An experimental investigation on diffusion of water in haplogranitic melts

  The diffusivity of water has been investigated for a haplogranitic melt of anhydrous composition Qz₂₈Ab₃₈Or₃₄ (in wt %) at temperatures of 800–1200°C and at pressures of 0.5–5.0 kbar using the diffusion couple technique. Water contents of the starting glass pairs varied between 0 and 9 wt %. Concentration-distance profiles for the different water species (molecular water and hydroxyl groups) were determined by near-infrared microspectroscopy. Because the water speciation of the melt is not quenchable (Nowak 1995; Nowak and Behrens 1995; Shen and Keppler 1995), the diffusivities of the individual species can not be evaluated directly from these profiles. Therefore, apparent chemical diffusion coefficients of water (D _water) were determined from the total water profiles using a modified Boltzmann-Matano analysis. The diffusivity of water increases linearly with water content <3 wt % but exponentially at higher water contents. The activation energy decreases from 64 ± 10 kJ/mole for 0.5 wt % water to 46 ± 5 kJ/mole for 4 wt % water but remains constant at higher water contents. A small but systematic decrease of D _water with pressure indicates an average activation volume of about 9 cm³/mole. The diffusivity (in cm²/s) can be calculated for given water content (in wt %), T (in K) and P (in kbar) by

The pressure and temperature conditions and timing of glass formation in mantle-derived xenoliths from Baarley, West Eifel, Germany: the case for amphibole breakdown, lava infiltration and mineral – melt reaction

Summary Mantle-derived xenoliths from Baarley in the Quaternary West Eifel volcanic field contain six distinct varieties of glass in veins, selvages and pools. 1) Silica-undersaturated glass rich in zoned clinopyroxene microlites that forms jackets around and veins within the xenoliths. This glass is compositionally similar to groundmass glass in the host basanite. 2) Silica-undersaturated alkaline glass that contains microlites of Cr-diopside, olivine and spinel associated with amphibole in peridotites. This glass locally contains corroded primary spinel and phlogopite. 3) Silica-undersaturated glass associated with diopside, spinel ± olivine and rh?nite microlites in partly to completely broken down amphibole grains in clinopyroxenites. 4) Silica-undersaturated to silica-saturated, potassic glass in microlite-rich fringes around phlogopite grains in peridotite. 5) Silica-undersaturated potassic glass in glimmerite xenoliths. 6) Silica-rich glass around partly dissolved orthopyroxene crystals in peridotites. Geothermometry of orthopyroxene–clinopyroxene pairs (P = 1.5 GPa) gives temperatures of ∼ 850 °C for unveined xenoliths to 950–1020 °C for veined xenoliths. Clinopyroxene – melt thermobarometry shows that Cr-diopside – type 2 glass pairs in harzburgite formed at 1.4 to 1.1 GPa and ∼ 1250 °C whereas Cr-diopside – type 2 glass pairs in wehrlite formed at 0.9 to 0.7 GPa and 1120–1200 °C. This bimodal distribution in pressure and temperature suggests that harzburgite xenoliths may have been entrained at greater depth than wehrlite xenoliths. Glass in the Baarley xenoliths has three different origins: infiltration of an early host melt different in composition from the erupted host basanite; partial melting of amphibole; reaction of either of these melts with xenolith minerals. The composition of type 1 glass suggests that jackets are accumulations of relatively evolved host magma. Mass balance modelling of the type 2 glass and its microlites indicates that it results from breakdown of disseminated amphibole and reaction of the melt with the surrounding xenolith minerals. Type 3 glass in clinopyroxenite xenoliths is the result of breakdown of amphibole at low pressure. Type 4 and 5 glass formed by reaction between phlogopite and type 2 melt or jacket melt. Type 6 glass associated with orthopyroxene is due to the incongruent dissolution of orthopyroxene by any of the above mentioned melts. Compositional gradients in xenolith olivine adjacent to type 2 glass pools and jacket glass can be modelled as Fe–Mg interdiffusion profiles that indicate melt – olivine contact times between 0.5 and 58 days. Together with the clinopyroxene – melt thermobarometry calculations these data suggest that the glass (melt) formed over a short time due to decompression melting of amphibole and infiltration of evolved host melt. None of the glass in these xenoliths can be directly related to metasomatism or any other process that occurred insitu in the mantle. Received November 23, 1999; revised version accepted September 5, 2001     

Mineral chemistry of submarine lavas from Hilo Ridge, Hawaii: implications for magmatic processes within Hawaiian rift zones

