Edward M. Stolper

The role of water in the petrogenesis of Mariana trough magmas

Most variations in composition among primitive basalts from the Mariana back-arc trough can be explained by melting mixtures of an NMORB-type mantle source and an H_2O-rich component, provided the degree of melting is positively and approximately linearly correlated with the proportion of the H_2O-rich component in the mixture. We conclude that the degrees of melting by which Mariana trough magmas are generated increase from magmas similar to NMORB, through more H_2O-enriched basalts, to ‘arc-like’ basalts, and that this increase is due to the lowering of the solidus of mantle peridotite that accompanies addition of the H_2O-rich component is likely to be ultimately derived from fluid from a subducting slab, but we propose that by the time fluids reach the source regions of Mariana trough basalts, they have interacted with sufficient mantle material that for all but the most incompatible of elements (with respect to fluid-mantle interaction), they are in equilibrium with the mantle. In contrast, fluids added to the source regions of Mariana island-arc magmas have typically interacted with less mantle and thus retain the signature of slab-derived fluids to varying degrees for all but the most compatible elements. Primitive Mariana arc basalts can be generated by melting mixtures of such incompletely exchanged slab-derived fluids and sources similar to NMORB-type mantle sources, but the degrees of melting are typically higher than those of Mariana trough NMORB and the sources have been variably depleted relative to the back-arc sources by previous melt extraction. This depletion may be related to earlier extraction of back-arc basin magmas or may evolve by repeated fluxing of the sources as fluid is continually added to them in the regions of arc magma generation. If fluid with partitioning behavior relative to the solid mantle similar to that deduced for the H_2O-rich component involved in the generation of Mariana trough basalts were extracted from primitive mantle, the residual mantle would have many of the minor and trace element characteristics of typical oceanic upper mantle; primitive mantle enriched in such fluid would be a satisfactory source for the continental crust in terms of its trace and minor element chemical composition.

Water in silicate glasses: An infrared spectroscopic study

Infrared and near-infrared transmission spectra have been taken on 19 volcanic and synthetic silicate glasses with known H2O contents (0.06–6.9 wt. %). Absorption peaks were observed at wavelengths of 1.41 μm, 1.91 μm, 2.22 μm, 2.53 μm, and 2.8 μm. These peaks have been attributed to the first overtone of the OH stretching vibration, the combination stretching+bending mode of H2O molecules, the combination stretching+bending mode of X-OH groups, a combination mode of the fundamental OH stretch+a low energy lattice vibration, and the fundamental OH stretching mode, respectively. Molar absorptivities of the peaks have been determined to be 0.2, 1.8, 1.0, 0.9, and 67 l/mol-cm. These values apply over the full range of glass compositions studied (albite, rhyolite, basalt).Quantitative determinations of total H2O contents and of the concentrations of molecular water and hydroxyl groups in silicate glasses are possible using these molar absorptivities, although they are limited in their accuracy by the accuracy of the reported water contents of the glasses used to calibrate these molar absorptivities. The most important uses of this technique may stem from its applicability to microsamples (≧100 μm) and to the determination of the concentrations of hydroxyl groups and molecular water in quenched silicate melts.Hydroxyl groups are the dominant hydrogen-bearing species in water-bearing glasses at low total water contents, but molecular H2O was detected in all samples with ≧0.5 weight percent total water. The concentration of hydroxyl groups increases rapidly with total water content at low total water contents, but more slowly at higher (>3 wt. %) total water contents; it may level off or even decrease at high total water contents. The concentration of molecular water increases slowly at low total water contents and more rapidly at high total water contents. More water is dissolved as molecular water than as hydroxyl groups at total water contents greater than ∼4 wt. %. Molecular water in these glasses is probably structurally bound rather than present as fluid inclusions as a separate phase, since ice bands were not observed in spectra taken at 78K and since samples were free of visible bubbles.It is proposed that the speciation of water in silicate glass formed by rapid quenching from melt equilibrated at high temperatures reflects that of the melt. According to this hypothesis, neither high water contents nor high pressures are needed to stabilize substantial quantities of molecular water in melts. This hypothesis, that water dissolves in silicate melts as both molecular water and hydroxyl groups in proportions similar to those measured in waterbearing glasses, can explain the variations in viscosity, electrical conductivity, diffusivity of “water”, diffusivity of cesium, and phase relationships that are observed in melts as functions of total water content. It also explains the observation that at vapor-saturation at high pressures, where most of the dissolved water is expected to be present as molecular water, water solubilities are similar for all melts but that at low pressures and water contents, where most dissolved water is present in dissociated form as hydroxyl groups, vapor-saturated water solubilities differ for different melt compositions. The linear relationship between water fugacity and the square of the mole fraction of total dissolved water observed for silicate melts at low water contents and the observed deviations from this linear relationship at high total water contents can be accounted for by this hypothesis.

