving either molecular oxygen or vacancy diffusion are favored.     

Diffusion and partition coefficients of minor and trace elements in San Carlos olivine at 1,300°C with some geochemical implications

Lattice diffusion coefficients have been determined for 19 elements (Li, Be, Na, Ca, Sc, Ti, V, Cr, Mn, Fe, Co, Ni, Zn, Y, Zr, Eu, Gd, Lu and Hf) in a single crystal of San Carlos olivine as a function of crystallographic orientation, at 1,300°C, 1 bar and fO₂ = 10^−8.3 bars, by equilibration with a synthetic silicate melt. Results for Li, Na, V, Cr, Fe and Zn are from diffusion of these elements out of the olivine, starting from their indigenous concentrations; those for all other elements are from diffusion into the olivine, from the silicate melt reservoir. Our 25-day experiment produced diffusion profiles 50 to > 700 μm in length, which are sufficiently long that precise analyses could be achieved by scanning laser ablation inductively coupled plasma mass spectrometry, even at concentration levels well below 1 μg g⁻¹. For the divalent cations Ca, Mn, Fe and Ni, profiles were also obtained by electron microprobe analysis. The results of the two methods agree well with each other, and are consistent with divalent cation diffusion coefficients previously determined using different experimental methodologies. Olivine/melt partition coefficients retrieved from the data are also consistent with other published partitioning data, indicating that element incorporation and transport in olivine in our experiment occurred via mechanisms appropriate to natural conditions. Most of the examined trace elements diffuse through olivine at similar rates to the major octahedral cations Fe and Mg, showing that cation charge and radius have little direct influence on diffusion rates. Aluminium and P remain low and constant in the olivine, implying negligible transport at our analytical scale, hence Al and P diffusion rates that are at least two orders of magnitude slower than the other cations studied here. All determined element diffusivities are anisotropic, with diffusion fastest along the [001] axis, except Y and the REEs, which diffuse isotropically. The results suggest that element diffusivity in olivine is largely controlled by cation site preference, charge balance mechanisms and point-defect concentrations. Elements that are present on multiple cation sites in olivine (e.g. Be and Ti) and trivalent elements that are charge-balanced by octahedral site vacancies tend to diffuse at relatively fast rates.     

Oxygen chemical diffusion in three basaltic liquids at elevated temperatures and pressures

The diffusivity of oxygen has been measured in three basaltic liquids from 1280 to 1450°C and 4 to 21 kilobars using a solid media piston-cylinder apparatus. The measurements were done by monitoring the reduction of ferric iron in previously oxidized spheres of basalt melt. The compositions studied were olivine nephelinite, alkali basalt, and 1921 Kilauea tholeiite.The isobaric temperature dependence of oxygen diffusivity is adequately described by Arrhenius relationships for the three liquids studied. Arrhenius activation energies were determined at 12 kilobars for olivine nephelinite (62± 6 kcal/mole) and tholeiite (51 ± 4 kcal/mole) and at 4, 12, and 20 kilobars for alkali basalt (70 ± 7, 86 ± 6, and 71 ± 14 kcal/mole, respectively). The Arrhenius parameters for the three compositions define a compensation law which is indistinguishable from those for oxygen diffusion in simple silicate melts (DUNN, 1982) and for divalent cation diffusion in basaltic melts (Hofmann, 1980). These results suggest that the principal species contributing to the total diffusivity of oxygen is the oxide anion (O^2?).The isothermal pressure dependence of oxygen diffusion is complex and quite different from that observed for cationic diffusion in silicate melts. All three compositions show a sharp decrease in oxygen diffusivity at approximately the same pressure as the change in the liquidus phase from olivine to pyroxene, but otherwise the pressure dependence can be described by Arrhenius type equations. The equations yield negative activation volumes for the olivine nehpelinite and the alkali basalt. The activation volumes determined for the tholeiite are near zero at low pressure and positive at high pressure. A negative activation volume represents a decrease in the average size of the principal diffusing species.The results of this study are consistent with a melt model which includes both continuous changes in the relative proportions of the various anionic species in the melt with pressure and the occurrence of anionic disproportionation reactions within narrow pressure ranges.     

