Bubble growth in slightly supersaturated albite melt at constant pressure

Bubble growth experiments were performed in a piston-cylinder by hydrating albite melt with ∼11 wt.% H₂O at 550 MPa followed by rapid decompression at 1 MPa s⁻¹ to pressures of 450 or 400 MPa. At these conditions the melt was supersaturated with ∼0.5 or ∼1.5 wt.% H₂O, respectively, which caused rapid exsolution and bubble growth. Results at 1200 °C demonstrate that portions of the initial cumulative bubble-area distributions may be characterized by a power law with an exponent near 1, but they rapidly evolve to exponential distributions and approach a unimodal distribution after 32 h of growth. This evolution occurs by the growth of larger bubbles at the expense of smaller ones. The growth rate of the average bubble radius in these experiments is described by a power law whose exponent is 0.35, close to the theoretical exponent of 1/3 for phase growth in which coalescence is dominated by Ostwald ripening of the bubbles. Over the range of pressures and water contents investigated at 1200 °C, the bubble-size distributions and growth rate are not significantly affected by changes in the amount of exsolved water or by splitting the decompression path into two steps. Similar decompression experiments at 800 °C are dominated by smaller bubbles than in the 1200 °C experiments and also demonstrate exponential cumulative size distributions, but consistently contain a small fraction of larger bubbles. The growth rate of these bubble radii cannot be fit with a power law, but a logarithmic dependence of the bubble radii on time is possible, suggesting a difference in the growth mechanisms at low and high temperatures. This difference is attributed to the orders of magnitude changes in melt viscosity and water diffusion in the melt as the temperature varies from 800 to 1200 °C. At 1200 °C the transport properties of albite melt resemble those of natural basaltic melts whereas at 800 °C the properties are similar to those of andesitic to dacitic melts. The decompression rate used in this study exceeds natural rates by one to two orders of magnitude. Thus, these results indicate that natural mafic-to-intermediate magmas supersaturated with only a small excess of water should easily nucleate bubbles during ascent and that bubble growth in mafic magmas will proceed much more rapidly than in andesitic to dacitic magmas. Intermediate composition magmas also may be capable of forming bimodal bubble-size distributions even in the case when only one nucleation event occurred. The rapid evolution of the bubble-size distribution from a power law to an exponential may be useful in constraining the time duration between bubble nucleation and the quenching of natural samples.     

Ascent-driven crystallisation of dacite magmas at Mount St Helens, 1980–1986

We introduce a novel scheme that enables natural silicic glasses to be projected into the synthetic system Qz-Ab-Or-H₂O in order to relate variations in volcanic glass chemistry to changing pressure (P) and temperature (T) conditions in the sub-volcanic magma system. By this means an important distinction can be made between ascent-driven and cooling-driven crystallisation under water-saturated or undersaturated conditions. In samples containing feldspar and a silica phase (quartz or tridymite), quantitative P-T estimates of the conditions of last equilibrium between crystals and melt can be made. Formation of highly silicic melts (i.e. >77 wt% SiO₂) is a simple consequence of the contraction of the silica phase volume with decreasing pressure, such that high silica glasses can only form by crystallisation at low pressure. Resorption of quartz crystals appears to be a further diagnostic feature of decompression crystallisation. Groundmass and inclusion glasses in dacites from the 1980-1986 eruption of Mount St Helens volcano (WA) span a wide range in SiO₂ (68-80 wt%, anhydrous). The compositions of the least evolved (SiO₂-poor) inclusions in amphibole phenocrysts record entrapment of silicic liquids with Е.4 wt% water, corresponding to a water saturation pressure of ~200 MPa at 900 °C. The compositions of more evolved (higher SiO₂) plagioclase-hosted inclusions and groundmass glasses are consistent with extensive ascent-driven fractional crystallisation of plagioclase, oxide and orthopyroxene phenocrysts and microlites to low pressures. During this polybaric crystallisation, plagioclase phenocrysts trapped melts with a wide range of dissolved water contents (3.5-5.7 wt%). Magmas erupted during the Plinian phase of the 18 May 1980 eruption were derived from a large reservoir at depths of ̈́ km. Subsequent magmas ascended to varying depths within the sub-volcanic system prior to extraction. From glass chemistry and groundmass texture two arrest levels have been identified, at depths of 0.5-1 and 2-4 km. A single dome sample from February 1983 contains groundmass plagioclase, tridymite and quartz, testifying to temperatures of at least 885 °C at 11 MPa. These shallow storage conditions are comparable to those in the cryptodome formed during spring 1980. The corresponding thermal gradient, А.2 °C MPa^-1, is consistent with near-adiabatic magma ascent from ~8 km. We argue that the crystallisation history of Mount St Helens dacite magma was largely a consequence of decompression crystallisation of hot magma beyond the point of water saturation. This challenges the conventional view that phenocryst crystallisation occurred by cooling in a large magma chamber prior to the 1980-1986 eruption. Because the crystallisation process is both polybaric and fractional, it cannot be simulated directly using isobaric equilibrium crystallisation experiments. However, calculation of the phase proportions in water-saturated 910ᆣ °C experiments by Rutherford et al. (1985) over the pressure range 220-125 MPa reproduces the crystallisation sequence and phenocryst modes of Mount St Helens dacites from 18 May 1980. By allowing for the effects of fractional versus equilibrium crystallisation, entrained residual source material, and small temperature differences between nature and experiment, phase compositions can also be matched to the natural samples. We conclude that decompression of water-saturated magma may be the dominant driving force for crystallisation at many other silicic volcanic centres.     