Geochemical Consequences of Melt Percolation: The Upper Mantle as a Chromatographic Column

As magmas rise toward the surface, they traverse regions of the mantle and crust with which they are not in equilibrium; to the extent that time and the intimacy of their physical contact permit, the melts and country rocks will interact chemically. We have modeled aspects of these chemical interactions in terms of ion-exchange processes similar to those operating in simple chromatographic columns. The implications for trace element systematics are straightforward: the composition of melt emerging from the top of the column evolves from close to that of the incipient melt of the column matrix toward that of the melt introduced into the base of the column. The rate of evolution is faster in the incompatible than the compatible elements and, as a result, the abundance ratios of elements of different compatibilities can vary considerably with time. If diffusion and other dispersive processes in the melt are negligible and if exchange between melt and solid rock is rapid, extreme fractionations may occur, and the change from initial to final concentration for each element can be through an abrupt concentration front. Integration and mixing of the column output in a magma chamber or dispersive processes within the column, in particular the incomplete equilibration between matrix and fluid due to the slow diffusion in the solid phases, may lead to diffuse fronts and smooth trace element abundance patterns in the column output. If the matrix material is not replenished, the chromatographic process is a transient phenomenon. In some geological situations (e.g., under island arcs and oceanic islands), fresh matrix may be fed continuously into the column, leading to the evolution of a steady state. Aspects of the geochemistry of ultramafic rocks, island arc lavas, and comagmatic alkaline and tholeiitic magmas may be explained by the operation of chromatographic columns.

Determining the composition of high-pressure mantle melts using diamond aggregates

We present a new experimental technique for circumventing the quenching problems that have plagued high-pressure peridotite melting studies. A thin layer of ~50 μm diamonds is placed above a layer of peridotite powder. Partial melt extracted from the peridotite layer collects in the pore spaces between the diamonds and equilibrates diffusively with the residual peridotite mineralogy. Isolated from the crystalline residue, the melt quenches to a glass that records the composition of the liquid coexisting with the residual crystalline phases under the conditions of the experiment. We have used this technique to investigate partial melting of a fertile mantle composition at 10 kbar and a temperature range of 1270–1390°C. Oxide concentrations in the liquids from the longest duration runs (up to 151 hours) vary systematically with increasing temperature: TiO_2, Al_2O_3, and Na_2O decrease monotonically, while Cr_2O_3, FeO^∗, and MgO increase steadily. CaO shows more complicated behavior, first increasing and then decreasing, with the crest in the temperature-CaO trend approximately coincident with the disappearance of clinopyroxene from the residue between 1330 and 1350°C. Overall variation in silica content with temperature is small, and there appears to be a minimum at about 12% melting. The compositions of liquids produced in time series, temperature reversal, and two-stage experiments (conducted to test the technique) all indicate that our experimentally determined liquid compositions represent close approaches to equilibrium. Calculated melt fractions (F) also vary systematically with temperature. The slope of the T(°C)-F curve is not constant over the spinel lherzolite melting interval, but decreases as temperature increases from 1270 to 1330°C. Extrapolating the curve back to zero melt suggests that the anhydrous solidus temperature for our peridotite starting composition is ~ 1240°C. At temperatures below the cpxout curve, melt generation occurs via the reaction, 0.38 opx + 0.71 cpx + 0.13 sp → 0.22 oliv + 1.0 liq, and the proportions of minerals that enter the melt appear to be independent of temperature. At temperatures above cpx-out, the less well constrained melting reaction is: 1.06 opx + 0.04 sp = 0. 1 oliv + 1 liq. The fact that all of the 10 kbar melts have FeO^∗ contents that are substantially lower than those reported in any primitive MORB glasses further strengthens the conclusions that these glasses are not 10 kbar primary melts, that they involve a component of higher pressure partial melting, and that they have evolved by significant olivine fractionation from more primitive liquids. Our experimental data also provide an independent check of the results of recent peridotite partial melting calculations. Efforts to parameterize the experimental database on peridotite melting, and to calculate melt compositions as a function of P, T, and F are partially successful in reproducing the compositional trends determined in this study.