The crustal history of volcanic rocks can be inferred from the mineralogy and compositions of their phenocrysts which record episodes of magma mixing as well as the pressures and temperatures when magmas cooled. Submarine lavas erupted on the Hilo Ridge, a rift zone directly east of Mauna Kea volcano, contain olivine, plagioclase, augite ±orthopyroxene phenocrysts. The compositions of these phenocryst phases provide constraints on the magmatic processes beneath Hawaiian rift zones. In these samples, olivine phenocrysts are normally zoned with homogeneous cores ranging from ∼ Fo₈₁ to Fo₉₁. In contrast, plagioclase, augite and orthopyroxene phenocrysts display more than one episode of reverse zoning. Within each sample, plagioclase, augite and orthopyroxene phenocrysts have similar zoning profiles. However, there are significant differences between samples. In three samples these phases exhibit large compositional contrasts, e.g., Mg# [100 × Mg/(Mg+Fe⁺²)] of augite varies from 71 in cores to 82 in rims. Some submarine lavas from the Puna Ridge (Kilauea volcano) contain phenocrysts with similar reverse zonation. The compositional variations of these phenocrysts can be explained by mixing of a multiphase (plagioclase, augite and orthopyroxene) saturated, evolved magma with more mafic magma saturated only with olivine. The differences in the compositional ranges of plagioclase, augite and orthopyroxene crystals between samples indicate that these samples were derived from isolated magma chambers which had undergone distinct fractionation and mixing histories. The samples containing plagioclase and pyroxene with small compositional variations reflect magmas that were buffered near the olivine + melt ⇒Low-Ca pyroxene + augite + plagioclase reaction point by frequent intrusions of mafic olivine-bearing magmas. Samples containing plagioclase and pyroxene phenocrysts with large compositional ranges reflect magmas that evolved beyond this reaction point when there was no replenishment with olivine-saturated magma. Two of these samples contain augite cores with Mg# of ∼71, corresponding to Mg# of 36–40 in equilibrium melts, and augite in another sample has Mg# of 63–65 which is in equilibrium with a very evolved melt with a Mg# of ∼30. Such highly evolved magmas also exist beneath the Puna Ridge of Kilauea volcano. They are rarely erupted during the shield building stage, but may commonly form in ephemeral magma pockets in the rift zones. The compositions of clinopyroxene phenocryst rims and associated glass rinds indicate that most of the samples were last equilibrated at 2–3 kbar and 1130–1160 °C. However, in one sample, augite and glass rind compositions reflect crystallization at higher pressures (4–5 kbar). This sample provides evidence for magma mixing at relatively high pressures and perhaps transport of magma from the summit conduits to the rift zone along the oceanic crust-mantle boundary. Received: 8 July 1998 / Accepted: 2 January 1999     

Compositional variation and primary water contents of differentiated interstitial and included glasses in boninites

 Interstitial glasses and glasses in small inclusions in Mg-rich phenocrysts of 14 boninites from New Caledonia, the Mariana Trench, Cape Vogel and Chichi-jima were analysed by electron microprobe and the water contents measured in situ by ion microprobe. The glasses are remarkably fresh and abundant (∼30–90 vol.%), and the phenocrysts are often skeletal with glass inclusions. Broad-beam analyses (∼1030) of interstitial glasses and ∼180 point analyses of glass inclusions were carried out, as well as ∼100 hydrogen analyses. Most glasses have low water-free totals, high water contents, very low MgO, and low total iron; they are almost entirely quartzofeldspathic and with few exceptions (Q+or+ab+ an+C) lies in the range 83–96. The interstitial glasses from New Caledonia, the Marianas and most of the glasses from Chichi-jima are dacitic, those from Cape Vogel straddle dacitic and andesitic compositions, whereas the glasses in a highly glassy sample from Chichi-jima are high-Mg andesitic or boninitic with up ∼9 wt% MgO, and are, with the exception of a few high-Ca boninites from Tonga, the most Mg-rich interstitial glasses so far described in boninites. Glasses included in orthopyroxene, olivine or clinoenstatite are boninitic or high-Mg andesitic in the highly glassy rock and dacitic to high-silica dacitic in the others. They are in general slightly more differentiated than the interstitial glasses, because of more-extensive crystallization on the host crystal in small inclusions. The interstitial glass compositions show a direct relationship between silica and Al₂O₃ and, for most glasses, alkalis, and inverse relationships between silica and CaO, FeO and MgO; alkalis and TiO₂ show, however, a broad spread in values in glasses from the Marianas and New Caledonia. Included glasses show similar variations. Water contents in interstitial glasses are ∼2 wt% for the highly glassy high-Mg andesitic glasses from Chichi-jima, ∼5.4 wt% for the more differentiated andesitic to dacitic glasses from Cape Vogel, and ∼6.7–7.0 wt% in the most differentiated dacitic ones from the Marianas and New Caledonia. Water contents in glass inclusions in olivine, orthopyroxene and clinoenstatite are in the range ∼1.9–3.3 wt%. The interstitial glasses are black and not vesicular, showing that the liquids did not reach supersaturation after eruption on or intrusion near the sea floor, or were insufficiently so to allow nucleation of water vapour bubbles. The water is inferred to be primary and to increase strongly with crystallization in the residual liquid down to the glass-transition T. Received: 19 December 1994 / Accepted: 5 October 1995     

Evolution of parental magmas of Miocene shield basalts of Gran Canaria (Canary Islands): constraints from crystal,melt and fluid inclusions in minerals

