Magmatic Volatile Phase Research Articles

Effects of magmatic volatile influx in mafic VMS hydrothermal systems: Evidence from the Troodos ophiolite, Cyprus

The Troodos ophiolite, Cyprus is the principal on- land analogue for mafic-hosted volcanogenic massive sulfide (VMS) deposits. This study, for the first time, presents sulfur isotope (δ34S) data on a regional scale from VMS deposits and other mineralised zones across the Troodos ophiolite. In combination δ34S, Se/S ratios and trace element chemistry (e.g., Se, Cu and Au) of different hydrothermal sulfides are used to assess variations in magmatic volatile influx and the source of metals and sulfur in ancient hydrothermal systems.Sulfur isotope analyses (n=180) across 19 mineral deposits indicate a variable source of sulfur in the Troodos VMS hydrothermal system, this in turn allows a variable source of metals to be inferred. Pyrite δ34S range from −5.5‰ to +13.2‰ with an average of +4.6‰ (n=160) for all deposits investigated. These δ34S variations cannot only be explained by variable proportions of thermochemical seawater sulfate reduction (δ34S +18 to +19‰) and leaching of primary magmatic sulfur from igneous rocks (δ34S 0-1‰). Consequently, two processes are proposed, explaining the trace metal and δ34S variation across the Troodos ore-forming systems including, i) a variable source of metals in the sheeted dyke complex and ii) the addition of a magmatic volatile-rich phase to the VMS hydrothermal systems.Two distinct lava units exist in the Troodos stratigraphy, namely the upper and lower pillow lavas (UPL and LPL). The more primitive UPL are enriched in Au, Se and Cu relative to As, Sb and Zn that are concentrated in the LPL. Some VMS deposits pre-date the formation of the UPL (e.g., Agrokipia A) suggesting Se, Cu and Au depleted source rocks. Hence, the stratigraphic position of VMS deposits and the ratio of UPL:LPL affinity elements (e.g., As +Zn+Sb vs. Cu+Se+Au) imply a systematic relationship between trace element distribution and stratigraphic depth; this relates to the relative proportion of UPL and LPL affinity lavas within the metal source region.δ34S values <0‰ recorded in some VMS deposits that are less than the Troodos magmatic mean of 0- 1‰ may be related to anhydrite buffering during fluid ascent, microbial sulfate reduction or the direct contribution of magmatically derived sulfur, to the hydrothermal system from an underlying magma chamber via volatile exsolution. We propose that negative δ34S values combined with Se/S × 106 ratios >500 in pyrite suggest the contribution of a magmatic volatile component (e.g., Apliki and Skouriotissa). We demonstrate that the source of metals and sulfur in the Troodos VMS hydrothermal system is affected by regional scale processes related to i) variable source lithologies and, ii) the contribution of a magmatic volatile phase to some Troodos VMS hydrothermal systems.

Origin of Cu in the PACMANUS hydrothermal field from the eastern Manus back-arc basin: evidence from mass balance modeling

Hydrothermal precipitates associated with active vents in the eastern Manus Basin, an actively opening back-arc basin in the Bismarck Sea, Papua New Guinea, are among the most Cu-rich on the modern seafloor. The volcanic rocks associated with this mineralization may be insufficiently enriched in Cu to account for the Cu content of the sulfides by simple leaching. The PACMANUS hydrothermal field lies in the eastern portion of the eastern Manus Basin. Mass balance modeling of the PACMANUS hydrothermal system indicates that simple leaching of a stationary reaction zone (0.144 km3) by hydrothermal fluids cannot yield the Cu found in associated sulfide deposits because unacceptably high leaching, transportation and precipitation efficiencies are required to derive the Cu in sulfides by leaching processes. With 100% leaching, transport and precipitating efficiency, 0.166 km3 of volcanic rocks would need to be leached to account for the Cu budget of hydrothermal sulfide deposits. The key requirement for forming metal-rich magmatic fluids is a large amount of metals available to enter the exsolved vapor phase. Magmas generated in the eastern Manus Basin inherently have high fO2 because of metasomatism of the mantle source by oxidized materials from the subducted slab, leading to copper enrichment in the magma chamber. Moreover, the presence of Cu in gas-rich melt inclusi on bubbles in Pual Ridge andesite is evidence that degassing and partitioning of Cu into the magmatic volatile phase has occurred in the eastern Manus Basin. Numerical mass balance modeling indicates that approximately 0.236 Mt Cu was potentially transferred to the hydrothermal system per cubic kilometer magma. Magmatic degassing seems to play a more significant role than leaching.