Segregating gas from melt: an experimental study of the Ostwald ripening of vapor bubbles in magmas

Diffusive coarsening (Ostwald ripening) of H₂O and H₂O-CO₂ bubbles in rhyolite and basaltic andesite melts was studied with elevated temperature–pressure experiments to investigate the rates and time spans over which vapor bubbles may enlarge and attain sufficient buoyancy to segregate in magmatic systems. Bubble growth and segregation are also considered in terms of classical steady-state and transient (non-steady-state) ripening theory. Experimental results are consistent with diffusive coarsening as the dominant mechanism of bubble growth. Ripening is faster in experiments saturated with pure H₂O than in those with a CO₂-rich mixed vapor probably due to faster diffusion of H₂O than CO₂ through the melt. None of the experimental series followed the time^1/3 increase in mean bubble radius and time⁻¹ decrease in bubble number density predicted by classical steady-state ripening theory. Instead, products are interpreted as resulting from transient regime ripening. Application of transient regime theory suggests that bubbly magmas may require from days to 100 years to reach steady-state ripening conditions. Experimental results, as well as theory for steady-state ripening of bubbles that are immobile or undergoing buoyant ascent, indicate that diffusive coarsening efficiently eliminates micron-sized bubbles and would produce mm-sized bubbles in 10²–10⁴ years in crustal magma bodies. Once bubbles attain mm-sizes, their calculated ascent rates are sufficient that they could transit multiple kilometers over hundreds to thousands of years through mafic and silicic melt, respectively. These results show that diffusive coarsening can facilitate transfer of volatiles through, and from, magmatic systems by creating bubbles sufficiently large for rapid ascent.     

In situ observations of bubble growth in basaltic,andesitic and rhyodacitic melts

Bubble growth strongly affects the physical properties of degassing magmas and their eruption dynamics. Natural samples and products from quench experiments provide only a snapshot of the final state of volatile exsolution, leaving the processes occurring during its early stages unconstrained. In order to fill this gap, we present in situ high-temperature observations of bubble growth in magmas of different compositions (basalt, andesite and rhyodacite) at 1,100 to 1,240 °C and 0.1 MPa (1 bar), obtained using a moissanite cell apparatus. The data show that nucleation occurs at very small degrees of supersaturaturation (<60 MPa in basalt and andesite, 200 MPa in rhyodacite), probably due to heterogeneous nucleation of bubbles occurring simultaneously with the nucleation of crystals. During the early stages of exsolution, melt degassing is the driving mechanism of bubble growth, with coalescence becoming increasingly important as exsolution progresses. Ostwald ripening occurs only at the end of the process and only in basaltic melt. The average bubble growth rate (G _R) ranges from 3.4 × 10^?6 to 5.2 × 10^?7 mm/s, with basalt and andesite showing faster growth rates than rhyodacite. The bubble number density (N _B) at nucleation ranges from 7.9 × 10⁴ mm^?3 to 1.8 × 10⁵ mm^?3 and decreases exponentially over time. While the rhyodacite melt maintained a well-sorted bubble size distribution (BSD) through time, the BSDs of basalt and andesite are much more inhomogeneous. Our experimental observations demonstrate that bubble growth cannot be ascribed to a single mechanism but is rather a combination of many processes, which depend on the physical properties of the melt. Depending on coalescence rate, annealing of bubbles following a single nucleation event can produce complex bubble size distributions. In natural samples, such BSDs may be misinterpreted as resulting from several separate nucleation events. Incipient crystallization upon cooling of a magma may allow bubble nucleation already at very small degrees of supersaturation and could therefore be an important trigger for volatile release and explosive eruptions.     

Decompression and H<Subscript>2</Subscript>O exsolution driven crystallization and fractionation: development of a new model for low-pressure fractional crystallization in calc-alkaline magmatic systems

Magma ascent, decompression-induced H₂O exsolution and crystallization is now recognized as an important process in hydrous subduction zone magmas. During the course of such a process calculations suggest that the ascent rate of a degassing and crystallizing mafic magma will be greater than crystal settling velocities. Thus, any crystals formed as a consequence of volatile exsolution will remain suspended in the magma. If the magma erupts before the percentage of suspended crystals reaches the critical crystallinity value for mafic magma (~55 vol.%) it will produce the commonly observed crystal rich island arc basalt lava. If the magma reaches its critical crystallinity before it erupts then it will stall within the crust. Extension of compaction experiments on a 55 vol.% sand-Karo syrup suspension at different temperatures (and liquid viscosities) to the likely viscosities of interstitial andesitic to dacitic liquid within such a stalled magma suggest that small amounts (up to ~10%) can be expelled on a time scale of 1–10 years. The expelled liquid can create a new intermediate to silicic body of magma that is related to the original mafic magma via fractional crystallization. The short time scale for liquid expulsion indicate that decompression-induced H₂O exsolution and crystallization can be an important mechanism for fractional crystallization. Based on this assumption a general model of decompression-induced crystallization and fractionation is proposed that explains many of the compositional, mineralogical and textural features of Aleutian (and other andesites).     