长白山天池老虎洞期火山活动地质特征及成因意义

长白山天池火山老虎洞期火山活动发生在更新世晚期白头山组碱性粗面岩喷发之后,火山活动的产物主要为玄武岩质火山碎屑岩和少量玄武岩质或粗面岩质熔岩;老虎洞组火山岩的稀土元素地球化学特征介于早期玄武岩和气象站组碱流岩两者之间,将二者有机地联系在一起,使整个天池火山岩的演化趋势更加清晰。老虎洞组火山岩的存在充分证明了天池火山的粗面岩类与该区早期的大量玄武岩具有成因联系。长白山天池火山活动的成因并非简单地用西太平洋板块的俯冲作用所能解释的。     

Diffusive equilibration between hydrous metaluminous-peraluminous haplogranite liquid couples at 200 MPa (H<Subscript>2</Subscript>O) and alkali transport in granitic liquids

This study examines the systematics and rate of alkali transport in haplogranite diffusion couples in which a chemical potential gradient in Al is established between near water-saturated metaluminous and peraluminous liquids that differ only in their initial content of normative corundum. At 800°C, measurable chemical diffusion of alkalis occurs throughout the entire length (∼1 cm) of the diffusion couples in 2–6 h, indicating long range diffusive communication through melt. Alkali transport results in homogenization of initially different Na/Al and ASI [=mol. Al₂O₃/(CaO + Na₂O + K₂O)] throughout the couples within ∼24 h, whereas initially homogenous K* evolves to become uniformly different between metaluminous and peraluminous ends. Calculated effective binary diffusion coefficients for alkalis in experiments that do not significantly violate the requirement of a semi-infinite chemical reservoir (0- to 2-h duration at 800°C) are similar to those observed in previous studies: in the range of (1–8) × 10⁻¹² m²/s. Such a magnitude of diffusivity, however, is inadequate to account for the observed changes of alkali concentrations and molecular ratios throughout the couples in 2- to 6-h experiments. The latter changes are consistent with diffusivities estimated via the x ² = Dt approximation, which yields effective values around 10⁻⁹ m²/s. These observations suggest that Fick’s law alone does not adequately describe the diffusive transport of alkalis in granitic liquids. In addition to simple ionic diffusion associated with local gradients in concentration or chemical potential of the diffusing component described by Fick’s second law (local diffusion), alkali transport through melt involves system-wide diffusion (field diffusion) driven by chemical potential gradients that also include components with which the alkalis couple or complex (e.g., Al). Field diffusion involves the coordinated migration of essentially all alkali cations, resembling a positive ionic current that drives the system to a metastable state having a minimum energy configuration with respect to alkali distribution. The net result is effective transport rates perhaps three orders of magnitude faster than simple local alkali diffusion, and at least seven to eight orders of magnitude faster than the diffusive equilibration of Al and Si.     

Ion mobility and dynamic disordering of water in analcime

Ions and water molecules mobilities were studied by the ⁷Li, ²³Na, ¹H nuclear magnetic resonance in the framework aluminosilicate mineral analcime and in its cation-exchanged (lithium) form. It has been found that cation exchange as well as dehydration alter radically dynamics of particles interposed in channels of the zeolite framework. The highest mobility of alkali cations with the diffusion activation energy of 0.52 eV was detected in hydrated Li-analcime. The ion diffusion increase with temperature is accompanied by an increase in the dynamic orientational disordering of water sublattice. The energy of formation of structural defects responsible for ion diffusion and water disordering was deduced to be 0.2 eV. As a possible defect, the occupation of water positions by cations is discussed (model of bridging-type ion diffusion).     

