Annite stability revised. 1. Hydrogen-sensor data for the reaction annite = sanidine + magnetite + H2

In P - T - logfO₂ space, the stability of annite (ideally KFe ₃ ²⁺ (OH)₂AlSi₃O₁₀) at high fO₂ (low fH₂) is limited by the reaction: annite = sanidine + magnetite + H₂. Using the hydrogen-sensor technique, the equilibrium fH₂ of this reaction was measured between 500 and 800° C at 2.8 kbar in 50° C intervals. Microbrobe analyses of the reacted annite+sanidine+magnetite mixtures show that tetrahedral positions of annite have a lower Si/Al ratio than the ideal value of 3/1. Silicon decreases from 2.9 per formula unit at low temperatures to 2.76 at high temperatures. As determined by Mössbauer spectroscopy in three experimental runs, the Fe³⁺ content of annite in the equilibrium assemblage is 11%±3. A least squares fit to the hydrogensensor data gives H _R ⁰ = 50.269 ± 3.987 kJ and S _R ⁰ = 83.01 ± 4.35 J/K for equilibrium (1). The hydrogene-sensor data are consistent with temperature half brackets determined in the classical way along the nickel-nickel oxide (NNO) and quartz-fayalite-magnetite (QFM) buffers with a mixture of annite+sanidine+magnetite for control. Compared to published oxygen buffer reversals, agreement is only found at high temperature and possible reasons for that discrepancy are discussed. The resulting slope of equilibrium (1) in logfO₂ – T dimensions is considerably steeper than previously determined and between 400 and 800°C only intersects with the QFM buffer curve. Based on the hydrogen-sensor data and on the thermodynamic dataset of Berman (1988, and TWEEQ data base) for sanidine, magnetite and H₂, the deduced standard-state properties of annite are: H _f ⁰ =-5127.376±5.279 kJ and S ⁰=422.84±5.29 J/(mol K). From the recently published unit cell refinements of annites and their Fe³⁺ contents, determined by Mössbauer spectroscopy (Redhammer et al. 1993), the molar volume of pure annite was constrained as 15.568±0.030 J/bar. A revised stability field for annite is presented, calculated between 400 and 800°C.     

Differentiation and crystallization conditions of basalts from the Kerguelen large igneous province: an experimental study

Phase relations of basalts from the Kerguelen large igneous province have been investigated experimentally to understand the effect of temperature, fO₂, and fugacity of volatiles (e.g., H₂O and CO₂) on the differentiation path of LIP basalts. The starting rock samples were a tholeiitic basalt from the Northern Kerguelen Plateau (ODP Leg 183 Site 1140) and mildly alkalic basalt evolved from the Kerguelen Archipelago (Mt. Crozier on the Courbet Peninsula), representing different differentiation stages of basalts related to the Kerguelen mantle plume. The influence of temperature, water and oxygen fugacity on phase stability and composition was investigated at 500 MPa and all experiments were fluid-saturated. Crystallization experiments were performed at temperatures between 900 and 1,160°C under oxidizing (log fO₂ ~ ΔQFM + 4) and reducing conditions (log fO₂ ~ QFM) in an internally heated gas-pressure vessel equipped with a rapid quench device and a Pt-Membrane for monitoring the fH₂. In all experiments, a significant influence of the fO₂ on the composition and stability of the Mg/Fe-bearing mineral phases could be observed. Under reducing conditions, the residual melts follow a tholeiitic differentiation trend. In contrast, melts have high Mg# [Mg²⁺/(Mg²⁺ + Fe²⁺)] and follow a calk-alkalic differentiation trend at oxidizing conditions. The comparison of the natural phenocryst assemblages with the experimental products allows us to constrain the differentiation and pre-eruptive conditions of these magmas. The pre-eruptive temperature of the alkalic basalt was about 950–1,050°C. The water content of the melt was below 2.5 wt% H₂O and strongly oxidizing conditions (log fO₂ ~ ΔQFM + 2) were prevailing in the magma chamber prior to eruption. The temperature of the tholeiitic melt was above 1,060°C, with a water content below 2 wt% H₂O and a log fO₂ ~ ΔQFM + 1. Early fractionation of clinopyroxene is a crucial step resulting in the generation of silica-poor and alkali-rich residual melts (e.g., alkali basalt). The enrichment of alkalis in residual melts is enhanced at high fO₂ and low aH₂O.     

Crystallization conditions in the Upper Pollara magma chamber, Salina Island, Southern Tyrrhenian Sea

Summary Pyroclastites erupted from the Upper Pollara magma chamber (13 ka, Salina Island, Aeolian Archipelago) resulted from mingling and mixing of rhyolitic and andesitic magmas. An experimental study has been conducted on the rhyolitic end-member to constrain the pre-eruptive conditions of the magma. In order to check for the role of mixing on the equilibrium phase assemblage, three different starting compositions, corresponding to three different mixing degrees, have been used. The crystallization experiments were conducted at two different oxygen fugacities and at variable temperature and fluid contents. The results indicate that the natural mineralogical assemblage can only be reproduced from a composition showing a certain degree of mixing. Assuming a pressure of 200 MPa (generally accepted for the Aeolian Islands), the pre-eruptive temperature of the magmas is estimated between 755 and 800 °C and the water content of the melt was higher than 4–4.5 wt.%. The Upper Pollara magma crystallized at relatively high fO₂ (ΔlogfO₂ = Ni–NiO + 1 log unit), compared to rhyolitic magmas from Lipari and Vulcano. As this difference has not been observed for the most primitive magmas the difference in fO₂ could be related to different degassing processes operating in Salina and Lipari – Vulcano magmas.     