Magma storage prior to the 1912 eruption at Novarupta, Alaska

New analytical and experimental data constrain the storage and equilibration conditions of the magmas erupted in 1912 from Novarupta in the 20th century's largest volcanic event. Phase relations at H₂O+CO₂ fluid saturation were determined for an andesite (58.7 wt% SiO₂) and a dacite (67.7 wt%) from the compositional extremes of intermediate magmas erupted. The phase assemblages, matrix melt composition and modes of natural andesite were reproduced experimentally under H₂O-saturated conditions (i.e., P_H2O=P_TOT) in a negatively sloping region in T-P space from 930 °C/100 MPa to 960 °C/75 MPa with fO₂~NNO+1. The H₂O-saturated equilibration conditions of the dacite are constrained to a T-P region from 850 °C/50 MPa to 880 °C/25 MPa. If H₂O-saturated, these magmas equilibrated at (and above) the level where co-erupted rhyolite equilibrated (~100 MPa), suggesting that the andesite-dacite magma reservoir was displaced laterally rather than vertically from the rhyolite magma body. Natural mineral and melt compositions of intermediate magmas were also reproduced experimentally under saturation conditions with a mixed (H₂O + CO₂) fluid for the same range in P_H2O. Thus, a storage model in which vertically stratified mafic to silicic intermediate magmas underlay H₂O-saturated rhyolite is consistent with experimental findings only if the intermediates have X_H2O^fl=0.7 and 0.9 for the extreme compositions, respectively. Disequilibrium features in natural pumice and scoria include pristine minerals existing outside their stability fields, and compositional zoning of titanomagnetite in contact with ilmenite. Variable rates of chemical equilibration which would eliminate these features constrain the apparent thermal excursion and re-distribution of minerals to the time scale of days.     

Heterogeneous bubble nucleation on Fe-Ti oxide crystals in high-silica rhyolitic melts

The nucleation of H₂O bubbles in magmas has been proposed as a trigger for volcanic eruptions. To determine how bubbles nucleate heterogeneously in silicate melts, experiments were carried out in which high-silica rhyolitic melts were hydrated at 740-800°C and 50-175 MPa, decompressed by 20-70 MPa, and held at the lower pressures for ≥10 s before being quenched. The hydration conditions were subliquidus, and all samples contain blocky magnetite + needle-shaped hematite ± plagioclase. Magnetite is abundant at 800°C and high pressures, whereas hematite becomes more abundant at lower temperatures and pressures. Bubbles nucleated in a single event in all samples, with the number density (N_T) of bubbles varying between 2 × 10⁷ and 1 × 10⁹ cm⁻³. At low degrees of supersaturation, one to a few bubbles nucleate on faces of magnetite, but at medium to high degrees of supersaturation, multiple bubbles nucleate on single magnetite grains. On hematite, one to a few bubbles nucleated at the ends of the needle-shaped crystals at medium supersaturations, but formed along their entire lengths at high supersaturations. N_T increases as water diffusivity decreases, indicating that the number of bubbles nucleated is influenced by their growth, which depletes the melt with respect to H₂O and lowers supersaturation. If volcanic eruptions are triggered by bubble formation in magmas stored in shallow-level magma chambers, then the supersaturations needed for heterogeneous nucleation suggest that only small amounts of crystallization are needed, whereas homogeneous nucleation is unlikely to trigger eruptions.     

An Experimental Study of Water and Carbon Dioxide Solubilities in Mid-Ocean Ridge Basaltic Liquids. Part II: Applications to Degassing

Degassing processes in basaltic magmas rich in both water andcarbon dioxide can be modeled using the solubilities of theendmember systems and the assumption of Henry's law. Suitesof vapor-saturated basaltic melts having a range of initialCO₂/H₂O ratios and erupted over a narrow depth interval willdefine negatively sloped arrays on an H₂O vs CO₂ plot. It isimportant that all of the major volatile species be consideredsimultaneously when interpreting trends in dissolved volatilespecies concentrations in magmas. Based on measured concentrations of water and carbon dioxidein basaltic glasses, the composition of the vapor phase at 1200°Cthat could coexist with a basaltic melt and the pressure atwhich it would be vapor saturated can be calculated. The rangein vapor compositions in equilibrium with submarine basaltsreflects the range in water contents in the melts characteristicof each environment. The ranges in the molar proportion of CO₂in vapor phases (X_CO2) calculated to be in equilibrium withsubmarine tholeiitic glasses are 0•93–1•00 formid-ocean ridge basalts (MORB), 0•60–0•99 forglasses from Kilauea [representative of ocean island basalts(OIB)] and 0–0•94 for glasses from back-arc basins(BABB). MORB glasses from spreading centers ranging from slow(e.g. the Mid-Atlantic Ridge) to fast (e.g. East Pacific Rise,9–13°N) are commonly supersaturated with respect toCO₂-rich vapor, resulting from magma ascent rates so rapid thatmagmas erupt on the sea-floor without having been fully degassedby bubble nucleation and growth during ascent. In contrast tothe MORB glasses, volatile contents in submarine glasses fromKilauea are consistent with having been in equilibrium witha vapor phase containing 60–100 mol% CO₂ at the pressureof eruption, reflecting differences in average magma transportrates during eruptions at mid-ocean ridges and hotspot volcanoes. Degassing during decompression of tholeiitic basaltic magmais characterized by strong partitioning of CO₂ into the vaporphase. During open system degassing, CO₂ is rapidly removedfrom the melt with negligible loss of water, until a pressureis reached at which the melt is in equilibrium with nearly purewater vapor. From this pressure downward, the water contentof the melt follows the water solubility curve. During closedsystem degassing, water and CO₂ contents in vapor-saturatedbasaltic magmas will depend strongly on the vapor compositionas determined by the initial volatile concentrations. Deviationfrom open system behavior, toward lower dissolved H₂O and CO₂saturation concentrations at a given pressure, will be greatestin melts having high total volatile concentrations and highCO₂:H₂O ratios. Closed system degassing of basaltic melts havingthe low initial H₂O and CO₂ contents typical of MORB and OIB,however, are similar to the open system case. KEY WORDS: mid-ocean ridge basalts; water and carbon dioxide solubility; degassing     