 Picritic units of the Miocene shield volcanics on Gran Canaria, Canary Islands, contain olivine and clinopyroxene phenocrysts with abundant primary melt, crystal and fluid inclusions. Composition and crystallization conditions of primary magmas in equilibrium with olivine Fo_90-92 were inferred from high-temperature microthermometric quench experiments, low-temperature microthermometry of fluid inclusions and simulation of the reverse path of olivine fractional crystallization based on major element composition of melt inclusions. Primary magmas parental for the Miocene shield basalts range from transitional to alkaline picrites (14.7–19.3 wt% MgO, 43.2–45.7 wt% SiO₂). Crystallization of these primary magmas is believed to have occurred over the temperature range 1490–1150° C at pressures ≈5 kbar producing olivine of Fo_80.6-90.2, high-Ti chrome spinel [Mg/ (Mg+Fe²⁺)=0.32–0.56, Cr/(Cr+Al)=0.50–0.78, 2.52–8.58 wt% TiO₂], and clinopyroxene [Mg/(Mg+Fe)=0.79–0.88, Wo_44.1-45.3, En_43.9-48.0, Fs_6.8-11.0] which appeared on the liquidus together with olivine≈Fo₈₆. Redox conditions evolved from intermediate between the QFM and WM buffers to late-stage conditions of NNO+1 to NNO+2. The primary magmas crystallized in the presence of an essentially pure CO₂ fluid. The primary magmas originated at pressures >30 kbar and temperatures of 1500–1600° C, assuming equilibrium with mantle peridotite. This implies melting of the mantle source at a depth of ≈100 km within the garnet stability field followed by migration of melts into magma reservoirs located at the boundary between the upper mantle and lower crust. The temperatures and pressures of primary magma generation suggest that the Canarian plume originated in the lower mantle at depth ≈900 km that supports the plume concept of origin of the Canary Islands. Received: 23 October 1995/Accepted: 21 February 1996     

Rates of grain boundary diffusion through enstatite and forsterite reaction rims

 The growth rates of enstatite rims produced by reaction of Fo₉₂ and SiO₂ were determined at 250–1500 MPa and 900–1100°C for a wide range of water contents. Growth rates were also determined for forsterite rims between MgO and Mg₂Si₂O₆ and between MgO and SiO₂. Rim growth rates are parabolic indicating diffusion-controlled growth of the polycrystalline rims which are composed of ˜ 2 μm diameter grains. Rim growth rates were used to calculate the product of the grain boundary diffusion coefficient (D'_A) times the effective grain boundary thickness (δ) assuming in turn that MgO, SiO₂, and Mg₂Si₋₁ are the diffusing components (coupled diffusion of a cation and oxygen or interdiffusion of Mg and Si). The values for D'_MgOδ, D', and D' for enstatite at 1000°C and 700 MPa confining pressure with about 0.1 wt %  water are about five times larger than the corresponding D'_Aδ values for samples initially vacuum dried at 250°C. Most of the increase in D'_Aδ occurs with the first 0.1 wt %  water. The activation energy for diffusion through the enstatite rims (1100–950°C) is 162 ± 30 kJ/mole. The diffusion rate through enstatite rims is essentially unchanged for confining pressures from 210–1400 MPa, but the nucleation rate is greatly reduced at low confining pressure (for  ≤ 1.0 wt % water present) and limits the conditions at which rim growth can be measured. The corresponding values for D'_Aδ through forsterite rims are essentially identical for the two forsterite-producing reactions when 0.1 wt % water is added and similar to the D'_Aδ values for enstatite at the same conditions. The D'_Aδ values for forsterite are ˜ 28 times larger for samples starting with 0.1 wt %  water compared to samples that were first vacuum dried. Thus water enhances these grain boundary diffusion rates by a factor of 5–30 depending on the mineralogy, but the total range in D'_Aδ is only slightly more than an order of magnitude for as wide a range of water contents as expected for most crustal conditions. Received: 1 July 1995 / Accepted: 1 August 1996     

Experimental reconstruction of sodic dolomitic carbonatite melts from metasomatised lithosphere

 Investigations of peridotite xenolith suites have identified a compositional trend from lherzolite to magnesian wehrlite in which clinopyroxene increases at the expense of orthopyroxene and aluminous spinel, and in which apatite may be a minor phase. Previous studies have shown that this trend in mineralogy and chemical composition may result from reaction between sodic dolomitic carbonatite melt and lherzolite at pressures around 1.7 to 2 GPa. This reaction results in decarbonation of the carbonatite melt, releasing CO₂-rich fluid. In this study, we have experimentally reversed the decarbonation reaction by taking two natural wehrlite compositions and reacting them with CO₂ at a pressure of 2.2 GPa and temperatures from 900 to 1150° C. Starting materials were pargasite-bearing wehrlites, one with minor apatite (composition 71001^*) and one without apatite (composition 70965^*). At lower temperatures (900° C) the products were apatite+pargasite+magnesite harzburgite for runs using composition 71001^*, and pargasite+dolomite lherzolite for runs using composition 70965^*. At and above 1000° C, carbonatite melt with harzburgite residue (olivine+orthopyroxene+spinel) and with lherzolite residue (olivine+orthopyroxene+clinopyroxene+ spinel) were produced respectively. Phase compositions in reactants and products are consistent with the documented carbonatite/lherzolite reactions, and also permit estimation of the carbonatite melt compositions. In both cases the melts are sodic dolomitic carbonatites. The study supports the hypothesis of a significant role for ephemeral, sodic dolomitic melts in causing metasomatic changes in the lithosphere at P≤2 GPa. The compositions of wehrlites imply fluxes of CO₂, released by metasomatic reactions, which are locally very large at around 5 wt% CO₂. Received: 15 December 1995/Accepted: 14 February 1996     