Iron in monticellite as an oxygen barometer for kimberlite magmas

Monticellite is a common magmatic mineral in the groundmass of kimberlites. A new oxygen barometer for kimberlite magmas is calibrated based on the Fe content of monticellite, CaMgSiO₄, in equilibrium with kimberlite liquids in experiments at 100 kPa from 1,230 to 1,350°C and at logfO₂ from NNO-4.1 to NNO+5.3 (where NNO is the nickel–nickel oxide buffer). The XFe_Mtc/XFe_liq was found to decrease with increasing fO₂, consistent with only Fe²⁺ entering the monticellite structure. Although the XFe-in-monticellite varies with temperature and composition, these dependencies are small compared to that with fO₂. The experimental data were fitted by weighted least square regression to the following relationship: \Updelta \textNNO = \frac{ log[ 0.858( ±0.021)\fracX_\textFe^\textLiq X_\textFe^\textMtc ] - 0.139( ±0.022) }0.193( ±0.004) \Updelta {\text{NNO}} = \frac{{\left\{ {\log \left[ {0.858( \pm 0.021)\frac{{X_{\text{Fe}}^{\text{Liq}} }}{{X_{\text{Fe}}^{\text{Mtc}} }}} \right] - 0.139( \pm 0.022)} \right\}}}{0.193( \pm 0.004)} where ΔNNO is the fO₂ relative to that of the Nickel-bunsenite (NNO) buffer and XFe_liq/XFe_Mtc is the ratio of mole fraction of Fe in liquid and Fe-in-monticellite (uncertainties at 2σ). The application of this oxygen barometer to natural kimberlites from both the literature and our own investigations, assuming the bulk rock FeO is that of their liquid FeO, revealed a range in fO₂ from NNO-3.5 to NNO+1.7. A range of Mg/(Mg + Fe²⁺) (Mg#) for kimberlite melts of 0.46–0.88 was derived from the application of the experimentally determined monticellite-liquid Kd Fe²⁺–Mg to natural monticellites. The range in Mg# is broader and less ultramafic than previous estimates of kimberlites, suggesting an evolution under a wide range of petrologic conditions.     

High pressure experimental calibration of the olivine-orthopyroxene-spinel oxygen geobarometer: implications for the oxidation state of the upper mantle

Synthetic spinel harzburgite and lherzolite assemblages were equilibrated between 1040 and 1300° C and 0.3 to 2.7 GPa, under controlled oxygen fugacity (f _{O 2}). f _{O 2} was buffered with conventional and open double-capsule techniques, using the Fe−FeO, WC-WO₂-C, Ni−NiO, and Fe₃O₄−Fe₂O₃ buffers, and graphite, olivine, and PdAg alloys as sample containers. Experiments were carried out in a piston-cylinder apparatus under fluid-excess conditions. Within the P-T-X range of the experiments, the redox ratio Fe³⁺/ΣFe in spinel is a linear function of f _{O 2} (0.02 at IW, 0.1 at WCO, 0.25 at NNO, and 0.75 at MH). It is independent of temperature at given Δlog(f _{O 2}), but decreases slightly with increasing Cr content in spinel. The Fe³⁺/ΣFe ratio falls with increasing pressure at given Δlog(f _{O 2}), consistent with a pressure correction based on partial molar volume data. At a specific temperature, degree of melting and bulk composition, the Cr/(Cr+Al) ratio of a spinel rises with increasing f _{O 2}. A linear least-squares fit to the experimental data gives the semi-empirical oxygen barometer in terms of divergence from the fayalite-magnetite-quartz (FMQ) buffer:

Synthesis, crystal structure and crystal chemistry of ferri-clinoholmquistite, ?Li 2Mg 3Fe 3+ 2Si 8O 22(OH) 2