Diffusion in fluid-bearing and slightly-melted rocks: experimental and numerical approaches illustrated by iron transport in dunite

Bulk diffusion of iron in synthetic dunites containing 1–6 vol.% fluid or melt at 10 kbar (1 GPa) and 900°–1300° C was examined by encapsulating the samples in platinum, which served as a sink for iron. The rate of iron loss from the dunite was found to depend strongly upon the identity of the fluid, which was varied from CO₂ and H₂O to melts of basaltic and sodium carbonate composition. Carbon dioxide in amounts up to 4 vol.% has no effect upon bulk iron diffusion because it exists in the dunite are isolated pores. The interconnected nature of H₂O, basaltic melt, and carbonate melt, on the other hand, results in marked enhancement of bulk-rock Fe diffusion that is correlated with the diffusivity and solubility of olivine components in the fluid. At 1300° C, 4–5 vol.% of either water or basaltic melt increases the effective bulk diffusivity from the fluidabsent value of 10^-10 cm²/s to 10^-8 cm²/s. A single experiment involving a similar volume fraction of carbonate melt yielded a minimum bulk diffusivity of 10^-7–10^-6 cm²/s. This remarkably high value is attributable to the concurrent high diffusivity and high solubility of olivine components in molten carbonate H₂O has a high diffusivity, estimated at 10^-4 cm²/s in this study, and basaltic melt can dissolve large amounts of olivine, but neither possesses these two qualities in combination. Bulk transport of Fe in dunite containing <2 vol.% of pure H₂O is independent of olivine grain size for samples having an average grain diameter of <10 m to 60 m. This is probably because bulk diffusion specifically in these H₂O-bearing samples is ratelimited by the flux (which is proportional to concentration) of olivine components in the fluid. Given a constant fluid volume fraction, the effect of reducing the grain size is to increase the number of fluid-filled channels, but at the same time to decrease their average aperture, thus keeping constant the cross-sectional area through which the diffusional flux occurs. (Independence of bulk diffusivity from grain size is not anticipated for rocks containing melt, in which the silicate components are much more soluble.) In numerical (finite difference) simulations of selected laboratory experiments, the bulk Fe transport process was modeled as diffusion in fluid-filled tubules of triangular cross-section that are supplied by volume diffusion from contacting olivine grains with which they are in surface equilibrium. Applying a tortuosity factor of 1.7 brings the numerically computed diffusional loss profiles for experiments containing basaltic melt into near-coincidence with the experimentally-determined curves. This success in reproducing the experimental results lends credence to the interpretation of the bulk diffusional loss profiles as composites of gradients due to volume, grain-boundary and fluid-phase diffusion.     

Observations of Li isotopic variations in the Trinity Ophiolite: Evidence for isotopic fractionation by diffusion during mantle melting

The Trinity peridotite (northern CA) contains numerous lithologic sequences consisting of dunite to harzburgite to spinel lherzolite to plagioclase lherzolite. Previous workers have documented geochemical gradients in these sequences consistent with melt-rock reaction processes occurring around dunites, interpreted to reflect conduits for melt ascent. We have undertaken a study of Li isotope compositions of clinopyroxene and some olivine within these sequences using ion probe techniques to test the hypothesis that the geochemical gradients are related to diffusive fluxing of alkali elements into or away from the melt conduit.Results show large variations in ⁷Li/⁶Li occurring in a consistent pattern across three transects from dunite to plagioclase lherzolite within the Trinity peridotite. Specifically, measurements of average δ⁷Li for single thin sections along the traverse reveal a low in δ⁷Li in the harzburgite adjacent to the dunite returning to higher values farther from the dunite with a typical offset of ∼10 per mil in the low δ⁷Li trough. This pattern is consistent with a process whereby Li isotopes are fractionated during diffusion through a melt either from the dunite conduit to the surrounding peridotite, or from the surrounding peridotite into the dunite conduit. The patterns in ⁷Li/⁶Li occur over a length scale similar to the decrease in REE concentration in these same samples. Explaining both the trace element and Li isotopic gradients requires a combined process of alkali diffusion and melt extraction.We develop a numerical model and examine several scenarios of the combined diffusion-extraction process. Using experimentally constrained values for the change in Li diffusion coefficient with isotope mass, large changes in δ⁷Li as a function of distance can be created in year to decade timescales. The addition of the melt extraction term allows larger Li concentration gradients to be developed and thus produces larger isotopic fractionations than diffusion only models. The extraction aspect of the model can also account for the observed decrease in rare earth element concentrations across the transects.     