Optimal depth of subvolcanic magma chamber growth controlled by volatiles and crust rheology

Storage pressures of magma chambers influence the style, frequency and magnitude of volcanic eruptions. Neutral buoyancy or rheological transitions are commonly assumed to control where magmas accumulate and form such chambers. However, the density of volatile-rich silicic magmas is typically lower than that of the surrounding crust, and the rheology of the crust alone does not define the depth of the brittle–ductile transition around a magma chamber. Yet, typical storage pressures inferred from geophysical inversions or petrological methods seem to cluster around 2 ± 0.5 kbar in all tectonic settings and crustal compositions. Here, we use thermomechanical modelling to show that storage pressure is controlled by volatile exsolution and crustal rheology. At pressures $$\lesssim$$ 1.5 kbar, and for geologically realistic water contents, chamber volumes and recharge rates, the presence of an exsolved magmatic volatile phase hinders chamber growth because eruptive volumes are typically larger than recharges feeding the system during periods of dormancy. At pressures $$>rsim$$ 2.5 kbar, the viscosity of the crust in long-lived magmatic provinces is sufficiently low to inhibit most eruptions. Sustainable eruptible magma reservoirs are able to develop only within a relatively narrow range of pressures around 2 ± 0.5 kbar, where the amount of exsolved volatiles fosters growth while the high viscosity of the crust promotes the necessary overpressurization for eruption. Volatile exsolution and crustal viscosity dictate that the optimum pressure for the growth of an eruptible magma reservoir is 2 kbar in all tectonic settings and crustal compositions, according to thermomechanical modelling.

An accurate model to predict sulfur concentration at anhydrite saturation in silicate melts

Magmatic sulfur is an essential constituent for the genesis of porphyry-type ore deposits and also significantly affects the Earth’s climate when emitted into the atmosphere. As an increasing number of studies indicate that anhydrite commonly occurs as primary igneous mineral in arc magmas, any quantitative model aiming to assess the efficiency of sulfur degassing from magmas must take the potential presence of anhydrite into account. To facilitate this, we present a new empirical model to predict the solubility of anhydrite in silicate melts as a function of pressure (P), temperature (T) and silicate melt composition. The model is based on 189 experimental data points, encompassing T from 750 to 1325 °C, P from 30 to 3000 MPa and basaltic to rhyolitic melt compositions. Fifteen of these experiments were conducted as part of this study in rapid-quench Molybdenum – Hafnium Carbide pressure vessel assemblies to obtain tighter constraints on the effect of water concentration, P and T on anhydrite solubility. These experiments show that anhydrite solubility in the silicate melt rapidly increases with increasing T following Arrhenius relationship. The solubility of anhydrite also increases with increasing dissolved water concentration, with the relative increase being the most significant for felsic melt compositions. The new model predicts sulfur concentration at anhydrite saturation (SCAS) in the silicate melt with median and mean absolute percentage errors of 19 and 25 relative%, respectively. The model is equally successful over the entire P, T and compositional space included in its calibration. This is a significant improvement over most previously published models, which were calibrated based on a more limited P, T and compositional range, and predict SCAS with significantly larger errors when applied to the entire dataset utilized in this study. The implementation of the new anhydrite saturation model in an example scenario of the crystallization of a hydrous andesite magma at P = 200 MPa suggests that anhydrite saturation may severely limit the amount of S that can be transferred into the exsolving magmatic volatile phase during crystallization-driven degassing of upper crustal magma reservoirs.

Addressing ore formation and exploration

Common base and noble metals represent an important economic factor in the actual industrial development. For instance the world resources for copper are actually estimated for about the next 30 years only. The situation requires rethinking the way major ore deposits form, leading to new guides for exploration. The present paper briefly examines the processes leading to ore formation, in relation with granitic or granodioritic intrusions. It identifies the importance of metal enrichment during the magmatic stage. Within the magma chamber that forms porphyry intrusions, metals may incorporate to first formed crystals, becoming inert; concentrate into the residual melt of a mush; or segregate by diffusion into the exsolved magmatic volatile phase (MVP) into which they are transported and precipitated. A competition results between elements partitioning and diffusivity. Hence, a specific Péclet number for each metal (Cu, Au, Ag, Mo, W, Sn, and REE) controls the ratio between the diffusive and the advective flux. Metals diffusivity in the melt shows differential behavior relative to a threshold of about 10−13 m2/s. Metals with slower diffusivity values (e.g. As) will not concentrate. Conversely, fast diffusive metals (Au, Ag, Cu) may rapidly incorporate the MVP, provided an adequate component (halogens or S) is attractive for metals. The chemistry of the MVP escaping the magma induces different alteration patterns. Their relative content in F, Cl or S, attested by the composition of biotites and apatites, links with the preferential content of metals in ore deposits, representing a valuable tool for exploration. Finally the model is replaced in a set of coupled mechanical-chemical instabilities, within a three phase material.

How do metals escape from magmas to form porphyry-type ore deposits?