This work reports the synthesis of ferri-clinoholmquistite, nominally _Li2(Mg₃Fe³⁺₂)Si₈O₂₂(OH)₂, at varying f_O2 conditions. Amphibole compositions were characterized by X-ray (powder and single-crystal) diffraction, microchemical (EMPA) and spectroscopic (FTIR, Mössbauer and Raman) techniques. Under reducing conditions ( NNO+1, where NNO = Nickel–Nickel oxide buffer), the amphibole yield is very high (>90%), but its composition, and in particular the FeO/Fe₂O₃ ratio, departs significantly from the nominal one. Under oxidizing conditions ( NNO+1.5), the amphibole yield is much lower (<60%, with Li-pyroxene abundant), but its composition is close to the ideal stoichiometry. The exchange vector of relevance for the studied system is ^M2(Mg,Fe²⁺) ^M4(Mg,Fe²⁺) ^M2Fe³⁺_–1 ^M4Li_–1, which is still rather unexplored in natural systems. Amphibole crystals of suitable size for structure refinement were obtained only at 800 °C, 0.4 GPa and NNO conditions (sample 152), and have C2/m symmetry. The X-ray powder patterns for all other samples were indexed in the same symmetry; the amphibole closest to ideal composition has a = 9.428(1) Å, b = 17.878(3) Å, c = 5.282(1) Å, = 102.06(2)°, V = 870.8(3) ų. Mössbauer spectra show that Fe³⁺ is strongly ordered at M2 in all samples, whereas Fe²⁺ is disordered over the B and C sites. FTIR analysis shows that the amount of ^CFe²⁺ increases for increasingly reducing conditions. FTIR data also provide strong evidence for slight but significant amounts of Li at the A sites.     

Fe-Ti oxide geothermometry: thermodynamic formulation and the estimation of intensive variables in silicic magmas

A new thermodynamic formulation of the Fe–Ti oxide geothermometer/oxygen barometer is developed. The method is based upon recently calibrated models for spinel solid solutions in the quinary system (Fe²⁺, Mg)(Al,Fe³⁺,Cr)₂O₄–(Fe²⁺, Mg)₂TiO₄ by Sack and Ghiorso, and rhombohedral oxides in the quaternary system (Fe²⁺,Mg,Mn)TiO₃–Fe₂O₃ (this paper). The formulation is internally consistent with thermodynamic models for (Fe²⁺,Mg)-olivine and -orthopyroxene solid solutions and end-member thermodynamic properties tabulated by Berman. The constituent expressions account for compositional and temperature dependent cation ordering and reproduce miscibility gap features in all of the component binaries. The calibration does not account for the excess Gibbs energy resulting from compositional and temperature dependent magnetic ordering in either phase. This limits application of the method to assemblages that equilibrated at temperatures above 600° C. Practical implementation of the proposed geothermometer/oxygen barometer requires minimal use of projection algorthms in accommodating compositions of naturally occurring phases. The new formulation is applied to the estimation of temperature and oxygen fugacity in a wide variety of intermediate to silicic volcanic rocks. In combination with previous work on olivine and orthopyroxene thermodynamics, equilibration pressures are computed for a subset of these volcanics that contain the assemblage quartz, oxides and either ferromagnesian silicate. The calculated log₁₀ f _{O 2}-T relations are reflected in coexisting ferromagnesian mineral assemblages. Volcanics with the lowest relative oxygen fugacity (log₁₀ f _{O 2}) are characterized by the assemblage olivine-quartz, those with slightly higher log₁₀ f _{O 2} s, by the assemblage orthopyroxene-quartz. The sequence proceeds with the necessary phases biotite-feldspar, then hornblende-quartz-clinopyroxene, and finally at the highest log₁₀ f _{O 2} s, sphene-quartz-clinopyroxene. Quantitative analysis of these trends, utilizing thermodynamic data for the constituent phases, establishes that, in most cases, the T-log₁₀ f _{O 2}value computed from the oxides is consistent with the compositions of coexisting silicate phases, indicating that phenocryst equilibrium was achieved prior to eruption. There is, however, considerable evidence of oxide-silicate disequilibrium in samples collected from more slowly cooled domes and obsidians. In addition, T-log₁₀ f _{O 2}trends from volcanic rocks that contain biotite and orthopyroxene are interpreted to imply a condition of Fe²⁺–Mg exchange disequilibrium between orthopyroxene and coexisting ferromagnesian silicates and melt. It is suspected that many biotite-feldspar-quartz-orthopyroxene bearing low temperature volcanic rocks inherit orthopyroxene xenocrysts which crystallized earlier in the cooling history of the magma body.The problem is probably at least as complex as that of the feldspars... A.F. Buddington (1956)     

The stability of primary alluaudites in granitic pegmatites: an experimental investigation of the Na2(Mn2−2x Fe1+2x )(PO4)3 system

In order to assess the geothermometric potential of the Na₂(Mn_2−2x Fe_1+2x )(PO₄)₃ system (x = 0–1), which represents the compositions of natural weakly oxidized alluaudites, we performed hydrothermal experiments between 400 and 800°C, at 1 kbar, under an oxygen fugacity (f(O₂)) controlled by the Ni–NiO (NNO), Fe₂O₃–Fe₃O₄ (HM), Cu₂O–CuO (CT), and Fe–Fe₃O₄ (MI) buffers. When f(O₂) is controlled by NNO, single-phase alluaudites crystallize at 400 and 500°C, whereas the association alluaudite + marićite appears between 500 and 700°C. The limit between these two fields corresponds to the maximum temperature that can be reached by alluaudites in granitic pegmatites, because marićite has never been observed in these geological environments. Because alluaudites are very sensitive to variations of oxygen fugacity, the field of hagendorfite, Na₂MnFe²⁺Fe³⁺(PO₄)₃, has been positioned in the f(O₂)–T diagram, and provides a tool that can be used to estimate the oxygen fugacity conditions that prevailed in granitic pegmatites during the crystallization of this phosphate.     

