Bernard W. Evans

Phase relations of epidote-blueschists

Phase diagrams have been computed for minerals in the system NCMASH, using the optimised thermodynamic data-base of Berman (1988) augmented by data on glaucophane and pumpellyite, to illustrate the relationships between rocks containing epidote and sodic amphibole and those containing assemblages characteristic of neighbouring metamorphic facies. The stoichiometries of stable bounding equilibria yield a simple list of mineral parageneses that serves to define an epidote-blueschist facies and surrounding metamorphic facies. The calculations suggest that when iron-free the pairs zoisite+glaucophane and clinozoisite+glaucophane do not possess a stability field. The natural Fe-bearing paragenesis of epidote+sodic amphibole occupies a closed field in the PT-diagram, whose size increases dramatically as the ratio Fe3+/Al in the minerals is increased. Where garnet is involved in bounding reactions, such as with respect to eclogite and (garnet-bearing) albite-epidote-amphibolite, the epidote-blueschist field decreases in size as the ratio Fe2+/Mg is increased. The classic blueschist-to-greenschist transition reaction is best developed in rocks with large Fe3+ and Fe2+ contents and is possibly metastable in MgAl-rich rocks. In the latter, a wedge of albite-epidote-(actinolite/barroisite) amphibolite separates the epidote-blueschist from the greenschits paragenesis field. A first occurrence of garnet divides the epidote-blueschist field into a lower-temperature garnet-free zone and a higher-temperature garnet-bearing zone, although the temperatures calculated for this garnet “isograd” seem high. The locations of all the facies boundaries are influenced by H2O-activity; however, even minor amounts of CO2 in the fluid phase (X(CO2) > 0.005) render the epidote+sodic amphibole paragenesis metastable, so that facies boundaries cannot be shifted by varying the CO2-content of the fluid phase. Below 600°C no mutual boundaries exist between the amphibolite facies and the epidote-blueschist and eclogite facies respectively. The calculated low-pressure limit for the eclogite paragenesis pyralspite+omphacite (Jd 50%)+quartz typical metabasite compositions is 12 to 14 kbars, with H2O-activity=0.9 or 1.5 to 2 kbar lower for Jd 30%.

Chrome-spinel in progressive metamorphism—a preliminary analysis

Progressive metamorphism of serpentinite and allied rocks causes systematic changes in the composition of the chrome-spinel phase, when the latter is in equilibrium with chlorite and two magnesium silicate minerals. The stable spinel in antigorite-serpentinites is Al-poor magnetite, Cr-magnetite, or ferrit-chromite, depending on the local CrFe3+ ratio in the rock. With increasing metamorphic grade up to middle amphibolite facies conditions (talc + olivine, or Ca-poor amphibole + olivine stable) more chromiferous spinels (chromites) are encountered, containing modest amounts of Al. Further increase in grade (enstatite + olivine stable) extends the range of possible spinel compositions to green MgAl2O4-rich spinel. The Al-content is governed by P-, T-, and ƒH2O-sensitive equilibria involving chlorite. MgFe2+ ratios in the spinel are a function of Cr, Al and Fe3+ in the spinel, and the ratio MgFe2+ in coexisting silicates. Solid solution in natural chromites on the magnetite-chromite join is complete at ≈500°C, and close to the join chromite-Al spinel (with variable Fe/Mg) it is complete at ≈700°C. The temperature dependence of KD, the olivine-spinel Fe-Mg partition coefficient, is greater than implied by the jackson (1969) geothermometer. A tentative, revised, graphical calibration is offered, based on microprobe-analyzed high-grade metamorphic pairs and pairs from basaltic pumice. This new plot gives results which are broadly consistent with relative temperatures of equilibration inferred from other geothermometers, for alpine peridotites, peridotite nodules and meteorites.