A Phase Diagram for Mid-Ocean Ridge Basalts: Preliminary Results and Implications for Petrogenesis

Samples of a primitive mid-ocean ridge basalt (MORB) glass were encapsulated in a mixture of ol (Fo90) and opx (En90) and melted at 10, 15, and 20 kbar. After quenching, the basaltic glass was present as a pool within the ol+opx capsule, but its composition had changed so that it was saturated with ol and opx at the conditions of the experiment. By analyzing the quenched liquid, the location of the ol+opx cotectic in the complex, multicomponent system relevant to MORB genesis was determined.As pressure increases from 1 atm to 10 kbar, the dry ol+opx cotectic moves from quartz tholeiitic to olivine tholeiitic compositions. With further increases in pressure, the cotectic continues to move toward the ol-di-plag join (i.e., toward alkalic compositions). Between 15 and 20 kbar, ol+opx+di-saturated liquids change from tholeiitic to alkalic in character, although part of the ol+opx cotectic is still in the tholeiitic (i.e, hy-normative) part of composition space. At pressures of 10–15 kbar, tholeiitic liquids may be able to fractionate to alkalic liquids on the ol+di cotectic.Primitive MORB compositions come close to but do not actually lie on the ol+opx cotectic under any conditions studied. This suggests that not even the most primitive of known MORBs are primary melts of the mantle. The correspondence of most MORBs to the 1 atm ol+di+plag cotectic suggests that low pressure fractionation was involved in their genesis from parent liquids. Picritic liquids that have been proposed as parents to the MORB suite could equilibrate with harzburgite (or Iherzolite) at 15–20 kbar and thus could be primary. Fractionation of ol from these liquids could yield primitive MORB liquids, but other primary liquids or more complex fractionation paths involving others phases in addition to ol cannot be ruled out. The possibility that these picritic liquids could equilibrate with ol+opx at 25–30 kbar cannot be ruled out.

A Habitable Fluvio-Lacustrine Environment at Yellowknife Bay, Gale Crater, Mars

The Curiosity rover discovered fine-grained sedimentary rocks, which are inferred to represent an ancient lake and preserve evidence of an environment that would have been suited to support a martian biosphere founded on chemolithoautotrophy. This aqueous environment was characterized by neutral pH, low salinity, and variable redox states of both iron and sulfur species. Carbon, hydrogen, oxygen, sulfur, nitrogen, and phosphorus were measured directly as key biogenic elements; by inference, phosphorus is assumed to have been available. The environment probably had a minimum duration of hundreds to tens of thousands of years. These results highlight the biological viability of fluvial-lacustrine environments in the post-Noachian history of Mars.

The speciation of water in silicate melts

Previous models of water solubility in silicate melts generally assume essentially complete reaction of water molecules to hydroxyl groups. In this paper a new model is proposed that is based on the hypothesis that the observed concentrations of molecular water and hydroxyl groups in hydrous silicate glasses reflect those of the melts from which they were quenched. The new model relates the proportions of molecular water and hydroxyl groups in melts via the following reaction describing the homogeneous equilibrium between melt species: H_2O_(molecular) (melt) + oxygen (melt) = 2OH (melt). An equilibrium constant has been formulated for this reaction and species are assumed to mix ideally. Given an equilibrium constant for this reaction of 0.1–0.3, the proposed model can account for variations in the concentrations of molecular water and hydroxyl groups in melts as functions of the total dissolved water content that are similar to those observed in glasses. The solubility of molecular water in melt is described by the following reaction: H_2O (vapor) = H_2O_(molecular) (melt). These reactions describing the homogeneous and heterogeneous equilibria of hydrous silicate melts can account for the following observations: the linearity between f_(H2O) and the square of the mole fraction of dissolved water at low total water contents and deviations from linearity at high total water contents; the difference between the partial molar volume of water in melts at low total water contents and at high total water contents; the similarity between water contents of vapor-saturated melts of significantly different compositions at high pressures versus the dependence on melt composition of water solubility in silicate melts at low pressures; and the variations of viscosity, electrical conductivity, the diffusivity of “water,” the diffusivity of cesium, and phase relationships with the total dissolved water contents of melts. This model is thus consistent with available observations on hydrous melt systems and available data on the species concentrations of hydrous glasses and is easily tested, since measurements of the concentrations of molecular water and hydroxyl groups in silicate glasses quenched from melts equilibrated over a range of conditions and total dissolved water contents are readily obtainable.