The oxidation state of lamproitic magmas

Summary Olivine leucitites from the Gaussberg volcano, Antarctica are primitive members of the lamproite group of ultrapotassic rocks. They are glass rich, have an Mg number of around 70, and carry spinel lherzolite xenoliths. Liquidus phase fields and compositions were studied experimentally at atmospheric pressure with controlled oxygen fugacities. Chrome-spinels occur as inclusions in olivines in the natural rock, but it was necessary to add Cr₂O₃ to the experiments to stabilize spinels at the liquidus, indicating that some fractionation of spinel has almost certainly occurred.Experimental results show thatfO₂ conditions of crystallization can be characterized by ferric value (100 Fe³⁺/(Fe³⁺+Fe²⁺)) of spinel, ferric iron content of leucite, and Mg-number (100 Mg/(Mg+Fe²⁺)) of olivine. The results demonstrate that the liquidus phases of the Gaussberg rocks crystallized atfO₂ slightly below that of the NNO buffer. Application of the results to other lamproites indicates that they began to crystallize at oxygen fugacities varying from well above NNO (Leucite Hills) to around MW (West Kimberley and the Spanish fortunites). The Gaussberg olivine leucitite contains leucite cores poor in ferric iron with rims richer in ferric iron, indicating oxidation during emplacement. The ferric value of spinel is very sensitive to changes in oxygen fugacity and recognizing that some lamproitic magmas are known to contain diamonds, it may prove to be useful as a diamond survival indicator. The preservation of diamonds in lamproitic rocks will depend critically onfO₂: diamonds are not likely to be preserved in rocks which reach the surface as liquids atfO₂ near NNO or above.

---

Der Oxidationsgrad lamproitischer Magmen

Zusammenfassung Die Olivin-Leuzitite vom Gaussberg Vulkan, Antarktis, gehören zu den primären Gliedern der Lamproit-Gruppe der ultra-kalireichen Gesteine. Sie sind reich an Glas, weisen Mg-Werte um 70 auf, und führen Spinell-Lherzolith Xenolithe. Die Liquidus-Phasengleichgewichte bei Atmosphärendruck unter kontrollierten Sauerstoff-Partialdruckbedingungen wurden untersucht. Um Chromit als Liquidusphase zu stabilisieren, war es erforderlich, Cr₂O₃ zu den experimentellen Zusammensetzungen hinzuzufügen, obwohl Chromspinell als Einschluß in Olivin in den Ausgangsgesteinen auftritt; dies deutet eine Spinell-Franktionierung an.Die Resultate zeigten, daß diefO₂-Bedingungen bei der Kristallisation durch das Fe³⁺-Verhältnis (100 Fe³⁺/(Fe³⁺+Fe²⁺)) im Spinell, durch den Fe³⁺-Gehalt im Leuzit, und durch den Mg-Wert (100 mg/(Mg+Fe²⁺)) im Olivin charakterisiert sind. Die Liquidusphasen der Gaussberg Gesteine kristallisierten beifO₂-Bedingungen etwas unterhalb NNO. Die Anwendung der Resultate auf andere Lamproite zeigt, daß diese unter Sauerstoff-Partialdruckbedingungen kristallisierten, die von oberhalb NNO (Leucite Hills) bis etwa MW (West Kimberley und die Spanish Fortunites) variierten. Der Gaussberg Olivin-Leuzitit enthält Leuzitkristalle mit deutlicher Zonierung; die Kernbereiche sind arm an dreiwertigem Eisen, während die Randzonen deutlich an Fe³⁺ angereichert sind. Dies weist auf eine Oxidation des Magmas während der Platznahme hin. Das Fe³⁺-Verhältnis der Spinelle ist sehr empfindlich gegenüber Änderungen des Sauerstoff-Partialdruckes; da einige lamproitische Magmen Diamanten führen, könnte sich dieses als ein sinnvoller Indikator für eine mögliche Diamantführung erweisen. Die Überlebensfähigkeit von Diamanten in lamproitischen Gesteinen wird sehr stark abhängig sein vonfO₂: Magmen, die die Oberfläche unterfO₂-Bedingungen entsprechend NNO oder darüber erreichten, werden keine Diamanten mehr erhalten.