The fidelity of melt inclusions as records of melt composition

A series of experiments created melt inclusions in plagioclase and pyroxene crystals grown from a basaltic melt at 1,150°C, 1.0 GPa to investigate diffusive fractionation during melt inclusion formation; additionally, P diffusion in a basaltic melt was measured at 1.0 GPa. Melt inclusions and melts within a few 100 microns of plagioclase–melt interfaces were analyzed for comparison with melt compositions far from the crystals. Melt inclusions and melt compositions in the boundary layer close to the crystal–melt interface were similar, but both differ significantly in incompatible element concentrations from melt found greater than approximately 200 microns away from the crystals. The compositional profiles of S, Cl, P, Fe, and Al in the boundary layers were successfully reproduced by a two-step model of rapid crystal growth followed by diffusive relaxation toward equilibrium after termination of crystal growth. Applying this model to investigate possible incompatible element enrichment in natural melt inclusions demonstrated that at growth rates high enough to create the conditions for melt inclusion formation, ∼10⁻⁹–10⁻⁸ m s⁻¹, the concentration of water in the boundary layer near the crystal was similar to that of the bulk melt because of its high diffusion coefficient, but sulfur, with a diffusivity similar to major elements and CO₂, was somewhat enriched in the boundary layer melt, and phosphorus, with its low diffusion coefficient similar to other high-field strength elements and rare earth elements, was significantly enriched. Thus, the concentrations of sulfur and phosphorus in melt inclusions may over-estimate their values in the bulk melt, and other elements with similar diffusion coefficients may also be enriched in melt inclusions relative to the bulk melt. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Diffusion of tetravalent cations in zircon

Diffusion rates for the three tetravalent cations U, Th and Hf have been measured in synthetic zircon. Diffusant sources included oxide powders and ground pre-synthesized silicates. Rutherford backscattering spectrometry (RBS) was used to measure depth profiles. Over the temperature range 1400–1650 °C, the following Arrhenius relations were obtained (diffusion coefficients in m²sec⁻¹): log D _Th = (1.936 ± 0.9820) + (− 792 ± 34 kJ mol⁻¹ /2.303 RT) log D _U = (0.212 ± 2.440) + (− 726 ± 83 kJ mol⁻¹ /2.303 RT) log D _Hf = (3.206 ± 1.592) + (− 812 ± 54 kJ mol⁻¹ /2.303 RT) The data show a systematic increase in diffusivity with decreasing ionic radius (i.e., faster diffusion rates for Hf than for U or Th), a trend also observed in our earlier study of rare earth diffusion in zircon. Diffusive fractionation may be a factor in the Lu-Hf system given the much slower diffusion rates of tetravalent cations when compared with the trivalent rare earths. The very slow diffusion rates measured for these tetravalent cations suggest that they are essentially immobile under most geologic conditions, permitting the preservation of fine-scale chemical zoning and isotopic signatures of inherited cores. Received: 12 July 1996 / Accepted: 2 December 1996     