Many precious and base metals (Cu, Au, Ag, Mo, W, Sn) in porphyry-type deposits are extracted from a granitic to granodioritic silicate melt and transported by magmatic-hydrothermal fluids to the site of ore deposition. Metals abundance increases from some ppb in the melt to % in ore bodies. Assuming a unique magmatic source for metals, it implies an exceptional combination of enrichment process acting over up to 4 orders of magnitude. Previous models of magma differentiation and hydrothermal circulations omitted to consider the role of metal segregation during magma chamber evolution. Models for the formation and evolution of intrusions have recently shifted from the so-called melting-storage-assimilation-homogenization (MASH) paradigm, basically steady state, to the mantle-melting-segregation-ascent-emplacement (m(M-SAE)) paradigm, dynamic and discontinuous in time. Successive magma injections of variable compositions progressively build the magma chamber. This is supported by field relations (e.g. cross-cutting dykes and stocks), textures (e.g. partially resorbed enclaves, or mineral fabrics), geochemistry (e.g. hybridization signatures), and age dating aswell as istopic studies. As melt crystallizes and forms a mush (>50% crystals), volatiles also exsolve forming a magmatic volatile phase (MVP) while melt motion slows down. Metals partition between the three phases: melt; crystals; and MVP. They generally prefer the MVP owing to more favorable partition coefficients. This is indicated by metal content in coeval melt and fluid inclusions with homogenization temperatures above 600 °C, or in volcanic fumaroles. Here, by simulating the physical interactions between the three phases we suggest a model of fluid sparging, during which metals segregate towards the MVP by diffusion and are further transported by advection as metal complexes. The model also provides estimates of metal enrichment. An undimensional Péclet number rules the competition between diffusion and advection, basically scaling with respect to the inverse of the product of melt viscosity (η) by metal diffusivity (D). The threshold value of the Péclet number between diffusion and advection is roughly 10−9. Fast diffusive metals (Au, Cu, Ag, W) readily diffuse from silicate melt and towards bubbles of the MVP. To escape the mush through tubular structures, the MVP must overcome a critical gas saturation level (about 20% vol.) usually reached after several magma injections. Advection then takes over and transports the metal-enriched MVP towards the top of the magma chamber. This leads to nearly coeval (1) separation between a high salinity-liquid phase and a low-salinity vapor phase, (2) fluid-rock interactions resulting in potassic, advanced argillic and phyllic alterations and (3) metal deposition. The metal enrichment scales as the ratio between partition coefficient (i.e. related to the gradient in chemical potential), diffusivity (i.e. related to the gradient of concentration), and melt viscosity (i.e. related to the gradient of momentum). The rates at which all such gradients relax determe metals enrichment, inducing chemical and physical instabilities, leading to a cyclic process. The whole cycle also encompasses the case of a partial, non-completed, full enrichment yielding to barren intrusions. A tentative model generalizes the sparging fluid model to other metal deposits linked to an intrusion. Such generalization should be interpreted as predicting metal enrichment by 3–4 orders of magnitude, rather than predicting an exact value.

Feldspar-brine alteration at high-temperatures

A magmatic volatile phase (MVP) may be exsolved from a silicate melt through decompression during ascent and/or the crystallization of anhydrous mineral phases. MVPs are commonly chloride-rich fluids and are hypothesized to induce hydrothermal alteration in and around igneous intrusions. Feldspars are common alteration products in these systems with K-feldspar (potassic alteration zone) often found in areas directly above cupolas of the igneous intrusions. Albite-bearing alteration assemblages are found generally on the flanks of the intrusions, below the cupolas, or more distally. Mineral alteration reactions near the intrusion are likely influenced wholly or partially by a subcritical, vapor undersaturated high-salinity liquid phase (brine). Whereas the K-feldspar-MVP system is well studied, there is a dearth of data on albite-brine systems at elevated temperatures and pressures. This study addressed mineral phase stability and equilibrium brine compositions in the albite-andalusite-quartz-brine system (65 wt% NaCl equivalent; NaCl-HCl-H2O) at 600, 650, and 700 °C and 80 MPa. The stable mineral phases were ascertained through an examination of crystal morphology, optical microscopy, and by chemical compositions. The albite + andalusite + quartz + brine phase boundary was found to be at CNaClbrine/CHClbrine molality values of 87.5 ± 12.5, 67.5 ± 7.5, and 40 ± 10 at 600, 650, and 700 °C, respectively. Apparent equilibrium constants, not accounting for activity coefficients, for the reaction: albite + HCl = 0.5 andalusite + NaCl +2.5 quartz +0.5 H2O were calculated as 70 ± 10, 54 ± 6, and 32 ± 8 at 600, 650, and 700 °C, respectively. The alteration of albite to andalusite occurs at much lower HCl concentrations than is required to alter K-feldspar to andalusite for equivalent concentrations of chloride salts. Thus, HCl concentrations of brines within the potassic and/or sericitic alteration zones can be much higher than in albite-bearing alteration zones. The data presented in this study indicate that albite-rich mineral zones on the flanks of igneous intrusions are most likely produced by a low-acidity fluid with an elevated CNaClfluid/CKClfluid. Fluids that meet these specifications, and may be significant sources of sodic alteration, include non-magmatic fluids flowing towards the intrusion or a descending magmatic brine depleted in cations and metals by mineral alteration reactions near the cupola. Although the sodic alteration zone is unlikely to host significant concentrations of metals in porphyry ore systems, it can provide significant constraints on metasomatic reactions and fluid flow pathways. Sodic and sodic-calcic alteration sequences in iron-oxide‑copper‑gold deposits are likely produced by a low-acidity, saline fluid and, unlike porphyry systems, correlate with high metal contents which indicate that the precipitation mechanisms controlling ore distribution in these systems may not be as dependent on neutralization reactions as in potassic and sericitic alteration sequences.