Crystal nucleation and growth produced by continuous decompression of Pinatubo magma

High-temperature decompression experiments demonstrate that crystal textures preserve a record of the style and rate of magmatic ascent. To reinforce this link, we performed a suite of isothermal decompression experiments using starting material from the climactic 1991 Pinatubo eruption. We decompressed experiments from 220 MPa to final, quench pressures of 75 or 30 MPa using continuous decompression rates of 100, 30, 10, 3, 1, and 0.3 MPa h^?1. Amphibole, clinopyroxene, and plagioclase crystallized during the experiments, with plagioclase microlites dominating the assemblage. Total microlite number densities range from 10^7.6±0.4 up to 10^8.2±0.2 cm^?3, with plagioclase accounting for up to 65% of the total number. Plagioclase microlite area increased systematically from 19?±?8 to 937?±?487 µm² with increasing experiment duration. Our textures provide time-integrated records of crystal kinetics. Average nucleation and areal growth rates of plagioclase are highest in the fastest decompressions (~?10^7.5 cm^?3 h^?1 and 10.1?±?4.1 µm² h^?1, respectively) and more than an order of magnitude lower in the slowest experiments (~?10^5.5 cm^?3 h^?1 and 0.8?±?0.2 µm² h^?1, respectively). Both nucleation and growth rates are highest at high degrees of disequilibrium. We find that peak supersaturation-dependent instantaneous rates are generally more than an order of magnitude faster than average rates. We use those instantaneous nucleation and growth rates to introduce an iterative model to evaluate the effects of different decompression rates, decompression paths (continuous, single-step or multistep), and the presence of phenocrysts on final crystallinity and microlite size distribution.     

Shallow-level decompression crystallisation and deep magma supply at Shiveluch Volcano

Recent petrological studies indicate that some crustal magma chambers may be built up slowly by the intermittent ascent and amalgamation of small packets of magma generated in a deep-seated source region. Despite having little effect on whole-rock compositions, this process should be detectable as variable melt trace element composition, preserved as melt inclusions trapped in phenocrysts. We studied trace element and H₂O contents of plagioclase- and hornblende-hosted melt inclusions from andesite lavas and pumices of Shiveluch Volcano, Kamchatka. Melt inclusions are significantly more evolved than the whole rocks, indicating that the whole rocks contain a significant proportion of recycled foreign material. H₂O concentrations indicate trapping at a wide range of pressures, consistent with shallow decompression-driven crystallisation. The variation of trace element concentrations indicates up to ∼30% decompression crystallisation, which accounts for crystallisation of the groundmass and rims on phenocrysts. Trace element scatter could be explained by episodic stalling during shallow magma ascent, allowing incompatible element concentrations to increase during isobaric crystallisation. Enrichment of Li at intermediate pH₂O reflects influx and condensation of metal-rich vapours. A set of “exotic melts”, identified by their anomalous incompatible trace element characteristics, indicate variable source chemistry. This is consistent with evolution of individual magma batches with small differences in trace element chemistry, and intermittent ascent of magma pulses. Electronic supplementary material The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Generation of CO2-rich melts during basalt magma ascent and degassing