Chemical diffusion of fluorine in jadeite melt at high pressure

The chemical diffusion of fluorine in jadeite melt has been investigated from 10 to 15 kbars and 1200 to 1400°C using diffusion couples of Jadeite melt and fluorine-bearing jadeite melt (6.3 wt.% F). The diffusion profile data indicate that the diffusion process is concentration-independent, binary, F-O interdiffusion. The F-O interdiffusion coefficient ranges from 1.3 × 10^?7 to 7.1 × 10^?7 cm²/sec and is much larger than those obtained by Kushiro (1983) for Si-Ge and Al-Ga interdimision in jadeitic melts. The Arrhenius activation energy of diffusion is in the range of 36 to 39 kcal/mole as compared with 19 kcal/mole for fluorine tracer diffusion in a lime-aluminosilicate melt. The diffusivity and activation energy of F-O interdiffusion vary slightly with pressure, but the pressure dependence of F-O, Al-Ga and Si-Ge interdiffusion may be related to the relative volumes of the interdiffusing species for each pair. The magnitude of chemical diffusivity of fluorine is comparable to that of the chemical diffusivity of water in obsidian melts. The diffusivities of various cations are significantly increased by the addition of fluorine or water to a silicate melt. This fact, combined with the high diffusivity of fluorine, suggests that the F^? ion is the principal diffusing species in dry aluminosilicate melts and that dissolved fluorine will accelerate chemical equilibration in dry igneous melts.     

Low frequency electrical impedance of partially molten gabbro: The effect of melt geometry on electrical properties

Electrical impedance of partially molten gabbro was measured as a function of the frequency of the applied electric field. The impedance of the partially molten gabbro was mainly controlled by the volume and geometry of partial melt, which were observed with an optical microscope study of the quenched sample. The experimental results are interpreted by a theory that formulates the electrical property of rocks containing partial melt, taking ionic diffusion in the melt into account. The theory evaluates the contributions from two types of melt-geometry distributed in the rock matrix, namely:

1. (a) nearly isolated melt pockets, and

2. (b) connected melt along grain boundaries.

According to the theory, connected melt mainly contributes to the increase of net conductivity, while melt pockets increase displacement current and cause the frequency dependence of impedance. The characteristic ionic diffusivity in melt pockets was determined from the peak frequency of the imaginary part of impedance. The magnitude of diffusion coefficient is estimated to be of the order of 10⁻⁷ to 10⁻⁶ cm²/sec. This suggests that sodium, iron, magnesium and/or calcium ions mainly carry the electric charge in partial melt. The observed frequency dependence of impedance is considered to reflect ionic polarization in melt pockets due to the migration of these alkali ions toward solid-melt interfaces.The present experimental and theoretical work suggests that electrical properties of the earth's upper mantle may involve an intrinsic frequency dependence associated with ionic diffusion in the frequency range of geomagnetic observation, if a partial melt zone exists. The analysis of the earth's conductivity structure may be inadequate if such intrinsic frequency dependence of mantle media is neglected.     

Sulfur diffusion in rhyolite melts

 Diffusion rates for sulfur in rhyolite melt have been measured at temperatures of 800–1100° C, water contents of 0–7.3 wt%, and oxygen fugacities from the quartz-fayalite-magnetite buffer to air. Experiments involved dissolution of anhydrite or pyrrhotite into rhyolite melt over time scales of hours to days. Electron microprobe analysis was used to measure sulfur concentration profiles in the quenched glasses. Regression of the diffusion data in dry rhyolite melt gives D_sulfur=0.05·exp{−221±80RT}, which is one to two orders of magnitude slower than diffusion of other common magmatic volatiles such as H₂O, CO₂ and Cl^-. Diffusion of sulfur in melt with 7 wt% dissolved water is 1.5 to 2 orders of magnitude faster than diffusion in the anhydrous melt, depending on temperature. Sulfur is known to dissolve in silicate melts as at least two different species, S²⁻ and S⁶⁺, the proportions of which vary with oxygen fugacity; despite this, oxygen fugacity does not appear to affect sulfur diffusivity except under extremely oxidizing conditions. This result suggests that diffusion of sulfur is controlled by one species over a large range in oxygen fugacity. The most likely candidate for the diffusing species is the sulfide ion, S²⁻. Re-equilibration between S²⁻ and S⁶⁺ in oxidized melts must generally be slow compared to S²⁻ diffusion in order to explain the observed results. In a silicic melt undergoing degassing, sulfur will tend to be fractionated from other volatile species which diffuse more rapidly. This is consistent with analyses of tephra from the 1991 eruption of Mount Pinatubo, Philippines, and from other high-silica volcanic eruptions. Received: 26 April 1995 / Accepted: 1 November 1995     