Partial melting of metagreywackes, Part II. Compositions of minerals and melts

A series of experiments on the fluid-absent melting of a quartz-rich aluminous metagreywacke has been carried out. In this paper, we report the chemical composition of the phases present in the experimental charges as determined by electron microprobe. This analytical work includes biotite, plagioclase, orthopyroxene, garnet, cordierite, hercynite, staurolite, gedrite, oxide, and glass, over the range 100–1000 MPa, 780–1025 °C. Biotites are Na- and Mg-rich, with Ti contents increasing with temperature. The compositions of plagioclase range from An₁₇ to An₃₅, with a significant orthoclase component, and are always different from the starting minerals. At high temperature, plagioclase crystals correspond to ternary feldspars with Or contents in the range 11–20 mol%. Garnets are almandine pyrope grossular spessartine solid solutions, with a regular and significant increase of the grossular content with pressure. All glasses are silicic (SiO₂ = 67.6–74.4 wt%), peraluminous, and leucocratic (FeO + MgO = 0.9–2.9 wt%), with a bulk composition close to that of peraluminous leucogranites, even for degrees of melting as high as 60 vol.%. With increasing pressure, SiO₂ contents decrease while K₂O increases. At any pressure, the melt compositions are more potassic than the water-saturated granitic minima. The H₂O contents estimated by mass balance are in the range 2.5–5.6 wt%. These values are higher than those predicted by thermodynamic models. Modal compositions were estimated by mass balance calculations and by image processing of the SEM photographs. The positions of the 20 to 70% isotects (curves of equal proportion of melt) have been located in the pressure-temperature space between 100 MPa and 1000 MPa. With increasing pressure, the isotects shift toward lower temperature between 100 and 200 MPa, then bend back toward higher temperature. The melting interval increases with pressure; the difference in temperature between the 20% and the 70% isotects is 40 °C at 100 MPa, and 150 °C at 800 MPa. The position of the isotects is interpreted in terms of both the solubility of water in the melt and the nature of the reactions involved in the melting process. A comparison with other partial melting experiments suggests that pelites are the most fertile source rocks above 800 MPa. The difference in fertility between pelites and greywackes decreases with decreasing pressure. A review of the glass compositions obtained in experimental studies demonstrates that partial melting of fertile rock types in the crust (greywackes, pelites, or orthogneisses) produces only peraluminous leucogranites. More mafic granitic compositions such as the various types of calk-alkaline rocks, or mafic S-type rocks, have never been obtained during partial melting experiments. Thus, only peraluminous leucogranites may correspond to liquids directly formed by partial melting of metasediments. Other types of granites involve other components or processes, such as restite unmixing from the source region, and/or interaction with mafic mantle-derived materials. Received: 11 July 1995 / Accepted: 27 February 1997     

Stability of phlogopite in granitic melts, an experimental investigation

Water-saturated and water-undersaturated experiments (a _{H2 O} = 1.0 and 0.5) were performed in the temperature range 780–1040°C at 2 and 5 kbar in order to determine the upper thermal stability of phlogopite in granitic melts. Starting compositions were: (A) subaluminous mixtures of 20 wt % synthetic phlogopite and 80 wt % synthetic anhydrous haplogranitic glass; (B) peraluminous mixtures (normative corundum  = 4 %) of 20 wt % synthetic phlogopite and 80 wt % synthetic anhydrous peraluminous haplogranitic glass. The molar quartz: albite: orthoclase ratio of the glasses of the 2␣kbar runs was 35:39:26 and that of the 5 kbar runs 30:42:28. In the subaluminous system, phlogopite is stable up to 820°C at a _{H2 O} = 1.0 and up to 780°C at a _{H2 O} = 0.5. At higher temperatures, it is replaced by enstatite. In the peraluminous system phlogopite has a remarkably higher thermal stability (up to 1000°C at 5 kbar and a _{H2 O} = 1.0) and there is a temperature interval of 80°C at a _{H2 O} = 1.0, and 90–100°C at a _{H2 O} = 0.5 between the first appearance of enstatite and the disappearance of phlogopite. In the peraluminous system, phlogopite is a solid solution (ss) of phlogopite, muscovite, talc and eastonite components. The crystalline product of the phlogopite_ss breakdown reaction is an aluminous enstatite. The MgO-content of the melt depends on the normative corundum content of the starting material and the run temperature. It is independent of pressure. In the subaluminous system, the MgO-content ranges between 0.05 and 0.3 wt % in the temperature interval 780–880°C at both investigated water activities. The MgO-content of the peraluminous melts at a _{H2 O} = 1.0 ranges between 0.4 and 1.7 wt % and at a _{H2 O} = 0.5 between 0.2 and 1.4 wt % in the temperature range 780–980°C. Received: 28 August 1995 / Accepted: 6 August 1996     

Phenocryst complexity in andesites and dacites from the Tequila volcanic field,Mexico: resolving the effects of degassing vs. magma mixing