---

---

With 7 Figures     

The origin of the high-K latites from Camp Creek,Arizona: constraints from experiments with variable fO2 and a_{{\text{H}}_{\text{2}} {\text{O}}}

Near-liquidus melting experiments were performed on a high-K latite at fO₂'s ranging from iron-wustite-graphite (IWG) to nickel-nickel oxide (NNO) in the presence of a C-O-H fluid phase. Clinopyroxene is a liquidus phase under all conditions. At IWG , the liquidus at 10 kb is about 1,150° C but is depressed to 1,025° C at NNO and . Phlogopite and apatite are near-liquidus phases, with apatite crystallizing first at pressures below 10 kb. Phlogopite is a liquidus phase only at NNO and high . Under all conditions the high-K latites show a large crystallization interval with phlogopite becoming the dominant crystalline phase with decreasing temperature. Increasing fO₂ affects phlogopite crystallization but the liquidus temperature is essentially a function of . The chemical compositions of the near-liquidus phases support formation of the high-K latites under oxidizing conditions (NNO or higher) and high . It is concluded from the temperature of the H₂O-saturated liquidus at 10 kb, the groundmass: crystal ratio and presence of chilled latite margins around some xenoliths that the Camp Creek high-K latite magma passed thru the lower crust at temperatures of 1,000° C or more.     

Petrological and Experimental Constraints on the Pre-eruption Conditions of Holocene Dacite from Volcan San Pedro (36{degrees}S, Chilean Andes) and the Importance of Sulphur in Silicic Subduction-related Magmas

We present an experimental and petrological study aimed at estimatingthe pre-eruptive conditions of a Holocene dacitic lava fromVolcán San Pedro (36°S, Chilean Andes). Phase-equilibriumexperiments were performed at temperatures (T) from 800 to 950°C,and mainly at 200 MPa, but also at 55, 150, and 406 MPa. Oxygenfugacity (fO₂) ranged from the Ni–NiO buffer (NNO) to3·5 log units above (NNO + 3·5), and water contentsfrom     

Pre-eruptive Conditions of the Huerto Andesite (Fish Canyon System, San Juan Volcanic Field, Colorado): Influence of Volatiles (C-O-H-S) on Phase Equilibria and Mineral Composition

Crystallization experiments at 400 MPa, oxidized condition (logfO₂= NNO + 1, where NNO is nickel–nickel oxide buffer) andover a range of temperatures (850–950°C) and fluidcomposition (XH₂O_in = 0·3–1) have been carriedout to constrain the storage conditions of the sulphur-richmagma of the Huerto Andesite (an anhydrite, pyrrhotite, andS-rich apatite-bearing, post-Fish Canyon Tuff mafic lava). Theresults are used to evaluate the role of fluids released fromthe crystallization of magmas such as the Huerto Andesite onthe remobilization of the largely crystallized dacitic FishCanyon magma body. Experiments were performed using the naturalandesitic bulk composition with and without added sulphur. Thepresence of sulphur slightly affects the phase equilibria bychanging the phase proportions, stability fields of plagioclase,pyroxenes and ilmenite, and also affects the plagioclase composition.Phase equilibria and mineral composition data indicate thatthe magma may have contained 4·5 wt % water in the meltand that the pre-eruptive temperature was 875 ± 25°C.Assuming that the magma was in equilibrium with a fluid phase,the CO₂ concentration of the melt is estimated to be in therange 2000–4000 ppm (at 400 MPa). Before eruption, theandesite had an oxidation state very close to, or slightly within,the co-stability field of anhydrite–pyrrhotite at NNO+ 1·1. At these conditions, the sulphur content in themelt is 500 ppm. Assuming open-system degassing resulting fromcontinuing crystallization at depth, most of the CO₂ dissolvedin the andesitic melt should be released after the crystallizationof <10 vol. % of the magma, corresponding to a cooling from875 to 825–850°C. Thus, the fluids released owingto crystallization processes should be mainly composed of waterat temperatures below 825°C. KEY WORDS: experimental study; andesite; volatile; Fish Canyon Tuff; Huerto Andesite     

Macrocryst Fe - Ti oxides in olivine melilitites from Namaqualand-Bushmanland South Africa

Ilmenite macrocrysts in olivine melilitites from Namaqualand-Bushmanland, South Africa, have decomposed by subsolidus reduction to form oriented Mg-titanomagnetite along {0001} ilmenite planes. Residual ilmenite contains 10–11 wt% MgO, 1 wt% MnO, and 0.1 wt% Cr₂O₃. This macrocryst assemblage is mantled by an annulus of Mg-titanomagnetite, followed by an overgrowth of radiating magnetite + perovskite. Terminal compositions of these magnetites are similar to groundmass spinels, and to the outermost margins of magnetite macrocrysts that have very high Fe³⁺ core contents. The assemblages are remarkably similar to oxide intergrowths in kimberlites and an upper mantle derivation is proposed for ilmenite macrocrysts in these melilitites. Oxidation states in the source regions are also very similar, whether on-or off-craton, being slightly above FMQ (NNO), but reduced to FMQWM with the onset of decompression, volatile loss, and carbonate immiscibility. In the case of the melilitites, late stage, low pressure crystallization above NNO precipated abundant magnetite + perovskite. The oxide fO₂ data are consistent with, and refine the fO₂ estimates obtained previously for the behavior of Fe/Mg and Ni contents in olivine from the same suite of samples.     