Cu-Mo partitioning between felsic melts and saline-aqueous fluids as a function ofXNaCleq,fO2, andfS2

The formation of porphyry Cu-Mo deposits hinges critically on the ability of an exsolving magmatic volatile phases (MVP) to efficiently extract the available Cu and Mo from the silicate melt and transport them to the site of mineralization. There has been substantial debate about the relative importance of the critical parameters likely to control metal partitioning among silicate melts and supercritical fluids, vapors, and brines. To explore the relative contributions of key MVP parameters ( X NaCleq , f O 2 , f S 2 ), we present felsic magmatic Cu-Mo partitioning experiments at both reduced ( f O 2 = NNO+0.6) and oxidized conditions ( f O 2 = NNO+2), at high f S 2 , and over the full range of salinities ( X NaCleq ) relevant to porphyry deposit formation. The experiments demonstrate that fluid-melt Cu partition coefficients ( D Cu f/m ) have a dominantly second-order exponential relationship with X NaCleq at relevant salinities, consistent with a (Na,K)CuCl 2 ion-pair complexation mechanism. We find a strong linear dependence of Cu partitioning on Cl partitioning between coexisting brine and vapor, in good agreement with limited data from unmodified natural fluid inclusions. Whereas H 2 S can increase Cu partitioning via (Na,K)CuCl(HS) formation, SO 2 has no measureable effect on Cu partitioning. These data allow for quantifying the strong partitioning of Cu out of silicate melts at MVP salinities above ~5 wt%, which will become further enriched into tardo-magmatic brines on phase separation. Our data also highlight that low-salinity ( eq ) oxidized MVPs are not capable of substantially extracting Cu from evolved silicate melts or transporting it to the site of mineralization. We also show that D Mo f/m is a linear function of X NaCleq , consistent with mono-chloride (e.g., {Na,K}MoO 3 Cl), Na-K molybdate (e.g., {Na,K}HMoO 4 ), or thio-molybdate ({Na,K}HMoO 2 S 2 ) complexation mechanisms at modest salinities (>3 wt%) rather than the Mo-oxy-hydroxy [MoO 2 (OH) 2 ] complexation observed at lower salinities. The f O 2 of the magmatic system has a subordinate effect on D M f/ o m , with enhanced partitioning at higher f O 2 . We use the combined data set to produce expressions for D Cu f/m and D Mo f/m as functions of X NaCleq , X H 2 S , and f O 2 D Cu f / m = 8.0 ( ± 1.4 ) × 10 4 [ ( X NaCleq ) 2 ( X H 2 O ) 14 ± 2 { 1 + 180 ( ± 60 ) ( X H 2 S ) } ] + 380 ( ± 50 ) ( X NaCleq ) + 0.8 ( ± 0.5 ) D Mo f / m ( at NNO + 2 ) = 430 ( ± 60 ) · ( X NaCleq ) + 1.5 ( ± 0.7 ) · ( X H 2 O ) . These equations provide Cu-Mo fluid-melt partition coefficients for common arc melt-MVP assemblages as their composition evolves through time and space. Quantitative modeling of the contrasting partitioning behavior of Cu and Mo using these equations will allow for significant improvement in understanding metal extraction and porphyry deposit formation.

The mechanics of shallow magma reservoir outgassing

AbstractMagma degassing fundamentally controls the Earth's volatile cycles. The large amount of gas expelled into the atmosphere during volcanic eruptions (i.e., volcanic outgassing) is the most obvious display of magmatic volatile release. However, owing to the large intrusive:extrusive ratio, and considering the paucity of volatiles left in intrusive rocks after final solidification, volcanic outgassing likely constitutes only a small fraction of the overall mass of magmatic volatiles released to the Earth's surface. Therefore, as most magmas stall on their way to the surface, outgassing of uneruptible, crystal‐rich magma storage regions will play a dominant role in closing the balance of volatile element cycling between the mantle and the surface. We use a numerical approach to study the migration of a magmatic volatile phase (MVP) in crystal‐rich magma bodies (“mush zones”) at the pore scale. Our results suggest that buoyancy‐driven outgassing is efficient over crystal volume fractions between 0.4 and 0.7 (for mm‐sized crystals). We parameterize our pore‐scale results for MVP migration in a thermomechanical magma reservoir model to study outgassing under dynamical conditions where cooling controls the evolution of the proportion of crystal, gas, and melt phases and to investigate the role of the reservoir size and the temperature‐dependent viscoelastic response of the crust on outgassing efficiency. We find that buoyancy‐driven outgassing allows for a maximum of 40–50% volatiles to leave the reservoir over the 0.4–0.7 crystal volume fractions, implying that a significant amount of outgassing must occur at high crystal content (&gt;0.7) through veining and/or capillary fracturing.