DIffusion of Eu and Gd in basalt and obsidian

Tracer diffusion coefficients of ¹⁵³Gd and ¹⁵²Eu in olivine tholeiite have been determined at temperatures between 1150 and 1440°C. The results are identical for both tracers within experimental error. Between 1440 and 1320°C the diffusion coefficients are given by D(Eu, Gd) = 0.058 exp(?40,600/ RT). Between 1320 and 1210°C, the diffusion coefficients are constant at D = (1.4 ± 0.4) × 10^?7 cm²s^?1 and between 1210 and 1150°C, the D values drop irregularly to 4 × 10^?9 cm²s^?1. The liquidus temperature (1270°C) lies within the region of constant D. Such anomalous behavior has not been encountered in previous studies of Ca, Sr, Ba and Co diffusion in basalt. To explain the constant D value near the liquidus, we speculate that the structure of the melt changes as a function of temperature in such a way that the normal temperature dependence of the diffusivity is compensated. For example, the rare earth ions may be displaced from their (high temperature) octahedral coordination sites to other sites where they are more readily dissociated and therefore become progressively more mobile. The behavior below 1210°C may be the result of relatively stable complexes or molecules in the melt or of the formation of a REE bearing crystalline phase that has so far escaped detection. Preliminary results for Eu diffusion in obsidian are D (Eu, 800°C) = 5 × 10^?13 cm² s^?1 and D (Eu, 950°C) = 1.5 × 10^?11 cm² s^?1. These data are consistent with an activation energy of 59 Kcal mole^?1. These low diffusivities indicate that the partitioning of REE in crystallizing intermediate and acidic melts may be controlled by diffusion in the melt rather than equilibrium between the crystal surface and the bulk melt.The diffusion data are applied to partial melting in the mantle, in an attempt to explain how LREE enriched tholeiites may be derived from a LREE depleted mantle source. In this model LREE diffuse from garnet bearing regions that have small melt fractions into garnet free regions that have relatively large melt fractions. REE diffusion is so slow that this process is quantitatively significant only in small partially molten bodies (diameter ～1 km or less) or in larger, but strongly flattened bodies. Internal convective motion during diapiric rise would also increase the efficiency of the process.     

Rare earth element and gallium diffusion in yttrium aluminum garnet

Diffusion of four rare-earth elements and gallium has been measured in yttrium aluminum garnet (YAG). Sources of diffusant were mixtures of alumina and rare-earth element oxides for REE diffusion, and mixtures of gallium and yttrium oxides for Ga diffusion. Diffusion profiles were measured with Rutherford backscattering spectrometry (RBS). For the rare-earth elements investigated, the following Arrhenius relations were obtained: D_La=6.87×10^–1 exp (–582±21 kJ mol^–1 /RT) m²s^–1 D_Nd=1.63×10^–1 exp (–567±15 kJ mol^–1 /RT) m²s^–1 D_Dy=2.70×10⁰ exp (–603±35 kJ mol^–1 /RT) m²s^–1 D_Yb=1.50×10^–2 exp (–540±26 kJ mol^–1 /RT) m²s^–1 Diffusion rates for the rare earths are quite similar, in contrast with trends noted for zircon. It is likely that these differences are a consequence of the relative ionic radii of the REE and the cations for which they substitute in the mineral lattice. For gallium, the following Arrhenius relation was determined: D_Ga=9.96×10^–6 exp (–404±19 kJ mol^–1 /RT) m²s^–1 Gallium diffuses faster than the REE in YAG and has a smaller activation energy for diffusion. These data mirror relative trends in diffusion rates for YIG, in which trivalent cations occupying tetrahedral and octahedral sites (i.e., Al, Ga, Fe) diffuse faster than trivalent cations occupying dodecahedral sites (i.e., Y and the REE), and suggest that the rate-limiting process in the diffusion-controlled regime of solid-state creep of YAG is the diffusion of yttrium. Received: 10 November 1997 / Revised; accepted: 13 March 1998     