To test mechanisms of basaltic magma degassing, continuous decompressions of volatile-bearing (2.7–3.8 wt% H₂O, 600–1,300 ppm CO₂) Stromboli melts were performed from 250–200 to 50–25 MPa at 1,180–1,140 °C. Ascent rates were varied from 0.25 to ~1.5 m/s. Glasses after decompression show a wide range of textures, from totally bubble-free to bubble-rich, the latter with bubble number densities from 10⁴ to 10⁶ cm^?3, similar to Stromboli pumices. Vesicularities range from 0 to ~20 vol%. Final melt H₂O concentrations are homogeneous and always close to solubilities. In contrast, the rate of vesiculation controls the final melt CO₂ concentration. High vesicularity charges have glass CO₂ concentrations that follow theoretical equilibrium degassing paths, whereas glasses from low vesicularity charges show marked deviations from equilibrium, with CO₂ concentrations up to one order of magnitude higher than solubilities. FTIR profiles and maps reveal glass CO₂ concentration gradients near the gas–melt interface. Our results stress the importance of bubble nucleation and growth, and of volatile diffusivities, for basaltic melt degassing. Two characteristic distances, the gas interface distance (distance either between bubbles or to gas–melt interfaces) and the volatile diffusion distance, control the degassing process. Melts containing numerous and large bubbles have gas interface distances shorter than volatile diffusion distances, and degassing proceeds by equilibrium partitioning of CO₂ and H₂O between melt and gas bubbles. For melts where either bubble nucleation is inhibited or bubble growth is limited, gas interface distances are longer than volatile diffusion distances. Degassing proceeds by diffusive volatile transfer at the gas–melt interface and is kinetically limited by the diffusivities of volatiles in the melt. Our experiments show that CO₂-oversaturated melts can be generated as a result of magma decompression. They provide a new explanation for the occurrence of CO₂-rich natural basaltic glasses and open new perspectives for understanding explosive basaltic volcanism.     

Subsolidus and Partial Melting Reactions in the Quartz-excess CaO+MgO+Al2O3+SiO2+H2O System under Water-excess and Water-deficient Conditions to 10 kb: Some Implications for the Origin of Peraluminous Melts from Mafic Rocks

Experimental results up to 10 kb pressure are presented on thestability of amphibole in the quartz-excess CaO+MgO+Al₂O₃ (CMASH)system under H₂O)-excess and H₂O deficient conditions. Amphiboleis stable above the solidus under H₂O-excess conditions whereasunder H₂O-deficient conditions dehydration melting of amphibole-bearingassemblages defines the solidus. The successive appearance ofamphibole, talc, and zoisite with increasing pressure considerablymodifies the plagioclase-pyroxene-garnet-kyanite reactions documentedexperimentally in the CaO+MgO+Al₂O₃+SiO₂ system for gabbro-granulite-eclogitetransitions. Although both clino pyroxene and cordierite (withanorthite+orthopyroxene+quartz) may melt eutectically at oneatmosphere to form diopside-normative and corundum-normativemelts respectively, at higher pressures under H₂O-excess conditionsthe peritectic melting of mafic rock compositions produces corundum-normativeliquids together with either clinopyroxene or amphibole. Dehydrationmelting produces melts which are not corundum-normative. Thesedata are used to discuss the origins and evolution of contrastingbasalt-andesite-dacite-rhyolite volcanic suites and graniticplutons, many of whose silicic variants are corundum-normativein character, such as the Toba luff ignimbrites, Indonesia (Beddoc-Stephenset al., 1983) and I-type granite minimum melts (White &Chappell, 1977). In contrast, it is proposed that for the Cascadesbasalt-andesite-dacite-rhyolite suite the ortho pyroxene-plagioclase-quartzthermal divide was maintained up to rhyolite compositions, therebyprohibiting the derivation of corundum-normative rocks fromdiopside-normative parent magmas. The deduced reaction relations between pyroxenes, amphibole,plagioclase, quartz, and liquid are used to explain the absenceor extreme scarcity of hydrous phases in some hydrous magmas.These phase relations can also explain the development of laterplagioclase overgrowths on resorbed plagioclase cores in graniticintrusives, and the general absence of resorption and overgrowthsin chemically equivalent extrusive rocks. A theoretical analysis of the partial melting of forsterite-bearingassemblages in the CaO+MgO+Al₂O₃+SiO₂+H₂O system shows thatunder H₂O-excess conditions partial melting may generate corundum-normative(but low SiO₂) melts from a peridotite source at shallow depths.     

Silicic Magma Formation in Overthickened Crust: Melting of Charnockite and Leucogranite at 15, 20 and 25 kbar

Two models of silicic magma formation have been experimentallytested: (1) generation of A-type granite magma by partial meltingof crustal source rocks at depths >50 km; (2) productionof syenite magma by partial melting of quartzofeldspathic rocksat pressures >15 kbar. Melting experiments at 15, 20 and25 kbar were performed on Archaean biotite-bearing charnockiteof opx-bearing granodiorite composition, and on leucogranite.Most experiments were conducted with oxalic acid (OA) to provide2 wt % H₂O and     

Major Element Chemical Variation in the Eocene Lavas of the Isle of Skye, Scotland