Oxidised phase relations of a primitive basalt from Grenada,Lesser Antilles

A series of liquidus determinations is reported for a primitive arc basalt (15.4 wt % MgO, 45.5 wt % SiO₂) from Grenada, Lesser Antilles, at anhydrous, H₂O-undersaturated and H₂O-saturated conditions in the pressure range 1 atm to 1.7 GPa. \(\hbox{Fe}^{3+}/\Upsigma\hbox{Fe}\) of high-pressure experimental glasses as measured by μXANES ranges from 0.44 to 0.86, corresponding to oxygen fugacities (fO₂) between 3.2 and 7.8 log units above the nickel–nickel oxide redox buffer (NNO). 1-atm experiments conducted from NNO ? 2.5 to + 3.8 show that increasing fO₂ mainly increases the forsterite content (Fo) of olivine and has little effect on phase relations. The crystallisation sequence at lower crustal pressures for all water contents is forsteritic olivine + Cr-rich spinel followed by clinopyroxene. The anhydrous liquidus is depressed by 100 and 120 °C in the presence of 2.9 and 3.8 wt % H₂O, respectively. H₂O-undersaturated experiments at NNO + 3.2 to + 4.5 produce olivine of equivalent composition to the most primitive olivine phenocrysts in Grenadan picrites (Fo_91.4). We conclude that direct mantle melts originating beneath Grenada could be as oxidised as ~NNO + 3, consistent with the uppermost estimates from olivine–spinel oxybarometry of high Mg basalts. μXANES analyses of olivine-bearing experimental glasses are used to develop a semi-empirical oxybarometer based on the value of \({{K}_{D}}_{\rm ol-melt}^{\rm Fe-Mg}\) when all Fe is assumed to be in the Fe²⁺ state (\({K}_{D}^{{\rm Fe}_T}\)). The oxybarometer is tested on an independent data set and is able to reproduce experimental fO₂ to ≤1.2 log units. Experiments also show that the geochemically and petrographically distinct M- and C-series lavas on the island can be produced from hydrous melting of a common picritic source. Low pressures expand the olivine stability field at the expense of clinopyroxene, enriching an evolving melt in CaO and forcing differentiation to take place along a C-series liquid line of descent. Higher pressure conditions allow early and abundant clinopyroxene crystallisation, rapidly depleting the melt in both CaO and MgO, and thus creating the M-series.     

Experimental Crystallization of a High-K Arc Basalt: the Golden Pumice, Stromboli Volcano (Italy)

The near-liquidus crystallization of a high-K basalt (PST-9golden pumice, 49·4 wt % SiO₂, 1·85 wt % K₂O,7·96 wt % MgO) from the present-day activity of Stromboli(Aeolian Islands, Italy) has been experimentally investigatedbetween 1050 and 1175°C, at pressures from 50 to 400 MPa,for melt H₂O concentrations between 1·2 and 5·5wt % and NNO ranging from –0·07 to +2·32.A drop-quench device was systematically used. AuPd alloys wereused as containers in most cases, resulting in an average Feloss of 13% for the 34 charges studied. Major crystallizingphases include clinopyroxene, olivine and plagioclase. Fe–Tioxide was encountered in a few charges. Clinopyroxene is theliquidus phase at 400 MPa down to at least 200 MPa, followedby olivine and plagioclase. The compositions of all major phasesand glass vary systematically with the proportion of crystals.Ca in clinopyroxene sensitively depends on the H₂O concentrationof the coexisting melt, and clinopyroxene Mg-number shows aweak negative correlation with NNO. The experimental data allowthe liquidus surface of PST-9 to be defined. When used in combinationwith melt inclusion data, a consistent set of pre-eruptive pressures(100–270 MPa), temperatures (1140–1160°C) andmelt H₂O concentrations is obtained. Near-liquidus phase equilibriaand clinopyroxene Ca contents require melt H₂O concentrations<2·7–3·6 and 3 ± 1 wt %, respectively,overlapping with the maximum frequency of glass inclusion data(2·5–2·7 wt % H₂O). For olivine to crystallizeclose to the liquidus, pressures close to 200 MPa are needed.Redox conditions around NNO = +0·5 are inferred fromclinopyroxene compositions. The determined pre-eruptive parametersrefer to the storage region of golden pumice melts, which islocated at a depth of around 7·5 km, within the metamorphicarc crust. Golden pumice melts ascending from their storagezone along an adiabat will not experience crystallization ontheir way to the surface. KEY WORDS: basalt; pumice; experiment; phase equilibria; Stromboli     

Resetting of oxybarometers and oxygen isotope ratios in granulite facies orthogneisses during cooling and shearing,Adirondack Mountains,New York