Hydrothermal alteration and fluid pH in alkaline-hosted epithermal systems

Epithermal gold mineralisation is found in a wide compositional range of host lithologies, but despite the diversity the alteration mineral assemblages are often similar between deposits. Notable exceptions are those gold deposits hosted in alkaline host rocks. Alkaline-hosted epithermal deposits are rare, but important, as they include some of the world’s largest known epithermal deposits by contained metal (e.g. Ladolam, Cripple Creek, Porgera). As well as the exceptional gold contents, the alkaline-hosted systems tend to exhibit different alteration mineral assemblages, with less quartz and widespread silicification than sub-alkaline-hosted equivalents, and greater enrichments in tellurium, and a scarcity of acid alteration (advanced argillic) types. In this study, geochemical modelling is used to demonstrate that 300°C hydrothermal fluids in equilibrium with alkali, silica-undersaturated host rocks at low water/rock ratios reach significantly higher pH than equivalents in sub-alkaline lithologies. A maximum, near-neutral pH (5.5–6) is buffered by reactions involving quartz in silica-saturated alkaline and calc-alkaline lithologies. In silica-undersaturated, alkaline host rocks, quartz is exhausted by progressive water-rock interaction, and pH increases to 7 and above. Both tellurium and gold solubility are favoured by neutral to high fluid pH, and thus there is a clear mechanism within these hydrothermal systems that can lead to effective transport and concentration to produce gold telluride ore deposits in alkaline igneous hosts. This modelling demonstrates that alkaline rocks can still be altered to advanced argillic assemblages; the paucity of this alteration type in alkaline hosts instead points to NaCl≫HCl in magmatic volatile phases at the initiation of hydrothermal alteration.

Cu and Fe diffusion in rhyolitic melts during chalcocite “dissolution”: Implications for porphyry ore deposits and tektites

Copper diffusion plays an important role in natural processes, such as metal transport during the formation of magmatic-hydrothermal porphyry-type ore deposits and Cu isotope fractionation during tektite formation. Copper diffusion data in natural silicate melts, however, are limited. In this study, chalcocite (Cu 2 S) “dissolution” experiments were carried out using chalcocite-rhyolite diffusion “couples” to study Cu (and S) diffusion in rhyolitic melts. Instead of chalcocite dissolution as initially expected, our experiments show that Cu is transferred from the chalcocite crystal to the rhyolitic melt, and Fe is transferred from the rhyolitic melt to chalcocite, whereas the S concentration profile in the rhyolitic melt is essentially flat. From the Cu and Fe exchange profiles in the rhyolitic melts, Cu diffusivities and Fe diffusivities are obtained and reported. Copper diffusivity in rhyolitic melts containing 0.10 to 5.95 wt% H 2 O at temperatures of 750 to 1391 °C and pressures of 0.5 to 1.0 GPa can be described as: D Cu Rhy = exp [ - ( 14.75 ± 0.35 ) - ( 0.23 ± 0.10 ) w - ( 11647 ± 491 ) - ( 698 ± 117 ) w T ] , which allows the estimation of an activation energy for diffusion in dry rhyolitic melts to be 96.8 ± 4.1 kJ/mol. In the above equation, diffusivity ( D ) is in m 2 /s, T is the temperature in K, w is the H 2 O concentration in the rhyolitic melts in wt% and all errors reported are at 1σ level. Combining Cu diffusion data from this study and previous data in basaltic melt gives a general equation for Cu diffusivity in natural silicate melts: D cu = exp [ - ( 17.3 ± 0.9 ) + ( 3.8 ± 1.5 ) ( Si + Al-H ) - ( 4403 ± 1094 ) + ( 9700 ± 1921 ) ( Si + Al-H ) T ] , where Si+Al-H is the cation mole fraction of Si plus Al minus H in the silicate melt on a wet basis. Iron diffusivities obtained in this study, in anhydrous to 6 wt% H 2 O rhyolite, are combined with previous data to get a general equation for Fe diffusion in rhyolitic melts: D Fe Rhy = exp [ - ( 16.1 ± 1.7 ) - ( 19859 ± 2541 ) - ( 1218 ± 135 ) w T ] . Our data demonstrate that Cu diffusion is faster than H 2 O or Cl in rhyolitic melts containing 6 wt% water, which indicates that the scavenging and transport of Cu by a magmatic volatile phase during formation of porphyry-type ore deposits is not limited by diffusion of Cu. Based on our experimental data, Cu diffusivity is almost four orders of magnitude higher than Zn in anhydrous rhyolitic melts, which supports the explanation of more diffusive loss of Cu leading to more fractionated Cu isotopes than Zn isotopes in tektites.

Bubble accumulation and its role in the evolution of magma reservoirs in the upper crust

Volcanic eruptions transfer huge amounts of gas to the atmosphere. In particular, the sulfur released during large silicic explosive eruptions can induce global cooling. A fundamental goal in volcanology, therefore, is to assess the potential for eruption of the large volumes of crystal-poor, silicic magma that are stored at shallow depths in the crust, and to obtain theoretical bounds for the amount of volatiles that can be released during these eruptions. It is puzzling that highly evolved, crystal-poor silicic magmas are more likely to generate volcanic rocks than plutonic rocks. This observation suggests that such magmas are more prone to erupting than are their crystal-rich counterparts. Moreover, well studied examples of largely crystal-poor eruptions (for example, Katmai, Taupo and Minoan) often exhibit a release of sulfur that is 10 to 20 times higher than the amount of sulfur estimated to be stored in the melt. Here we argue that these two observations rest on how the magmatic volatile phase (MVP) behaves as it rises buoyantly in zoned magma reservoirs. By investigating the fluid dynamics that controls the transport of the MVP in crystal-rich and crystal-poor magmas, we show how the interplay between capillary stresses and the viscosity contrast between the MVP and the host melt results in a counterintuitive dynamics, whereby the MVP tends to migrate efficiently in crystal-rich parts of a magma reservoir and accumulate in crystal-poor regions. The accumulation of low-density bubbles of MVP in crystal-poor magmas has implications for the eruptive potential of such magmas, and is the likely source of the excess sulfur released during explosive eruptions.