Alkalic magmas generated by partial melting of garnet pyroxenite

Many oceanic-island basalts (OIBs) with isotopic signatures of recycled crustal components are silica poor and strongly nepheline (ne) normative and therefore unlike the silicic liquids generated from partial melting of recycled mid-oceanic-ridge basalt (MORB). High-pressure partial-melting experiments on a garnet pyroxenite (MIX1G) at 2.0 and 2.5 GPa produce strongly ne-normative and silica-poor partial melts. The MIX1G solidus is located below 1350 and 1400 8C at 2 and 2.5 GPa, respectively, slightly cooler than the solidus of dry peridotite. Chemographic analysis suggests that natural garnet pyroxenite compositions straddle a thermal divide. Whereas partial melts of compositions on the silica-excess side of the divide (such as recycled MORB) are silica saturated, those from silica-deficient garnet pyroxenites can be alkalic and have similar- ities to low-silica OIB. Although the experimental partial melts are too rich in Al2O3 to be parental to highly undersaturated OIB suites, higher-pressure (4-5 GPa) partial melting of garnet pyrox- enite is expected to yield more appropriate parental liquids for OIB lavas. Silica-deficient garnet pyroxenite, which may originate by mixing of MORB with peridotite, or by recycling of other mafic lithologies, represents a plausible source of OIB that may resolve the apparent contradiction of strongly alkalic composition with iso- topic ratios characteristic of a recycled component.

Infrared spectroscopic measurements of CO2 and H2O in Juan de Fuca Ridge basaltic glasses

Dissolved H_2O and CO_2 contents in basaltic glasses from the Juan de Fuca Ridge and neighboring seamounts were determined by infrared spectroscopy. CO_2 contents range from about 45 to 360 ppm by weight, with carbonate ion complexes the only detectable form of dissolved carbon. Samples erupted at a given depth exhibit a large range in dissolved CO_2 contents that we interpret to be the result of variable amounts of degassing. The lowest CO_2 contents at each depth are in reasonable agreement with the experimentally determined CO_2 solubility curve for basalt at low pressures. All glasses with CO_2 values higher than the experimentally determined solubility at the eruption depth are oversaturated because of incomplete degassing. The highest CO_2 contents are spatially associated with the local topographic highs for each ridge segment. Lavas from relatively deep areas may have had greater opportunity to degas during ascent from a magma chamber or during lateral flow in dikes or seafloor lava flows. The highest observed CO_2 concentrations are from the axial seamount and lead to an estimate of a minimum depth to the magma chamber of 2.7 km beneath the ridge axis. H_2O contents vary from 0.07 to 0.48 wt.%, with hydroxyl groups the only detectable form of dissolved water. Water contents correlate positively with FeO^*/MgO and the highest water contents are found in the incompatible element-enriched Endeavour segment lavas. Variations in ratios of water to other incompatible elements suggest that water has a bulk partition coefficient similar to La during partial melting (D ∼ 0.01).

Experimental petrology of eucritic meteorites

Low pressure melting experiments on eucritic meteorites demonstrate that the compositions of most eucrites can be generated by low pressure fractionation of pigeonite and plagioclase from liquids similar in composition to the Sioux County and Juvinas eucrites. It is unlikely that the liquids with compositions similar to Sioux County and Juvinas were themselves residual liquids produced by extensive fractionation of more magnesian parental liquids. The compositions of Stannern and Ibitira cannot be produced by fractionation of liquids with compositions similar to other known eucrites. Liquid compositions similar to Stannern, Ibitira, and Sioux County could have been generated by increasing degrees of low pressure partial melting of source regions composed of olivine (~Fo65), pigeonite (~Wo5En65), plagioclase (~An94), Cr-rich spinel, and metal. These source assemblages may have been primitive, undifferentiated material of the basaltic achondrite parent body and the eucrites may represent melts produced in early stages of its melting and differentiation. Further melting in these source regions, after exhaustion of plagioclase, may have produced magnesian liquids from which the magnesian pyroxenes and olivines in howardites, diogenites, and mesosiderites crystallized in closed-system plutonic environments. Most of the cumulate eucrites (e.g. Moama, Moore County, Serra de Mage) could not have equilibrated with liquids similar in composition to known eucrites. These cumulates may have accumulated from liquids produced by extensive fractionation of advanced partial melts of the source regions of eucritic liquids. A depletion in Na, K, and Rb in Ibitira is noted.