The serpentinite multisystem revisited : chrysotile is metastable

The two rock-forming polymorphs of serpentine Mg3Si2O5(OH)4, lizardite and chrysotile, occur in nature in virtually identical ranges of temperature and pressure, from surficial or near-surficial environments to temperatures perhaps as high as 400°C. Laboratory evidence indicates that lizardite is the more stable at low temperatures, but the difference in their Gibbs free energies is not more than about 2 kJ in the 300-400°C range. Above about 300°C, antigorite + brucite is more stable than both; in other words, chrysotile is nowhere the most stable. The crystal structures of lizardite and chrysotile give rise to contrasting crystallization behaviors and hence modes of occurrence. The hydration of peridotite at low temperature results in the growth of lizardite from olivine, and (commonly topotactically) from chain and sheet silicates, although the MgO-SiO2-H2O (MSH) phase diagram predicts antigorite + talc in bastite. The activity of H2O during serpentinization may be buffered to low values by the solids, making the reaction of olivine to lizardite + brucite a stable one. Conservation of oxygen and inheritance of the Fe2+/Mg exchange potential of olivine lead predictably to the precipitation of a highly magnesian lizardite and magnetite, and to the evolution of H2. Volume expansion is made possible by lizardites force of crystallization, and it is tentatively suggested that this might account for the a-serpentine orientation (length normal to (001)) of lizardite pseudofibers in mesh rims and hourglass pseudomorphs after olivine. Whereas mineral replacements commonly conserve volume, in massif serpentinites the diffusive loss of Mg and Si needed for volume conservation during serpentinization requires chemical potential gradients that are unlikely to exist. For small bodies of serpentinite, sheared serpentinite, and systems of large water/rock ratio, volume expansion may be much less. Chrysotile is most conspicuously developed in tectonically active environments, where associated lithotypes show marginal greenschist-facies parageneses and antigorite tends to make its first appearance. Chrysotile growth is favored in isotropic stress microenvironments of fluid-filled voids and pores (where it may ultimately crystallize pervasively), and in veins, generally after active hydration in the immediate surroundings has ceased. This nevertheless allows the simultaneous growth of lizardite and chrysotile in adjacent partially and fully serpentinized peridotite, respectively, as in the cores and rims of kernel structures. Although prominent along shear surfaces, chrysotile growth as slip fiber is promoted by the presence of fluid rather than shear stress. Unlike lizardite, whose growth produces the stress associated with expansion, extreme flattening and shear might be expected to destroy the chrysotile structure. Thus, lizardite and chrysotile behave as though they were a stress-antistress mineral pair. Calorimetric, solubility, and reaction-reversal experiments on chrysotile integrate contributions to its free energy arising from surface properties and, most importantly, from its radius-dependent strain energy. Minimally strained chrysotile (r ≈ 90 Å) may in fact be more stable than lizardite, whereas a maximal-radius chrysotile (r ≈ 200 Å) is not. A model for chrysotile in veins cutting lizardite- or antigorite-bearing rock involves nucleation of low-strain chrysotile followed by kinetically favored crystallization of higher-energy layers driven by mild fluid supersaturation maintained by local potential gradients. It is not clear if this explanation adequately accounts for serrate veins and mass-fiber chrysotile. A revised phase diagram for lizardite and antigorite is offered, and possible stable and metastable reactions among the phases in serpentinites are followed on an isobaric diagram of reaction free energy (driving force) as a function of temperature. Composition-induced equilibrium shifts are believed unlikely to be determinative in most occurrences of Mg-rich lizardite and chrysotile. Circumstances of growth rather than temperature and pressure determine the occurrence of chrysotile vis-à-vis lizardite in serpentinites.

THERMODYNAMICS OF RHOMBOHEDRAL OXIDE SOLID SOLUTIONS AND A REVISION OF THE FE-TI TWO-OXIDE GEOTHERMOMETER AND OXYGEN-BAROMETER

A model for the thermodynamic properties of rhombohedral oxide solid solutions in the system Fe2O3-FeTiO3-MgTiO3-MnTiO3 (containing minor amounts of Al2O3) is presented. The model accounts for temperature and compositionally dependent long-range cation-order and the related high to low symmetry structural phase transition. The model is calibrated from the cation-ordering data of Harrison and others (2000; Harrison and Redfern, 2001) and experimental data on Fe+2Ti ⇔ (Fe+3)2 exchange between rhombohedral oxide and spinel from Lattard and others (2005) and Evans and others (2006). Successful calibration require introduction of an energetic contribution attributed to short-range cation-order, which reduces the configurational entropy of the solid solution. The resultant thermodynamic model for the rhombohedral oxides is internally consistent with the model for spinel solid solutions of Sack and Ghiorso (1991a, 1991b) and with the endmember thermodynamic properties database of Berman (1988); a new model equation for the isobaric heat capacity of ulvöspinel (cubic Fe2TiO4) is proposed and values of the enthalpy of formation, −1490.417 kJ/mol, and third law entropy, 184.199 J/K-mol, at 298.15 K and 105 Pa are recommended. The new model forms the basis of a revised FeTi-oxide geothermometer/oxygen barometer, which is applied to a newly compiled dataset of natural two oxide pairs from silicic volcanic rocks. Results are compared to previous formulations with the general conclusion that the new model gives a better estimate of oxidation state for magmas that equilibrated under conditions more oxidizing than the nickel-nickel oxide buffer. Estimates of oxygen fugacity are fairly insensitive to analytical uncertainties in oxide compositions. By contrast, temperature estimates are especially sensitive to analytical error and to the abundances of “minor” constituents. Application of the geothermometer to oxide pairs that grew under conditions where the rhombohedral phase was cation disordered (that is high temperature or at oxygen fugacities greater by about one log10 unit than the nickel-nickel oxide buffer) results in an uncertainty due solely to analytical error of at least 50°C and sometimes as high as 100 °C. Temperature estimates from the new geothermometer can be made using either the Fe+2Ti ⇔ (Fe+3)2 exchange or Fe+2 ⇔ Mg exchange between the two oxides. Comparison of the two temperature estimates provides a means of evaluating the internal consistency of coexisting oxide compositions and assessing the extent of disequilibrium. Temperatures calculated from the new model are found to be consistent with experimental phase relations for the stability of cummingtonite in silicic volcanics. Other petrologic constraints on derived temperatures are examined including limits on the width of the miscibility gap and the development of self-reversed remanent magnetization in the rhombohedral series. Software that implements the new thermodynamic model and the two-oxide geothermometer/oxygen barometer is available from http://www.ofm-research.org/.