Systematic variations of trace element and sulfur isotope compositions in pyrite with stratigraphic depth in the Skouriotissa volcanic-hosted massive sulfide deposit, Troodos ophiolite, Cyprus

The Troodos ophiolite represents one of the best-preserved fossil analogs of modern oceanic crust and includes numerous volcanic-hosted massive sulfide deposits. The Skouriotissa deposit can be separated into a stockwork ore zone and an overlying massive sulfide lens that is covered by metalliferous sediments representing the former sulfide–seawater interface. Pyrite is the dominant sulfide mineral within these ores. The trace element composition of pyrite varies systematically with stratigraphic depth (down to ~150mbsf) probably reflecting fluid temperature variations and effects of phase separation (Co, Ni, Se, Te, Bi and Cu). Metal remobilization due to hydrothermal zone refining (Zn, Sb and Pb) and fluid-seawater mixing at the seafloor (Mo) represent further important processes controlling the pyrite chemistry. Massive sulfide-hosted sphalerite and euhedral pyrite probably formed from hot fluids (~400°C), while the occurrence of colloform pyrite indicates lower precipitation temperatures (<400°C). Similar δ18O quartz-fluid equilibration temperatures (~400°C) in the stockwork zone suggest that the Skouriotissa fluids did not cool significantly during the final 150m of fluid ascent to the seafloor. The δ34S composition of deep stockwork pyrite (−1.4‰) suggests that an isotopically light magmatic volatile phase (<0‰) was added to the hydrothermal system of Skouriotissa. During further fluid ascent about 38% of Cretaceous seawater (δ34S=18–19‰) was added leading to the precipitation of stockwork pyrite with positive δ34S values (6.1‰). In addition, the chemical and textural similarities between Skouriotissa and modern seafloor vent systems and massive sulfide deposits suggest that the Skouriotissa hydrothermal system has a modern analog.

Abundances of platinum group elements in native sulfur condensates from the Niuatahi-Motutahi submarine volcano, Tonga rear arc: Implications for PGE mineralization in porphyry deposits

Some porphyry Cu–Au deposits, which are enriched in Pd, are potentially an economic source of Pd. Magmatic volatile phases are thought to transport the platinum group elements (PGEs) from the porphyry source magma to the point of deposition. However, the compatibilities of the PGEs in magmatic volatile phases are poorly constrained. We report PGE and Re contents in native sulfur condensates and associated altered dacites from the Niuatahi-Motutahi submarine volcano, Tonga rear arc, in order to determine the compatibility of PGEs and Re in magmatic volatile phases, and their mobility during secondary hydrothermal alteration. The native sulfur we analyzed is the condensate of a magmatic volatile phase exsolved from the Niuatahi-Motutahi magma. The PGEs are moderately enriched in the sulfur condensates in comparison to the associated fresh dacite, with enrichment factors of 11–285, whereas Au, Cu and Re are strongly enriched with enrichment factors of ∼20,000, ∼5000 and ∼800 respectively. Although the PGEs are moderately compatible into magmatic volatile phases, their compatibility is significantly lower than that of Au, Cu and Re. Furthermore, the compatibility of PGEs decrease in the order: Ru>Pt>Ir>Pd. This trend is also observed in condensates and sublimates from other localities.PGE mineralization in porphyry Cu–Au deposits is characterized by substantially higher Pd/Pt (∼7–60) and Pd/Ir (∼100–10,500) than typical orthomagmatic sulfide deposits (e.g. Pd/Pt ∼0.6 and Pd/Ir ∼20 for the Bushveld). It has previously been suggested that the high mobility of Pd, relative to the other PGEs, may account for the preferential enrichment of Pd in porphyry Cu–Au deposits. However, the low compatibility of Pd in the volatile phase relative to the other PGEs, shown in this study, invalidates this explanation. We suggest that the PGE geochemistry of Pd-rich Cu–Au deposits is principally derived from the PGE characteristics of the magma from which the ore-forming fluid exsolved. Pd-rich porphyry Cu–Au deposits are associated with highly oxidized magmas. Prior to sulfide saturation Pd, Au and Cu behave as incompatible elements and concentrate in the melt with fractional crystallization, whereas Pt is depleted by early crystallization of a Pt-rich alloy and the other PGEs by the co-crystallization of the Pt-rich alloy and Cr spinel. As a consequence the Pd/Pt and Pd/Ir in the evolving melt and the magmatic volatile phases that exsolve from that melt, increase with increased fractionation. The high Pd content and high Pd/Pt (∼7–60) of Cu–Au porphyry ores therefore require the parent magma to have undergone extensive sulfide-undersaturated fractional crystallization prior to volatile exsolution.Our study also showed that the altered dacites contain PGE abundances that are similar to those of fresh dacites although Pt and Rh are slightly enriched in the altered dacites, which indicates low mobility of PGEs during secondary hydrothermal alteration.