The petrography, petrology, and oxygenisotope geochemistry of granulite-facies granitic and syenitic orthogneisses of the Diana and Stark complexes, Adirondack Mountains, New York, show that the extent and nature of resetting of isotopic and mineralogic systems is highly variable. There is a strong correlation between retrogression and shearing, and the rocks may be divided texturally into: (1) unsheared lithologies that preserve little-retrogressed pyroxene-or hornblendebearing peak-metamorphic mineralogies; and (2) sheared rocks that underwent retrogression, marked by the growth of late biotite, in centimetre-to metre-wide shear zones after the peak of metamorphism. Oxygen fugacities in the unsheared lithologies were estimated for reintegrated mineral compositions from magnetiteilmenite (Mt-Ilm) and ferrosilite-magnetic-quartz (Fs-Mt-Qtz) equilibria. Mt-Ilm yields logfO_2Mt-Ilm values of-15.9 to-17.6 (0.6 to 1.3 log units below the fayalite-magnetite-quartz buffer, FMQ) and temperatures of 670–745°C that agree with those from other geothermometry and phase equilibria studies. These data suggest that, aside from oxyexsolution of ilmenite from magnetite, the Fe-Ti system underwent only minor resetting during cooling, and the Fe-Ti oxides yield good estimates of peak-metamorphic temperatures and fO₂. In unsheared ilmenite + magnetite + orthopyroxene + quartz assemblages, values of logfO_2Mt-Ilm are lower than logfO_2Fs-Mt-Qtz by an average of 0.6 when the orthopyroxene activity model of Sack and Ghiorso is used. Minor resetting of the Fe-Ti oxides, analytical errors, and errors in the placement of end-member reactions probably account for this relatively small difference in fO₂ values. Whole-rock ¹⁸O values of unsheared Diana and Stark lithologies range from 4.0 to 10.3 reflecting pre-regional metamorphic oxygen-isotope ratios. Peak-metamorphic minerals preserve high-temperature oxygen-isotope fractionations, and, in many samples, the effective diffusion of oxygen in minerals ceased at higher temperatures than predicted from wet experimental diffusion data. These data suggest that the rocks did not contain an aqueous fluid phase during cooling. The combination of petrologic, isotopic, and textural data also permits a detailed study of shearing and retrogression. Ilmenites in the sheared lithologies underwent greater degrees of hematite loss than in the unsheared rocks, resulting in logfO_2Mt-Ilm values as low as-24.1 (3.1 log units below FMQ) and Mt-Ilm temperatures that are up to 175°C below regional estimates. Sheared rocks also have higher ¹⁸O values (up to 13.3). During shearing, ¹⁸O values of biotite, K-feldspar, and magnetite reset readily, while the degree of isotopic resetting of quartz correlates with the intensity for recrystallization.This paper is a contribution to IGCP Project 304, Lower Crustal Processes     

Experimental and Thermodynamic Constraints on the Sulphur Yield of Peralkaline and Metaluminous Silicic Flood Eruptions

Many basaltic flood provinces are characterized by the existenceof voluminous amounts of silicic magmas, yet the role of thesilicic component in sulphur emissions associated with trapactivity remains poorly known. We have performed experimentsand theoretical calculations to address this issue. The meltsulphur content and fluid/melt partitioning at saturation witheither sulphide or sulphate or both have been experimentallydetermined in three peralkaline rhyolites, which are a majorcomponent of some flood provinces. Experiments were performedat 150 MPa, 800–900°C, fO₂ in the range NNO –2 to NNO + 3 and under water-rich conditions. The sulphur contentis strongly dependent on the peralkalinity of the melt, in additionto fO₂, and reaches 1000 ppm at NNO + 1 in the most stronglyperalkaline composition at 800°C. At all values of fO₂,peralkaline melts can carry 5–20 times more sulphur thantheir metaluminous equivalents. Mildly peralkaline compositionsshow little variation in fluid/melt sulphur partitioning withchanging fO₂ (D_S 270). In the most peralkaline melt, D_S risessharply at fO₂ > NNO + 1 to values of >500. The partitioncoefficient increases steadily for S_bulk between 1 and 6 wt% but remains about constant for S_bulk between 0·5 and1 wt %. At bulk sulphur contents lower than 4 wt %, a temperatureincrease from 800 to 900°C decreases D_S by 10%. These results,along with (1) thermodynamic calculations on the behaviour ofsulphur during the crystallization of basalt and partial meltingof the crust and (2) recent experimental constraints on sulphursolubility in metaluminous rhyolites, show that basalt fractionationcan produce rhyolitic magmas having much more sulphur than rhyolitesderived from crustal anatexis. In particular, hot and dry metaluminoussilicic magmas produced by melting of dehydrated lower crustare virtually devoid of sulphur. In contrast, peralkaline rhyolitesformed by crystal fractionation of alkali basalt can concentrateup to 90% of the original sulphur content of the parental magmas,especially when the basalt is CO₂-rich. On this basis, we estimatethe amounts of sulphur potentially released to the atmosphereby the silicic component of flood eruptive sequences. The peralkalineEthiopian and Deccan rhyolites could have produced 10¹⁷ and10¹⁸ g of S, respectively, which are comparable amounts to publishedestimates for the basaltic activity of each province. In contrast,despite similar erupted volumes, the metaluminous Paraná–Etendekasilicic eruptives could have injected only 4·6 x 10¹⁵g of S in the atmosphere. Peralkaline flood sequences may thushave greater environmental effects than those of metaluminousaffinity, in agreement with evidence available from mass extinctionsand oceanic anoxic events. KEY WORDS: silicic flood eruptions; sulphur; experiment; Ethiopia; Deccan     