The Lower Eocene lavas of northern Skye are preserved over anarea of approximately 1500 km² in a shallow faulted oval basin.Seventy-four new major element chemical analyses have been madeof specimens showing minimal post-consolidation alteration.These demonstrate that the early volcanics vary from hypersthene-tonepheline-normative basalts; the former containing less TiO₂and P₂O₅ but more K₂O than the latter. The compositions of thesebasalts straddle the low-pressure thermal divide near the criticalplane of silica under-saturation, the normative ‘join’OI–Pl–Cpx; implying that their variation was causedby high-pressure, upper mantle processes. The unusual incompatibleelement pattern of the lavas suggests that a K-rich mantle phase,tentatively taken to be phlogopite, was involved in their genesis.At subvolcanic pressures the spectrum of basaltic magmas wassplit by the join olivine-plagioclase-augite, producing twodivergent trends; from alkali basalts, via relatively Si- andK-poor and Fe- and Ti-rich hawaiites and mugearites to benmoreite,and from hypersthene-normative basalts, via relatively Si- andK-rich and Fe- and Ti-poor intermediates to trachyte. One-atmospheremelting experiments on 21 lavas are used to supplement the chemicalinvestigation of these low-pressure trends. Two flows of aphyric, low-alkali tholeiitic basalt, with compositionsquite distinct from all other Skye lavas, have been discoveredinterleaved among the alkali-rich rocks near the top of thevolcanic pile, at present exposure level. The almost patternlesschemical variation of all the lavas with time suggests thatmagma chambers, as conventionally envisaged, did not exist beneaththe Skye volcanic field, but rather that fissure eruptions werefed from a sponge-like plexus of conduits and small reservoirs,within which magma, affected to varying extents by upper mantleprocesses, remained for varying periods, pursuing low-pressurefractionation trends. The chemistry of the main Skye basaltsand the low-alkali tholeiites, plus available data on otherpost-lava Skye basic igneous rocks, such as the Cuillin LayeredIntrusion, other gabbros in the central complexes, and late-stagealkali dolerite dykes, are all combined in an attempt to ascribethe variation of the basic magmas which approached or reachedthe surface of Skye to the growth, culmination, and waning stagesof a thermal event in the upper mantle beneath that area.     

Depths of Partial Crystallization of H2O-bearing MORB: Phase Equilibria Simulations of Basalts at the MAR near Ascension Island (7 11{degrees}S)

Phase equilibria simulations were performed on naturally quenchedbasaltic glasses to determine crystallization conditions priorto eruption of magmas at the Mid-Atlantic Ridge (MAR) east ofAscension Island (7–11°S). The results indicate thatmid-ocean ridge basalt (MORB) magmas beneath different segmentsof the MAR have crystallized over a wide range of pressures(100–900 MPa). However, each segment seems to have a specificcrystallization history. Nearly isobaric crystallization conditions(100–300 MPa) were obtained for the geochemically enrichedMORB magmas of the central segments, whereas normal (N)-MORBmagmas of the bounding segments are characterized by polybariccrystallization conditions (200–900 MPa). In addition,our results demonstrate close to anhydrous crystallization conditionsof N-MORBs, whereas geochemically enriched MORBs were successfullymodeled in the presence of 0·4–1 wt% H₂O in theparental melts. These estimates are in agreement with direct(Fourier transform IR) measurements of H₂O abundances in basalticglasses and melt inclusions for selected samples. Water contentsdetermined in the parental melts are in the range 0·04–0·09and 0·30–0·55 wt% H₂O for depleted and enrichedMORBs, respectively. Our results are in general agreement (within±200 MPa) with previous approaches used to evaluate pressureestimates in MORB. However, the determination of pre-eruptiveconditions of MORBs, including temperature and water contentin addition to pressure, requires the improvement of magma crystallizationmodels to simulate liquid lines of descent in the presence ofsmall amounts of water. KEY WORDS: MORB; Mid-Atlantic Ridge; depth of crystallization; water abundances; phase equilibria calculations; cotectic crystallization; pressure estimates; polybaric fractionation     

Kinetics of heterogeneous bubble nucleation in rhyolitic melts: implications for the number density of bubbles in volcanic conduits and for pumice textures

We performed decompression experiments to simulate the ascent of a phenocryst-bearing rhyolitic magma in a volcanic conduit. The starting materials were bubble-free rhyolites water-saturated at 200 MPa–800°C under oxidizing conditions: they contained 6.0 wt% dissolved H₂O and a dense population of hematite crystals (8.7 ± 2 × 10⁵ mm⁻³). Pressure was decreased from the saturation value to a final value ranging from 99 to 20 MPa, at constant temperature (800°C); the rate of decompression was either 1,000 or 27.8 kPa/s. In all experiments, we observed a single event of heterogeneous bubble nucleation beginning at a pressure P _N equal to 63 ± 3 MPa in the 1,000 kPa/s series, and to 69 ± 1 MPa in the 27.8 kPa/s series. Below P _N, the degree of water supersaturation in the liquid rapidly decreased to a few 0.1 wt%, the nucleation rate dropped, and the bubble number density (BND) stabilized to a value strongly sensitive to decompression rate: 80 mm⁻³ at 27.8 kPa/s vs. 5,900 mm⁻³ at 1,000 kPa/s. This behaviour is like the behavior formerly described in the case of homogeneous bubble nucleation in the rhyolite-H₂O system and in numerical simulations of vesiculation in ascending magmas. Similar degrees of water supersaturation were measured at 27.8 and 1,000 kPa/s, implying that a faster decompression rate does not result in a larger departure from equilibrium. Our experimental results imply that BNDs in acid to intermediate magmas ascending in volcanic conduits will depend on both the decompression rate and the number density of phenocrysts, especially the number density of magnetite microphenocrysts (1–100 mm⁻³), which is the only mineral species able to reduce significantly the degree of water supersaturation required for bubble nucleation. Very low BNDs (≈1 mm⁻³) are predicted in the case of effusive eruptions ( ≈ 0.1 kPa/s). High BNDs (up to 10⁷ mm⁻³) and bimodal bubble size distributions are expected in the case of explosive eruptions: (1) a relatively small number density of bubbles (1–100 mm⁻³) will first nucleate in the lower part of the conduit ( ≈ 10 kPa/s), either at high pressure on magnetite or at lower pressure on quartz and feldspar (or by homogeneous nucleation in the liquid) and (2) then, extreme decompression rates near the fragmentation level ( ≈ 10³ kPa/s) will trigger a major nucleation event leading to the multitude of small bubbles, typically a few micrometers to a few tens of micrometers in diameter, which characterizes most silicic pumices.     