The petrology of five phenocryst-poor (2–5%) andesites and dacites, all of which were erupted from different short-lived, monogenetic vents, is compared to that of phenocryst-rich (10–25%) andesites erupted from the adjacent stratovolcano, Volcán Tequila, in the Mexican arc. Despite differences in phenocryst abundances, these magmas have comparable phase assemblages (plagioclase + orthopyroxene + titanomagnetite + ilmenite + apatite ± augite ± hornblende), and similarly wide variations in phenocryst compositions, coupled to complex zoning patterns. For the phenocryst-poor lavas, equilibrium pairs of two Fe–Ti oxides lead to a narrow range of calculated temperatures for each sample that range from 934 (±24) to 1,073 (±6)°C and oxygen fugacities that range from +0.1 to +0.7 log units relative to the Ni–NiO buffer. Application of the plagioclase-liquid hygrometer to each sample at these calculated temperatures leads to maximum melt water concentrations of 4.6–3.1 wt% during plagioclase crystallization, indicating that the magmas were fluid saturated at depths ≥6.4–4.5 km. There is a wide, continuous range in the composition of plagioclase (≤44 mol% An) and orthopyroxene (≤16% Mg#) phenocrysts in each sample, which is consistent with a loss of dissolved water (≤2.8 wt%) from the melt phase during degassing as the magmas ascended rapidly to the surface. Evidence is presented that shows the effect of dissolved water is to reduce the activity of MgO relative to FeO in the melt phase, which indicates that degassing will also affect the Mg# of pyroxene phenocrysts, with higher melt water concentrations favoring Fe-rich pyroxene. Both plagioclase and orthopyroxene commonly display diffusion-limited growth textures (e.g., skeletal and hopper crystals, large interior melt hollows, and swallow tails), which are consistent with large undercoolings produced by degassing-induced crystallization. Therefore, degassing is proposed as a possible cause for the phenocryst compositional diversity documented in the phenocryst-poor andesite and dacite lavas erupted from peripheral vents, including the coexistence of normally zoned plagioclase and reversely zoned orthopyroxene. Degassing-induced crystallization may also explain some of the phenocryst complexity in crystal-rich andesites erupted from large stratovolcanoes, including Volcán Tequila.     

The determination of partial melt compositions of peridotitic systems by melt inclusion synthesis

An experimental method of melt inclusion synthesis within olivine crystals has been developed to determine the composition of the melt present in a partially molten peridotite assemblage. Trace element doped peridotite was equilibrated with 5 wt% of a C-O-H volatile source at 20 kbar/1175 °C in a piston-cylinder apparatus under buffered oxygen and sulphur fugacity conditions [log(f _O2) ∼ IW +1 log unit, log (f _S2) ∼ Fe/FeS > +1 log unit]. A single crystal of olivine, which had been cut to a disc shape, was included in the sample capsule. At run conditions the peridotite charge formed olivine, orthopyroxene, clinopyroxene, Fe-Ni sulphide and a volatile-bearing melt. The melt phase is preserved as homogeneous glass inclusions up to 50 μm in size, trapped in situ in the olivine disc. The major element composition of the glass inclusions showed them to be of broadly basaltic character, but with a low Mg/(Mg + ΣFe), which is associated with precipitation of olivine from the melt inclusion onto the walls of the olivine disc during quenching. Thus the equilibrium melt composition has been calculated from the glass inclusion composition by addition of olivine component using the Fe/Mg exchange coefficient of Roeder and Emslie (1970); the desired Mg/(Mg + ΣFe) being determined from the composition of olivine formed at run conditions in the peridotite section of the charge. The melt composition obtained is close to the trend for dry melting established by Falloon and Green (1988), and it is evident that although the reduced volatiles in this case have induced a liquidus depression of some 250 °C, there has been only a small shift in melt composition. Trace element, carbon and hydrogen contents of thirteen melt inclusions have been determined by secondary ion mass spectrometry (SIMS). The trace element signature is consistent with ∼29% melting in equilibrium with a lherzolitic assemblage. The equilibrium melt has a C/H of 0.48 by weight. Carbon solubility in partial melts is thus significant under reducing conditions in the presence of dissolved “water components” and establishes a major melt fluxing role for carbon in the upper mantle. The ubiquitous presence of carbon and hydrogen in basaltic magmas underscores the importance of determining both the position of vapour-present solidi and the composition of melts generated, when developing petrogenetic models. Received: 1 July 1996 / Accepted: 25 June 1997     

An experimental study of the effects of melt composition on plagioclase-melt equilibria at 5 and 10 kbar: implications for the origin of magmatic high-An plagioclase

An experimental investigation of plagioclase crystallization in broadly basaltic/andesitic melts of variable Ca# (Ca/(Ca+Na)*100) and Al# (Al/(Al+Si)*100) values and H₂O contents has been carried out at high pressures (5 and 10 kbar) in a solid media piston-cylinder apparatus. The H₂O contents of glasses coexisting with liquidus or near-liquidus plagioclases in each experiment were determined via an FTIR spectroscopic technique. This study has shown that melt Ca# and Al#, H₂O content and crystallization pressure all control the composition of liquidus plagioclase. Increasing melt Ca# and Al# increase An content of plagioclase, whereas the effect of increasing pressure is the opposite. However, the importance of the role played by each of these factors during crystallization of natural magmas varies. Melt Ca# has the strongest control on plagioclase An content, but melt Al# also exerts a significant control. H₂O content can notably increase the An content of plagioclase, up to 10 mol % for H₂O-undersaturated melts, and 20 mol % for H₂O-saturated melts. Exceptionally calcic plagioclases (up to An₁₀₀) in some primitive subduction-related boninitic and related rocks cannot be attributed to the presence of the demonstrated amounts of H₂O (up to 3 wt %). Rather, they must be due to the involvement of extremely refractory (CaO/Na₂O>18) magmas in the petrogenesis of these rocks. Despite the refractory nature of some primitive MORB glasses, none are in equilibrium with the most calcic plagioclase (An₉₄) found in MORB. These plagioclases were likely produced from more refractory melts with CaO/Na₂O = 12–15, or from melts with exceptionally high Al₂O₃(>18%). Magmas of appropriate compositions to crystallize these most calcic plagioclases are sometimes found as melt inclusions in near liquidus phenocrysts from these rocks, but are not known among wholerock or glass compositions. The fact that such melts are not erupted as discrete magma batches indicates that they are effectively mixed and homogenized with volumetrically dominant, less refractory magmas. The high H₂O contents (∼ 6 wt%) in some high-Al basaltic arc magmas may be responsible for the existence of plagioclases up to An₉₅ in arc lavas. However, an alternative possibility is that petrogenesis involving melts with abnormally high CaO/Na₂O values (> 8) may account for the presence of highly anorthitic plagioclases in these rocks. Received: 31 August 1993 / Accepted: 20 May 1994     

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.     