Redox control of sulfur degassing in silicic magmas

Explosive eruptions involve mainly silicic magmas in which sulfur solubility and diffusivity are low. This inhibits sulfur exsolution during magma uprise as compared to more mafic magmas such as basalts. Silicic magmas can nevertheless liberate large quantities of sulfur as shown by the monitoring of SO 2 in recent explosive silicic eruptions in arc settings, which invariably have displayed an excess of sulfur relative to that calculated from melt degassing. If this excess sulfur is stored in a fluid phase, it implies a strong preference of sulfur for the fluid over the melt under oxidized conditions, with fluid/melt partition coefficients varying between 50 and 2612, depending on melt composition. Experimentally determined sulfur partition coefficients for a dacite bulk composition confirm this trend and show that in volcanic eruptions displaying excess gaseous sulfur, the magmas were probably fluid-saturated at depth. The experiments show that in more reduced silicic magmas, those coexisting only with pyrrhotite, the partition coefficient decreases dramatically to values around 1, because pyrrhotite locks up nearly all the sulfur of the magma. Reevaluation of the sulfur yields of some major historical eruptions in the light of these results shows that for oxidized magmas, the presence of 1-5 wt % fluid may indeed account for the differences observed between the petrologic estimate of the sulfur yield and that constrained from ice core data. Explosive eruptions of very large magnitude but involving reduced and cool silicic magmas, such as the Toba or the Bishop events, release only minor amounts of sulfur and could have consequently negligible long-term (years to centuries) atmospherical effects. This redox control on sulfur release diminishes as the melt composition becomes less silicic and as temperature increases, because both factors favor more efficient melt sulfur degassing owing to the increased diffusivity of sulfur in silicate melts under such conditions.

Petrogenesis of garnet lherzolite, Cima di Gagnone, Lepontine Alps

Abstract Garnet lherzolite at Cima di Gagnone has chemical and mineralogical properties similar to those of other garnet lherzolites in the lower Pennine Adula/Cima Lunga Nappe (Alpe Arami, Monte Duria). The Cima di Gagnone occurrence encloses mafic boudins that belong to an eclogite-metarodingite suite common in the numerous neighboring ultramafic lenses. The ultramafic rocks at Cima di Gagnone, including the garnet lherzolite, are interpreted as tectonic fragments of an originally larger lherzolite body that underwent at least partial serpentinization prior to regional metamorphism. This lherzolite body cycled through at least three metamorphic facies: greenschist or blue-schist (as antigorite serpentinite) → eclogite (as garnet lherzolite), pre-Alpine or early Alpine → amphibolite facies (as chlorite-enstatite-tremolite peridotite), Lepontine metamorphism. Relics of titanoclinohumite in the garnet peridotite, as also recorded by Mockel near Alpe Arami, are consistent with this metamorphic history, since they indicate a possible connection with Pennine antigorite serpentinites, e.g., Liguria, Piedmont, Zermatt-Saas, Malenco, Pustertal, all of which have widespread titanoclinohumite belonging to the antigorite paragenesis. Estimated pressures in excess of 20 kbar and temperatures of 800°±50°C for the garnet lherzolite assemblage are not inconsistent with conditions inferred for Gagnone and Arami eclogites. These conditions could have been reached during deep subduction zone metamorphism. It is shown by calculation that the effects of Fe and Cr on the location of the garnet lherzolite/spinel lherzolite phase boundary largely counter-balance each other.