Experimental investigation of large-ion-lithophile-element-, high-field-strength-element- and rare-earth-element-partitioning between calcic amphibole and basaltic melt: the effects of pressure and oxygen fugacity

The effects of pressure and oxygen fugacity (fO₂) on trace element partitioning between pargasitic amphibole and alkali-basaltic melts have been determined at pressures from 1.5 to 2.5 GPa and oxygen fugacities at 2 log units above and below the nickel–nickel oxide buffer. Amphibole crystallization experiments were performed in a piston cylinder apparatus and partition coefficients between amphibole and quenched melt of large-ion-lithophile elements (LILE: Rb, Sr, Ba), high-field-strength elements (HFSE: Zr, Nb, Ta, Hf, U, Th) and rare-earth elements (REE: La to Lu; +Y) were measured with a LASER ablation inductively coupled plasma – mass spectrometer. Increasing pressure from 1.5 to 2.5 GPa at similar temperatures and approximately constant fO₂ increases D _Rb but decreases D _Zr and D _Hf and D _REE (D _La, D _Ce, D _Pr). An empirical relationship was observed between D _Zr and (Ti/Al)_M2 in the amphibole, which can be described by:

Ferric Iron in CaTiO3 Perovskite as an Oxygen Barometer for Kimberlitic Magmas I: Experimental Calibration

A method to estimate the oxygen fugacity (fO₂) during the crystallizationof kimberlites is developed using the Fe content of CaTiO₃ perovskite(Pv), a common groundmass phase in these rocks. With increasingfO₂, more Fe exists in the kimberlitic liquid as Fe³⁺, and thuspartitions into Pv. Experiments to study the partitioning ofFe between Pv and kimberlite liquid were conducted at 100 kPaon simple and complex anhydrous kimberlite bulk compositionsfrom 1130 to 1300°C over a range of fO₂ from NNO –5 to NNO + 4 (where NNO is the nickel–nickel oxide buffer),and at Nb and rare earth element (REE) contents in the startingmaterials of 0–5 wt % and 1500 ppm, respectively. Thepartitioning of Fe between Pv and kimberlite liquid is influencedmostly by fO₂, although the presence of Nb increases the partitionof Fe³⁺ into perovskite at a given T and fO₂. Multiple linearregression (MLR) of all the experimental data produces a relationshipthat describes the variation of Fe and Nb in Pv with fO₂ relativeto the NNO buffer:

(uncertaintiesat 2, and Nb and Fe as cations per three oxygens). Over therange of conditions of our experiments, this relationship showsno temperature (T) dependence, is not affected by the bulk Fecontent of the kimberlite starting material and reproduces experimentaldata to within 1 log fO₂ unit. KEY WORDS: kimberlites; oxygen fugacity; perovskite; ferric iron; magma     

Rhyodacite magma storage conditions prior to the 3430 yBP caldera-forming eruption of Aniakchak volcano, Alaska

This study presents a pre-eruptive magma storage model for the rhyodacite and andesite magmas erupted during the 3430 yBP caldera-forming eruption of Aniakchak volcano, Alaska, derived from phase equilibria experiments and petrological data. The compositions of Fe–Ti oxide pairs from the early erupted Plinian rhyodacite pumice yield core temperatures of 871–900°C, with rims up to ∼942°C, and fO₂ from −10.6 to −11.8 log units. Melt inclusions entrapped in plagioclase phenocrysts have H₂O contents between 3 and 5 wt%, estimated by FTIR and electron microprobe volatiles by difference methods, with no detectable CO₂. Assuming water saturation, this corresponds to entrapment pressures between ∼65 and 150 MPa. Phase equilibria results reproduce the natural phase assemblages at of 95–150 MPa at 870–880°C, assuming water saturation. A mismatch in experimental versus natural glass SiO₂ and Al₂O₃, and MELTS models for H₂O-undersaturated conditions indicate that the rhyodacite may not have been H₂O saturated. MELTS models with and P _total of 125–150 MPa at 870–880°C reproduce the natural groundmass glass Al₂O₃ composition best, indicating the magma may have been slightly H₂O undersaturated. Those pressures correspond to storage at 4.5–5.4 km depth in the crust. MELTS models and VBD estimates from melt inclusions in titanomagnetite grains from the andesite indicate pre-eruptive conditions of ∼1,000°C and > 110 MPa, corresponding to a minimum residence depth of ∼4.1 km assuming water saturation or greater if the magma was H₂O undersaturated. Previous geochemical studies indicate separate histories of the two magmas, though they retain some evidence that they are ultimately related through fractional crystallization processes. Analogous to the 1912 Novarupta magmas, the rhyodacite and andesite presumably originated within the same crystal mush zone beneath the edifice, yet were separated laterally and underwent different degrees of crustal assimilation. The andesite must have resided in close proximity, with ascent occurring in response to movement of the rhyodacite, and resulting in extensive syn-eruptive mingling.Electronic Supplementary Material Supplementary material is available for this article at and is accessible for authorized users.