Grain boundary diffusion of oxygen,potassium and calcium in natural and hot-pressed feldspar aggregates

 Grain boundary diffusion rates of oxygen, potassium and calcium in fine-grained feldspar aggregates were determined experimentally. The starting materials were a natural albite rock from the Tanco pegmatite and aggregates hot-pressed from fragments of Amelia albite or Ab, Or and An composition glasses. The technique employed isotopic tracers (¹⁸O, ⁴¹K, ⁴²Ca) either evaporated onto the surface or in an aqueous solution surrounding the sample, and depth profiling using an ion microprobe (SIMS). From the depth profiles, the product of the grain boundary diffusion coefficient (D′) and effective boundary width (δ) was calculated using numerical solutions to the appropriate diffusion equation. The experimental reproducibility of D′δ is a factor of 3. A separate determination of D′ independent of δ yields an effective grain boundary width of ∼3 nm, consistent with high resolution TEM observations of a physical grain boundary width <5 nm. Oxygen (as molecular water) grain boundary diffusion rates were determined in the Ab and Or aggregates at 450°–800° C and 100 MPa (hydrothermal), potassium rates in Or aggregates at 450°–700° C both at 0.1 MPa (in air) and at 100 MPa (hydrothermal), and calcium rates in An aggregates at 700°–1100° C and 0.1 MPa (in air). Oxygen grain boundary diffusion rates are similar in all three of the Ab aggregates and in the Or aggregate. Potassium and oxygen depth profiles measured in the same samples yield different D′δ values, confirming a diffusional transport mechanism. Potassium diffusion in the Or aggregate has a greater activation energy (216 vs 78 kJ/mol) than oxygen, and the Arrhenius relations cross at ∼625° C. Potassium D′δ values in Or aggregates are about a factor of five greater in hydrothermal experiments at 100 MPa than in experiments at 0.1 MPa in air. Calcium grain boundary diffusion rates in An aggregates are 4 to 5 orders of magnitude slower than potassium in Or and have a greater (291 kJ/mol) activation energy. This suggests that differences in formal charge and/or size of diffusing species may play an important role in their relative grain boundary diffusion rates. Received: 24 December 1993 / Accepted: 16 June 1994     

The effect of water on the viscosity of a haplogranitic melt under P-T-X conditions relevant to silicic volcanism

 The viscosities of hydrous haplogranitic melts synthesized by hydrothermal fusion at 2 kbar pressure and 800 to 1040° C have been measured at temperatures just above the glass transition and at a pressure of 1 bar using micropenetration techniques. The micropenetration viscometry has been performed in the viscosity range of 10⁹ Pa s to 10¹² Pa s. The samples ranged in water content from 0.4 to 3.5 wt%. For samples with up to 2.5 wt% H₂O, the water contents have been determined using infrared spectroscopy obtained before and after each viscometry experiment to be constant over the duration of the measurements. Above this water content a measurable loss of water occurs during the viscometry. The viscosity data illustrate an extremely nonlinear decrease in viscosity with added water. The viscosity drops drastically with the addition of 0.5 wt% of water and then shallows out at water contents of 2 wt%. An additional viscosity datum point obtained from the analysis of fluid inclusions in a water-saturated HPG8 confirms a near invariance of the viscosity with the addition of water between 2 and 6 wt%. These measurements may be compared directly with the data of Hess et al. (1995, in press) for the effects of excess alkali and alkaline earth oxides on the viscosity of HPG8 (also obtained at 1 bar). The viscosity of the melts, compared on an equivalent molar basis, increases in the order H₂O<(Li₂O<Na₂O< K₂O<Rb₂O,Cs₂O<BaO<SrO<CaO<MgO< BeO). The extraordinary decrease in melt viscosity with added water is poorly reproduced by the calculation scheme of Shaw (1972) for the range of water contents investigated here. The speciation of water in the quenched glasses can be used to quantify the dependence of the viscosity on hydroxyl content. Considering only the hydroxyl groups as active fluidizers in the hydrous melts the nonlinearity of the viscosity decrease and the difference with the effects of the alkali oxides becomes larger. Consequences for degassing calcalkaline rhyolite are discussed. Received: 17 August 1995/Accepted: 8 January 1996     