Metal (copper) segregation in magmas

Before precipitating to form porphyry-type deposits, metals are transported and concentrated into magmas. Ultimately, they can enter crystalline phases or segregate into the volatile phase. In both cases, partition coefficients determine the partitioning according to the ambient physico-chemical conditions. Metal partitioning between the melt and the magmatic volatile phase (MVP) is driven by their solubility. In this study, Cu has been selected as a test for metal segregation. We evaluate qualitatively the metal's behavior with respect to the melt or to the MVP by comparing the difference in chemical potential and polarizability between the fluid phase and dissolved copper compound. Maps of polarizability are drawn after computing the chemical reactivity parameters (electrophilicity, hardness, and polarizability) for various silicate melts; a synthetic fluid phase with water, CO2, S18 compounds and halogens; and Cu-compounds as a test metal. Cu-compounds show a better affinity with the fluid phase, enhanced by the presence of S in its reduced form. It explains how Cu could segregate into the fluid phase at the magmatic stage before being enriched by diffusion or melt/vapor partitioning, leading to late hydrothermal precipitation. The method should therefore be considered as a model for understanding the behavior of other metals and their segregation during the magmatic stage.

Copper partitioning between felsic melt and H2O–CO2 bearing saline fluids

Analysis of fluid inclusions from porphyry copper deposits reveals that magmatic vapor and brine are vital for the removal of copper from arc magmas and its transport to the site of ore deposition. Experiments in melt–vapor–brine systems allow for investigation of the partitioning of copper between silicate melts and volatile phases at magmatic conditions. The presence of CO2 affects both the pressure at which a melt saturates with respect to volatile phases. Therefore, the partitioning of copper among felsic (rhyolitic) melt, vapor and brine in CO2-bearing experiments was examined to provide insights into copper partitioning and the generation of porphyry copper and related deposits.We present results from experiments performed at 800°C and 100MPa in CO2-bearing melt–vapor–brine systems with XCO2v+b=0.10 and 0.38. The compositions of vapor and brine inclusions, and run-product glasses, were determined during the course of this investigation. Microthermometric measurements of fluid inclusions show an increase in the salinity of the magmatic brine (∼65 to ∼70wt% NaCleq) and decrease in the salinity of the vapor (∼3.5 to ∼1wt% NaCleq) as XCO2 is increased from 0.10 to 0.38. The partitioning of copper between brine and vapor (DCub/v±2σ) increases from 25 (±6) at XCO2=0.10, to 100 (±30) at XCO2=0.38. The partitioning of copper between vapor and melt (DCuv/m±2σ) decreases from 9.6 (±3.3) at XCO2=0.10, to 2 (±0.8) at XCO2=0.38. These data demonstrate that copper partitioning in sulfur-free, CO2-bearing systems is controlled by the changes in the salinity of the vapor and brine that, in turn, are functions of XCO2. No change in the apparent equilibrium constants for Cu–Na exchange was observed in Fe-bearing experiments which supports a salinity-dependent model for copper partitioning.An existing model (MVPart) for ore metal partitioning between melt and volatile phases was modified to incorporate partitioning data from CO2-bearing experiments. The model can be used to predict trends in efficiency of removal of copper from melts into exsolving CO2-bearing magmatic volatile phases at a fixed pressure and temperature. The CO2-MVPart model predicts that CO2-rich (XCO2=0.38) magmatic vapor will remove half of the available copper from the melt, without a contribution from the brine, compared to low CO2 (XCO2≤0.10) systems. Thus, periods of CO2-rich vapor exsolution are not expected to be associated with efficient removal of copper from shallow felsic melts.

Mush microphysics and the reactivation of crystal‐rich magma reservoirs

AbstractReactivation and eruption of upper crustal crystal‐rich magma reservoirs (“crystal mushes”) following recharge has recently been invoked in numerous volcanic systems worldwide. Over the last few years, several models have been proposed for the reactivation of such mushes prior to or during eruptions. These models vary significantly in terms of predicted timescales associated with reactivation, because they assume that different physical mechanisms control the dynamics of this process. A common limitation of all the proposed models is that they parameterize the complex nonlinear multiphase dynamics that govern the evolution of these magmas in their open system reservoirs and rely on simple empirical laws. We argue that microscale physical models are a necessity if one wants to better constrain the evolution of these complex systems and the conditions that lead to eruption. As petrological observations of erupted mushes strongly support a thermal and fluid input from wet magma recharges, we have developed a pore‐scale multiphase heat and fluid transport model to understand the effect of a percolating fluid phase on the partial melting and reactivation of crystal mushes. Specifically, we use lattice Boltzmann calculations to reveal a counterintuitive feedback between volatile transport and melting in crystal‐rich environments. We find that partial melting, even at a low degree, can significantly reduce the efficiency of the buoyant migration of exsolved volatiles in the mush and therefore negatively impact the heat transfer upward during reactivation. This negative feedback between melting and volatile transport is expected to significantly affect the distribution of exsolved volatiles in the reservoirs, as well as the transport of trace species carried by the volatile phase (e.g., S, metals). The presence of a disperse magmatic volatile phase (unconnected bubbles) will also affect the thermomechanical properties of the mush during reactivation, making it more compressible and thermally less conductive.