Element partitioning between olivine and silicate melt under high pressure

Element partitioning between olivine and silicate melt has been investigated at pressures 1–14 GPa, by using a 6–8 type multi-anvil high pressure apparatus. In order to observe systematics in the partitioning of trivalent ions, Li was added to the starting materials in order to increase the concentration of trivalent ions in olivine. With increasing pressure, it was found that partition coefficients of most of the elements gradually decreased. Trivalent ions generally showed parabolic pattern on partition coefficient — ionic radius diagram. When pyrolite-like material was used as the starting material, partition coefficient of Al, D_Al, gradually increased with increase in pressure while the partition coefficients of the other elements decreased, and the D_Al deviated from the parabolic pattern of other trivalent ions. The deviation of D_Al from the D pattern of the other trivalent ions was also found when olivine was employed as main component of the starting material. This result may be ascribed to the compositional change of coexisting silicate melt with increase in pressure.     

苏鲁皖新生代玄武岩中巨晶组合的稀土元素特征

我国东部苏鲁皖地区新生代碱性玄武岩中,除了含有大量地幔橄榄岩类捕虏体以外,尚含有一定数量的石榴石、普通辉石和歪长石巨晶。这些巨晶是在地幔不同深度上从玄武岩中晶出的。巨晶组合的分离结晶作用对熔体稀土元素含量有很大影响。赋存巨晶的碱性玄武岩所具有的LREE富集、HREE亏损的稀土元素分配型式是由地幔橄榄岩类部分熔融程度、石榴石巨晶和普通辉石巨晶的早期高压熔离和玄武岩的结晶分异作用等综合因素造成的。     