Lizardite versus antigorite serpentinite: Magnetite, hydrogen, and life(?)

The serpentinization of peridotite operates according to one or the other, or a combination, of two end-member mechanisms. In low-temperature environments (50–300 °C), where lizardite is the predominant serpentine mineral, olivine is consumed by reaction with H2O but its composition (Mg#) remains unchanged. Mg-rich lizardite, magnetite, and dihydrogen gas (±brucite) are products of the reaction. At higher temperatures (400–600 °C), rates of MgFe diffusion in olivine are orders of magnitude faster, with the result that the growth of Mg-rich antigorite can be accommodated by a compositional adjustment of olivine, eliminating the need to precipitate magnetite and evolve hydrogen. This latter end-member mechanism probably best reflects the situation in the forearc mantle wedge.

Titanian hydroxyl-clinohumite: Formation and breakdown in antigorite rocks (Malenco, Italy)

Petrographic evidence is presented for the breakdown of titanian hydroxyl-clinohumite to olivine+magnesian ilmenite (or geikielite)±magnetite in the outermost zone of the Bergell aureole in the Malenco Serpentinite, Prov. Sondrio, Italy. The breakdown coincides in the field with the isograd reaction: antigorite+diopside=olivine+tremolite+H2O. It is therefore concluded that this variety of clinohumite is unstable above approximately 520° C at a pressure of 3 kbars. Elsewhere in the Malenco Serpentinite, titanian hydroxyl-clinohumite may be found to have reacted with CO2 to produce antigorite, magnesian ilmenite and magnesite. Titanian hydroxyl-chondrodite was detected in one sample. Under crustal pressures, the stability field of F-free clinohumite is entirely contained inside that of antigorite. The stable occurrence of titanian clinohumite in high-grade metamorphic ultrabasic rocks may be attributed to the substitution of F for OH.

Geochemistry of high-grade eclogites and metarodingites from the Central Alps

Analyses for major, minor, and trace element contents of metamorphosed, variably rodingitized mafic rocks demonstrate substantial removal of Na and as much as three-fold gains in Ca as a consequence of rodingitization. Modest declines in Si and Fe can be explained in terms of dilution effects. Losses in K and Ba do not correlate with Ca% and may have been caused by an alteration process not related to the rodingitization. The Ca-metasomatism was not accompanied by a gain in Sr. The relative contents of Ti, Zr, Hf, Y, Co, Sc, and heavy REE show no readily detectable changes, despite the rodingitization (±other alteration) and subsequent metamorphisms, namely, eclogite facies (T≧800° C, P≧ 20kbar) followed by amphibolite facies, sillimanite zone. Protoliths were tholeiitic basalt or diabase, and gabbro, with trace element contents indicative of a spreading center origin. Trace element and REE patterns indicate low-pressure fractionation of this magma, with plagioclase stable. This petrogenesis is consistent with prior conclusions on the shallow crustal origin of the protolith of the eclogite-metarodingite-garnet lherzolite suite in the Cima Lunga-Adula nappe, Central Alps. Based on their bulk chemical composition, the mafic rocks in this suite could be the equivalent of Mesozoic ophiolitic rocks in the more external parts of the Alps.

Antigorite-ophicarbonates: Contact metamorphism in Valmalenco, Italy

Outside the Bergell tonalite contact aureole, ophicarbonate rocks consist of blocks of antigorite schist embedded in veins of calcite ± tremolite. An antigorite schistosity predates some of these calcite veins. Mono- and bimineralic assemblages occur in reaction zones associated with the veins. Within the aureole, the ophicarbonate veining becomes less distinct and polymineralic assemblages become more frequent. A regular sequence of isobaric univariant assemblages is found, separated by isograds corresponding to isobaric invariant assemblages. In order of increasing grade the invariant assemblages are: antigorite+diopside+olivine+tremolite+calcite antigorite+dolomite+olivine+tremolite+calcite antigorite+olivine+talc+magnesite antigorite+dolomite+olivine+tremolite+talc These assemblages match a previously derived topology in P-T-XCO2 space for the system CaO-MgO-SiO2-H2O-CO2; the field sequence can be used to adjust the relative locations of calculated invariant points with respect to temperature. Isobaric univariant and invariant assemblages are plotted along a profile map to permit direct comparison with the phase diagram.It is inferred that, during the formation of the ophicarbonate veins, calcite precipitated from fluid introduced into the serpentinite. During contact metamorphism, however, the compositions of pore fluids evolved by reaction in the ophicarbonate rocks were largely buffered by the solid phases. This control occurred on a small scale, because there are local variations in the buffering solid assemblages within a centimeter range.