Evidence for distinct stages of magma history recorded by the compositions of accessory apatite and zircon

Accessory minerals contain a robust and accessible record of magma evolution. However, they may reflect relatively late-stage conditions in the history of the host magmas. In the normally zoned Criffell granitic pluton (Scotland), whole-rock (WR) compositions reflect open system assimilation and fractional crystallisation at depths of >11 km, whereas amphibole barometry and the absence of inherited zircon suggest that the observed mineral assemblages crystallised following emplacement of magmas with little or no crystal cargo at depths of 4–6 km. The crystallisation history is documented by large trace-element variations amongst apatite crystals from within individual samples: decreasing LREE and Th concentrations in apatite crystals from metaluminous samples reflect broadly synchronous crystallisation of allanite, whereas lower LREE and Th, and more negative Nd anomalies in apatites from peraluminous samples reflect the effects of monazite crystallisation. WR evolution is likely to have occurred within a deep crustal hot zone where H₂O-rich (~6 wt%), low-viscosity magmas segregated and ascended adiabatically in a super-liquidus state, leading to resorption of most entrained crystals. Stalling, emplacement and crystallisation resulted from intersection with the H₂O-saturated liquidus at ~4 km. H₂O contents are as important as temperature in the development of super-liquidus magmas during ascent, blurring distinctions between apparently ‘hot’ and ‘cold’ granites. The trace-element contents of most accessory minerals are controlled by competitive crystallisation of other accessory minerals in small melt batches, consistent with the incremental assembly of large granitic plutons.     

Experimental study of bubble coalescence in rhyolitic and phonolitic melts

We have experimentally studied the process of bubble coalescence in rhyolite and phonolite melts of natural composition. The experiments involved decompression of water-saturated melts equilibrated at pressures and temperatures from 100 to 150 MPa and 775 to 840 °C in vertically oriented, rapid-quench capable, cold seal pressure vessels. One type of experiments (rhyolite MCR-100, 120, 150 and phonolite LSP-120 series') approximates a “static” bubble coalescence case, where we held the decompressed samples for ∼5 seconds to 4320 minutes (3 days) before quenching. The second type (rhyolite LPC-100 series) replicates an “expanding” bubble coalescence environment, where we continually decompressed the experiments at a rate of 0.5 MPa/s, examining samples quenched at ending pressures between 10 and 80 MPa. Our “static” case (MCR-100, 120, and 150, and LSP-120) results show significant increases in the modal bubble sizes and in the sizes of the largest bubbles, corresponding to measurable broadening in the size distributions. Their bubble number densities (N_V) decrease as a function of hold time at their final pressures (P_F), and can be fit well by power law functions. Our “expanding” case experiments (LPC-100) show a significant drop in N_V during the duration of the experiments that can be fit by an exponential equation as N_V vs. either time or P_F. Average estimates of bulk coalescence rates indicate a ∼1 order of magnitude drop in N_V for “static” case rhyolites in a 2-3 day period, and ∼2 orders of magnitude for phonolites within a 3 day period. Despite a ∼2 order of magnitude difference in viscosity, coalescene in the phonolite is only slightly faster than the rhyolite. The “expanding” case experiments show a ∼1 order of magnitude drop in N_V over 180 seconds. Thus, N_V's decrease 4 orders of magnitude faster in expanding vs. static bubbly rhyolite melts. Our results imply that significant bubble coalescence can occur in rhyolite magmas at relatively fast (∼20 m/s) ascent rates in the conduit. Thus, bubble interconnectivity, leading to high permeability, is possible during ascent. Bubble coalescence may occur during second boiling in magma bodies that are stalled in the crust. The timescales over which this occurs is much faster than the estimated rise rates for bubbles to reach the top of the magma chamber.     

The andesite aqueduct: perspectives on the evolution of intermediate magmatism in west-central (105–99°W) Mexico