Significance of rare hydrous alkaline melts in Hawaiian xenoliths

 Two types of melt pockets occur in Hawaiian mantle xenoliths: amphibole-bearing (AMP) and spinel-bearing (SMP). AMPs contain amphibole (kaersutite), olivine (Fo₉₂), clinopyroxene (with 7–11% Al₂O₃), vesicles and glass. SMPs contain olivine, clinopyroxene, spinel, glass, and vesicles. The glasses in SMPs (SiO₂=44–45%, 11–12% alkalis, La=90–110 ppm) and AMPs (SiO₂=49–54%, 6–8.5% alkalis, La=8–14 ppm) are distinct in color and composition. Both glasses are generally characterized by LREE-enriched (chondrite-normalized) patterns. Amphibole and clinopyroxene have gently convex upward-to-moderately LREE-enriched patterns. Mineral/glass trace element abundance ratio plots show a strong negative Ti anomaly and a gentle negative Zr anomaly for clinopyroxene/glass; whereas amphibole/glass patterns show a distinctive positive Ti spike. The amphibole/glass trace element ratios are similar to published megacryst/lava values. An earlier study showed that the Hawaiian spinel lherzolites (lithosphere) have largely been metasomatized during post-erosional Honolulu magmatic activity. REE abundances of SMP glasses (melts) overlap the REE abundances calculated for such metasomes. The occurrence of hydrous, alkaline, mafic melt pockets in Hawaiian upper mantle xenoliths implies that (1) such hydrous liquids are generated in the upper mantle, and (2) water plays a role in magmatic activity associated with the Hawaiian plume. Although we are uncertain about the source (plume, lithosphere, or asthenosphere) of this water, we speculate that such melts and other alkalic lavas erupted on Oahu and on the sea-floor over the Hawaiian arch were generated from a broad „wet“ rim of a radially layered Hawaiian plume, whose hot and „dry“ core supplied the shield-forming magmas. Received: 6 February 1995 / Accepted: 28 August 1995     

The hydrous phase equilibria (to 3 kbar) of an andesite and basaltic andesite from western Mexico: constraints on water content and conditions of phenocryst growth

We have conducted high pressure (to 3 kbar), water saturated melting experiments on an andesite (62 wt% SiO₂) and a basaltic andesite (55 wt% SiO₂) from western Mexico. A close comparison between the experimental phase assemblages and their compositions, and the phenocryst assemblages of the lavas, is found in water saturated liquids, suggesting that the CO₂ content was minimal in the fluid phase. Thus the historic lavas from Volcan Colima (with phenocrysts of orthopyroxene, augite, plagioclase, and hornblende) were stored at a temperature between 950–975 °C, at a pressure between 700–1500 bars, and with a water content of 3.0–5.0 wt%. A hornblende andesite (spessartite) from Mascota, of nearly identical composition but with only amphibole phenocrysts, had a similar temperature but equilibrated at a minimum of 2000 bars pressure with a dissolved water content of at least 5.5 wt% in the liquid. Experiments on the basaltic andesite show that the most common natural phenocryst assemblages (olivine, ±augite, ±plagioclase) could have precipitated at temperatures from 1000–1150 °C, in liquids with a wide range of dissolved water content (∼2.0–6.0 wt%) and a corresponding pressure range. A lava of the same bulk composition with phenocrysts of hornblende, olivine, plagioclase, and augite is restricted to temperatures below 1000 °C and pressures below 2500 bars, corresponding to <5.5 wt% water in the residual liquid. Although there is some evidence for mixing in the andesites (sporadic olivine phenocrysts), the broad theme of the history of both lava types is that the phenocryst assemblages for both the andesitic magmas and basaltic andesitic magmas are generated from degassing and reequilibration on ascent of initially hydrous parents containing greater than 6 wt% water. Indeed andesitic magmas could be related to a basaltic andesite parent by hornblende-plagioclase fractionation under the same hydrous conditions. Received: 10 December 1996 / Accepted: 21 August 1997     

The significance of phenocryst diversity in tephra from recent eruptions at Popocatepetl volcano (central Mexico)

Tephra lapilli from six explosive eruptions between April 1996 and February 1998 at Popocatepetl volcano (=Popo) in central Mexico have been studied to investigate the causes of magma diversification in thick-crusted volcanic arcs. The tephra particles are sparsely porphyritic (≈5 vol%) magnesian andesites (SiO₂=58–65 wt%; MgO=2.6–5.9 wt%) that contain phenocrysts of NiO-rich (up to 0.67 wt% NiO) magnesian olivine (Fo_89–91 cores) with inclusions of Cr-spinel (cr#=59–70), orthopyroxene (mg#=63–76), clinopyroxene (mg#=68–86), intermediate to sodic plagioclase (An_33–66), and traces of amphibole. Major and trace element systematics indicate magma mixing. The liquid mg#_melt ratios inferred from the ferromagnesian phenocrysts suggest the existence of a mafic (mg#_melt ≈ 72–76) and an evolved component magma (mg#_melt ≈ 35–40). These component magmas form a hybrid magnesian andesite with an intermediate range of mg#_melt=50–72. The mafic end member (mg#_melt ≈ 72–75) is saturated with olivine and spinel and crystallizes at temperatures ≈1170–1085 °C with oxygen fugacities close to the fayalite–magnetite–quartz buffer and elevated water contents of several wt% H₂O. A likely location of crystallization is at lower crustal levels, possibly at the Moho. Olivine is followed by high-mg# clinopyroxene which could start to crystallize during magma ascent. At depths of ≈4 to 13 km, the mafic magma mixes with an evolved composition containing low-mg# clino- and orthopyroxene and plagioclase at a temperature of ≈950 °C. The repetitive ascent of batches of mafic magmas spaced days to weeks apart implies multiple episodes of crystallization and magma mixing. The tephra is similar to the Popo magnesian andesites, suggesting similar generic processes for the common lavas of the volcano. The advantage of the tephra is that it can be used to reconstruct the composition of the mafic magma. Building on the elemental systematics of the tephra and a comparison to the near-primary basalts from the surrounding monogenetic fields, we infer that the Popo mafic end member is a magnesian andesite with variable, but high SiO₂ contents of ≈55–62 wt% and near-primary characteristics, such as high-mg#_melt of 72–75, FeO*/MgO ratios <1 (if extrapolated to an mg#_melt of 72–75), and high Ni contents (=200 ppm Ni). This model implies that the typical elemental signature of the Popo andesites, such as the low CaO, Al₂O₃, FeO*, high Na₂O contents, and the depletion in high-field strength elements (e.g., P, Zr, Ti), are mantle source phenomena. Thus, determining the elemental budget of the magnesian andesite, as it is prior to the modifications by crustal differentiation, is central to quantifying the subcrustal mass fluxes beneath Popo. Received: 13 December 1999 / Accepted: 11 August 2000     