Coupled lithium- and iron isotope fractionation during magmatic differentiation

In this study, we investigated Fe and Li isotope fractionation between mineral separates of olivine pheno- and xenocrysts (including one clinopyroxyene phenocryst) and their basaltic hosts. Samples were collected from the Canary Islands (Teneriffa, La Palma) and some German volcanic regions (Vogelsberg, Westerwald and Hegau). All investigated bulk samples fall in a tight range of Li and Fe isotope compositions (δ⁵⁶Fe_wr = 0.06–0.17‰ and δ⁷Li_ma = 2.5–5.2‰, assuming δ⁷Li of the olivine-free matrix is virtually identical to that of the bulk sample for mass balance reasons). In contrast, olivine phenocrysts display highly variable, but generally light Fe and mostly light Li isotope compositions compared to their respective olivine-free basaltic matrix, which was considered to represent the melt (with δ⁵⁶Fe_ol = ? 0.24 to 0.14‰ and δ⁷Li_ol = ? 10.5 to + 6.5‰, respectively). Single olivine crystals from one sample display even a larger range of δ⁵⁶Fe_ol between ? 0.7 and + 0.1‰. One single clinopyroxene phenocryst displays the lightest Li isotope composition (δ⁷Li_cpx = ? 17.7‰), but no Fe isotope fractionation relative to melt. The olivine phenocrysts show variable Mg# and Ni (correlated in most cases) that range between 0.89 and 0.74 and between 300 and 3000 μg/g, respectively. These olivines likely grew by fractional crystallization in an evolving magma. One sample from the Vogelsberg volcano contained olivine xenocrysts (Mg# > 0.89 and Ni > 3000 μg/g), in addition to olivine phenocrysts. This sample displays the highest Li- and the second highest Fe-isotope fractionation between olivine and melt (Δ⁷Li_ol-melt = ? 13; Δ⁵⁶Fe_ol-melt = ? 0.29).Our data, i.e. the variable olivine- at constant whole rock and matrix isotope compositions, strongly indicate disequilibrium, i.e. kinetic Fe and Li isotope fractionation between olivine and melt (for Li also between cpx and melt) during fractional crystallization. Δ⁷Li_ol-melt is correlated with the Li partitioning between olivine and melt (i.e. with Li_ol/Li_melt), indicating Li isotope fractionation due to preferential (faster) diffusion of ⁶Li into olivine during fractional crystallization. Olivine with low Δ⁷Li_ol-melt, also have low Δ⁵⁶Fe_ol-melt, indicating that Fe isotope fractionation is also driven by diffusion of isotopically light Fe into olivine, potentially, as Fe–Mg inter-diffusion. The lowest Δ⁵⁶Fe_ol-melt (? 0.40) was observed in a sample from Westerwald (Germany) with abundant magnetite, indicating relatively oxidizing conditions during magma differentiation. This may have enhanced equilibrium Fe isotope fractionation between olivine and melt or fine dispersed magnetite in the basalt matrix may have shifted its Fe isotope composition towards higher δ⁵⁶Fe. The decoupling of Li- and Fe isotope fractionation in cpx is likely due to faster diffusion of Li relative to Fe in cpx, implying that the large investigated cpx phenocryst resided in the magma for only a short period of time which was sufficient for Li- but not for Fe diffusion. The absence of any equilibrium Fe isotope fractionation between the investigated cpx phenocryst and its basaltic host may be related to the similar Fe^3 +/Fe^2 + of cpx and melt. In contrast to cpx, the generally light Fe isotope composition of all investigated olivine separates implies the existence of equilibrium- (in addition to diffusion-driven) isotope fractionation between olivine and melt, on the order of 0.1‰.     

A computer simulation study of natural silicate melts. Part II: High pressure properties

The thermodynamic, structural and transport properties of natural silicate melts under pressure are investigated by molecular dynamics simulation with the help of a force field recently introduced by us [Guillot B. and Sator N. (2007) A computer simulation study of natural silicate melts. Part I: low pressure properties. Geochim. Cosmochim. Acta71, 1249-1265]. It is shown that the simulation reproduces accurately the bulk moduli of a large variety of silicate liquids as evaluated from ultrasonic studies. The equations of state (EOS) of the simulated melts are in good agreement with the density data on mid-ocean ridge basalt, komatiite, peridotite and fayalite as obtained either by sink/float experiments or by shock-wave compression. From the structural point of view it is shown that the population of ^[5]Al and ^[6]Al species increases rapidly upon initial compression (0-50 kbar) whereas for Si these highly coordinated species are found in a significant abundance (>5%) only above ∼50 kbar for ^[5]Si and ∼100-150 kbar for ^[6]Si. This increase of the coordination of network formers is not the only response of the melt structure to the densification: there is also a large redistribution of the T-O-T (T = Si, Al) bond angles with the pressure and noticeably upon initial compression in rhyolitic and basaltic liquids. Furthermore, a detailed analysis of the population of bridging oxygens (BO) and nonbridging oxygens (NBO) points out that the polymerization of the melt generally increases when the pressure increases, the magnitude of this polymerization enhancement being all the more important that the melt is depolymerized at low pressure. The role of triclusters (threefold coordinated oxygens to network former cations) is particularly emphasized in acidic and basaltic liquids. The pressure-induced redistribution of the oxygen atoms through the melt structure is also stressed. Finally, the simulation predicts a nonmonotonic behavior of the diffusivity of network former ions when the pressure increases, a feature with depends on the melt composition. This could have a counterpart in the electrical conductivity at sufficiently high temperature when the viscosity of the liquid is low.