Intermediate calc-alkaline magma (52-65% SiO₂) in western-central Mexico is the focus of this paper, and the typically porphyritic andesites (57-65% SiO₂) form large central volcanoes, whereas basaltic andesites (52-57% SiO₂) are less porphyritic, and they are found as cones and flows but are absent from central volcanoes. Several studies of experimental phase equilibria on these lavas relate water concentration to the phenocryst assemblages and to the degree of crystallinity, so that the abundance, composition and variety of phenocrysts can be used to constrain the amount of water dissolved in the magmas. Thus, the plagioclase-rich andesites of Volcan Colima, Mexico, become so as a result of decompressional crystallisation at ~950 °C (the pyroxene phenocryst temperature), and lose their dissolved water (2.5 to 4.5 wt% H₂O) which is inversely proportional to the modal abundance of plagioclase. The feeding magma to V. Colima, North America's most productive central volcano, is represented by hornblende lamprophyre, a lava type without plagioclase phenocrysts which requires at least 6 wt% water to reproduce the phenocryst assemblage. Thus, degassing of the V. Colima magmas, and of those of the other central volcanoes in the western-central Mexican volcanic belt, contributes essentially all their dissolved water to the conduit or to the atmosphere. The source of this magmatic water is related to the source of the intermediate magmas. For some this must lie in the mantle, as the incorporation of hornblende-lherzolite nodules in a hydrous andesite with hornblende phenocrysts could only have occurred while ascending through the mantle. Consistent with a mantle source is the composition of the olivine phenocrysts in Mexican lavas with 10 to 5% MgO, which is in the mantle range of Fo_88-92. Accordingly, basaltic andesites and andesites with >5% MgO are candidates for a mantle source. The equilibration of intermediate magmas with the mantle, as illustrated by the experiments of various workers, requires that the magmas be hydrous at pressure. An additional constraint is that the activity of silica in the mantle must be equal to that in the hydrous magma at equilibrium. Using published and new experiments to define RTln%_SiO2 in hydrous liquids, this quantity is shown to vary as a function of liquid composition (H₂O, MgO, Na₂O+K₂O), and it approaches zero for quartz-saturated hydrous liquids. Using appropriate values of RTln%_SiO2 for three intermediate lavas, the amount of water required to equilibrate with an olivine-orthopyroxene mantle source is calculated, and within error indicates that only the most silica-rich magma is at water saturation in the mantle, in agreement with published experimental work. Hydrous intermediate magmas, ascending from their hornblende-lherzolite source regions (~1 to 1.5 GPa) along the hydrous adiabat, may not encounter any phase boundaries until 0.2-0.4 GPa because of the increase in the thermal stability of hornblende in water-undersaturated magmas. Therefore, the phenocryst assemblages of hornblende-free andesites equilibrate at low pressures. The virtual absence of basalt in west-central Mexico (<4 Ma) is considered to be related to the large increase in crystallinity found in isobaric hydrous experiments crystallising hornblende at pressures close to those at the base of the crust. As a large proportion of the ferromagnesian components of basalt is acceptable to hornblende, it does not take a significant cooling interval (~40-50 °C) below the liquidus for hydrous basaltic magma to acquire >50% crystallinity, evidently also an eruptible limit for V. Colima andesitic lavas. If the lower limit of water dissolved in Mexican intermediate magmas is accepted as that required for phenocryst equilibration (~6 wt% water), and the upper limit as saturation in the mantle source at 1 GPa (~16 wt%) then, with an estimate of the volcanic and plutonic magma delivery rate (km³/10⁶ year) per km of volcanic arc, the flux of water returned from the mantle along the 35,000-km, global subduction-related arc system can be estimated. Measurements of the volcanic flux are woefully few, and estimates from Mexico, the Lesser Antilles and central America show a range from 4 to 20 km³/10⁶ year2km which, if subtracted from the isotopically constrained continental growth rate, gives the plutonic flux rate. This suggests that, of the magma flux ascending to the continental crust, only about a fifth reaches the surface. If the dissolved magmatic water limits are coupled with the volcanic and plutonic emplacement rates, then the amount of water returned by magmatism to the crust is crudely in balance with that subducted.     

Experimental constraints on permeable gas transport in crystalline silicic magmas

The gas and fluid transport in magmas via permeable flow through interconnected bubble networks controls the rate of outgassing from magmas ascending in volcanic conduits and the fluid transport in the mushy boundary layer of magma reservoirs. Hence, clarifying its mechanism and rate is crucial to understanding the explosivity of volcanic eruptions and the evolution and dynamics of a magma reservoir. Recent experimental studies have determined the gas permeabilities in crystal-free rhyolite and basalt. However, no experimental study has investigated the effect of the crystal contents on the permeable gas transport in magmas. In this study, we performed decompression experiments for hydrous rhyolitic melts having crystallinities of 30 and 50 vol% to examine the effect of crystals on the bubble microstructure and gas permeability during magma vesiculation. Size-controlled (100-meshed) corundum crystals were used as an analog of the phenocrysts in silicic magmas. Microstructural analyses using X-ray CT showed that bubbles coalesce and their connectivity increases with a decrease in the final pressure after the decompression, that is, an increase in the vesicularity. As long as the vesicularities of melt part in the crystal-free basis (melt vesicularity) were similar, no clear effect of the crystallinity on the degree of bubble coalescence and connectivity was observed at melt vesicularities <68 vol%. The corundum showed a large contact angle with aqueous fluid as well as plagioclase and alkaline feldspar; this failed to induce the efficient heterogeneous nucleation and coalescence of bubbles on its surface. The gas permeabilities of all the run products were lower than the detection limits of the present analysis (the order of 10⁻¹⁶ m²) at melt vesicularities <68 vol%. These results show that silicic magmas containing 30 and 50 vol% phenocrysts with a large contact angle have low gas permeabilities until the vesicularity becomes large (at least >68 vol%). This result indicates that the permeable fluid transport through a deep volcanic conduit, which has been proposed on the basis of the observations of volcanic gases and natural products, is so slow that other processes, like shear deformation or magma convection, may be needed to explain the observations.