Magmatic–hydrothermal evolution and devolatilization beneath Merapi volcano, Indonesia

At Merapi volcano, Indonesia, explosive eruptions lasting several months result from injections of reduced mafic magmas and are interspersed by periods of lava dome growth and collapse, quiescent degassing, assimilation of wall rock and fractional crystallization lasting a few years. Silicate melt inclusions and volcanic gases were analyzed to study processes of magmatic volatile exsolution and volcanic degassing during periods of quiescent degassing at Merapi. Volcanic gases were sampled at Merapi in 1994 during a phase of quiescent degassing and in 2006 at the end of a dome-growth and collapse explosive eruption. Silicate melt inclusions were collected and analyzed from lavas and scorias for their major element and volatile components. Solubility relationships between H2O and CO2 in melt inclusions demonstrate that melt was being trapped beneath Merapi to depths reaching 19km. Host amphibole geobarometry indicates crystallization pressures reaching up to 26km. Hence there is evidence of deep pooling and crystallization of magma beneath Merapi at appreciable depths. Using a mass balance model of magmatic volatile exsolution, we demonstrate that the magma degassing from 19km to the surface approaches closed-system behavior. Compared to the bulk magmatic volatiles analyzed in melt inclusions, volcanic gases had similar amounts of CO2, were enriched in H2O and S and depleted in Cl and F. We thus propose that the magmatic volatile phase was initially exsolved from the magma as a supercritical fluid and that it subsequently exsolved into a H2O–Cl–F-rich brine and CO2–S-rich vapor. According to the H2O–NaCl model, brine-vapor exsolution occurred at ~5km in depth for a 5wt.% NaCl, 900°C to 1000°C fluid, although the presence of H2S, SO2 and CO2 may cause the supercritical fluid to unmix at greater depths. Magma at shallower depths exsolved a brine and a vapor directly from the melt. The volcanic gases at Merapi thus represent the vapor fraction of a magmatic volatile phase, as brine ponded below the surface. We further suggest that carbon dioxide liberated by the magmatic assimilation of limestone, as proposed by other studies, fluxed the melt thereby promoting CO2 enrichment of the melt and H2O degassing. Finally, we propose that Merapi volcanic gases are enriched in sulfur compared to the modeled volatile phase composition because deep, reduced mafic magma supplies sulfur to the shallow magmatic plumbing system.

The behavior of Cu, Zn and Pb during magmatic–hydrothermal activity at Merapi volcano, Indonesia

Understanding the behavior of Cu, Zn, and Pb in magmatic–hydrothermal systems is essential in developing models for the genesis of hydrothermal ore deposits in subduction environments. A commonly held view is that the metals originate in magmas of intermediate to felsic composition from which they partition directly into exsolving aqueous fluids. In this paper, we build on the results of an earlier study in which we showed that Fe, Cu, Co and Ni at Merapi volcano, Indonesia, were transferred from a mafic melt to an immiscible sulfide melt, and then to a magmatic volatile phase which carried them to the surface. Here we examine the pathways taken by the volatile phase and the behavior of Cu, Zn and Pb in the upper part of the magmatic system beneath Merapi volcano. Based on the composition of a suite of silicate melt inclusions and thermodynamic modeling, we show that the mafic melt did not hybridize with the resident felsic melt but instead transferred metals to the latter by exsolving a magmatic volatile phase which dissolved the Cu-rich sulfide melt and then percolated upwards through the felsic melt, enriching it in Cu and Pb. It is this fluid which transported the metals to the surface and potentially also to subsurface environments of ore formation.

A physical model for metal extraction and transport in shallow magmatic systems

The highest concentrations of metals (e.g., Cu, Au, Ag, Mo) in the Earth's crust are found in porphyry‐type deposits. The metals are ultimately sourced from magmas, and appear to be concentrated hundred to thousand‐fold from typical magmatic contents (ppm‐ppb) in the exsolved volatile phase. To better quantify the purging and transport of metals, we develop a physical model of volatile evolution in an incrementally built upper crustal magma reservoirs that considers (1) partitioning of metals from the melt to an exsolved volatile phase, and (2) advection of the buoyant volatile phase using a single dimensionless parameter, the Péclet number (Pe; ratio of advection rate over diffusion rate). We propose that metal extraction and segregation from magmas can occur in 3 stages with different Pe: (1) during exsolution of the magmatic volatile phase in shallow, crystal‐poor magma bodies (slow volatile advection; Pe ≪ 1), (2) during the growth of volatile channels that develop in the reservoir as crystallinity increases (Pe &lt; 1), and (3) during advection in connected channels (rapid volatile advection, high Pe ≥ 1). For each stage, a metal enrichment factor can be calculated, allowing insight into the optimal conditions to maximize metal mass flux into the overlying hydrothermal system. The model predicts that the most efficient purging of metals occurs for magmas with intermediate volatile contents and is enhanced during late‐stage magmatic activity, as the reservoirs reach high crystallinity and are not disturbed by volcanic venting, in agreement with natural observations suggesting that ore formation post‐dates volcanic activity.