The mechanisms of water diffusion in polymerized silicate melts

Diffusion of water was experimentally investigated for melts of albitic (Ab) and quartz-orthoclasic (Qz₂₉Or₇₁, in wt %) compositions with water contents in the range of 0 to 8.5 wt % at temperatures of 1100 to 1200 °C and at pressures of 1.0 and 5.0 kbar. Apparent chemical diffusion coefficients of water (D _water) were determined from concentration-distance profiles measured by FTIR microspectroscopy. Under the same P-T condition and water content the diffusivity of water in albitic, quartz-orthoclasic and haplogranitic (Qz₂₈Ab₃₈ Or₃₄, Nowak and Behrens, this issue) melts is identical within experimental error. Comparison to data published in literature indicates that anhydrous composition only has little influence on the mobility of water in polymerized melts but that the degree of polymerization has a large effect. For instance, D_water is almost identical for haplogranitic and rhyolitic melts with 0.5–3.5 wt % water at 850 °C but it is two orders of magnitude higher in basaltic than in haplogranitic melts with 0.2–0.5 wt % water at 1300 °C. Based on the new water diffusivity data, recently published in situ near-infrared spectroscopic data (Nowak 1995; Nowak and Behrens 1995), and viscosity data (Schulze et al. 1996) for hydrous haplogranitic melts current models for water diffusion in silicate melts are critically reviewed. The NIR spectroscopy has indicated isolated OH groups, pairs of OH groups and H₂O molecules as hydrous species in polymerized silicate melts. A significant contribution of isolated OH groups to the transport of water is excluded for water contents above 10 ppm by comparison of viscosity and water diffusion data and by inspection of concentration profiles from trace water diffusion. Spectroscopic measurements have indicated that the interconversion of H₂O molecules and OH pairs is relatively fast in silicate glasses and melts even at low temperature and it is inferred that this reaction is an active step for migration of water. However, direct jumps of H₂O molecules from one cavity within the silicate network to another one can not be excluded. Thus, we favour a model in which water migrates by the interconversion reaction and, possibly, small sequences of direct jumps of H₂O molecules. In this model, immobilization of water results from dissociation of the OH pairs. Assuming that the frequency of the interconversion reaction is faster than that of diffusive jumps, OH pairs and water molecules can be treated as a single diffusing species having an effective diffusion coefficient . The shape of curves of D_water versus water content implies that increases with water content. The change from linear to exponential dependence of D_water between 2 and 3 wt % water is attributed to the influence of the dissociation reaction at low water content and to the modification of the melt structure by incorporation of OH groups. Received: 26 March 1996 / Accepted: 23 August 1996     

The temperature dependence of the speciation of water in NaAlSi3O8-KAlSi3O8 melts: an application of fictive temperatures derived from synthetic fluid-inclusions

 The speciation of water dissolved in glasses along the join NaAlSi₃O₈-KAlSi₃O₈ has been investigated using infrared spectroscopy. Hydrous melts have been hydrothermally synthesized by chemical equilibration of cylinders of bubble-free anhydrous start glasses with water at 1040^° C and 2 kbar. These melts have been isobarically and rapidly (200^° C/s) “drop”-quenched to room temperature and then subsequently depressurized. The speciation of water in the quenched glasses reflects the state of water speciation at a temperature (the so-called fictive temperature) where the quenched-in structure of the glasses closely corresponds to the melt structure at equilibrium. This fictive temperature is detectable as the macroscopically measureable glass transition temperature of these melt compositions. A separate set of experiments using vesicular samples of the same chemistry has precisely defined the glass transition temperature of these melts (±5^° C) on the basis of homogenization temperatures for water-filled fluid inclusions (Romano et al. 1994). The spectroscopic data on the speciation of water in these quenched glasses has been quantified using experimentally determined absorptivities for OH and H₂O for each individual melt composition. The knowledge of glass transition temperatures, together with quantitative speciation data permits an analysis of the temperature dependence of the water speciation over the 113^° C range of fictive temperatures obtained for these water-saturated melts. The variation of water speciation, cast as the equilibrium constant K where K = [H₂O] [O _m ]/[OH]² is plotted versus the fictive temperature of the melt to obtain the temperature dependence of speciation. Such a plot describes a single linear trend of the logarithm of the equilibrium constant versus reciprocal temperature, implying that the exchange of K for Na has little influence on melt speciation of water. The enthalpy derived from temperature dependence is 36.5(±5) kJ/mol. The results indicate a large variation in speciation with temperature and an insensitivity of the speciation to the K–Na exchange. Received: 8 March 1995/Accepted: 6 June 1995