Journal of Petrology Advance Access originally published online on April 10, 2008
Journal of Petrology 2008 49(5):911-935; doi:10.1093/petrology/egn011
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Pre-eruptive Conditions of the Huerto Andesite (Fish Canyon System, San Juan Volcanic Field, Colorado): Influence of Volatiles (C–O–H–S) on Phase Equilibria and Mineral Composition
1Mineralogisch-Geochemisches Institut, Albertstrasse 23b, D-79104 Freiburg, Germany
2Institut für Mineralogie, Universität Hannover, Callinstrasse 3, D-30167 Hannover, Germany
RECEIVED NOVEMBER 8, 2007; ACCEPTED FEBRUARY 15, 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL AND ANALYTICAL...RESULTS: PHASE EQUILIBRIA AND...DISCUSSIONCONCLUSIONS: IMPLICATIONS FOR...REFERENCES |
|---|
Crystallization experiments at 400 MPa, oxidized condition (
logfO2 = NNO + 1, where NNO is nickel–nickel oxide buffer) and over a range of temperatures (850–950°C) and fluid composition (XH2Oin = 0·3–1) have been carried out to constrain the storage conditions of the sulphur-rich magma of the Huerto Andesite (an anhydrite, pyrrhotite, and S-rich apatite-bearing, post-Fish Canyon Tuff mafic lava). The results are used to evaluate the role of fluids released from the crystallization of magmas such as the Huerto Andesite on the remobilization of the largely crystallized dacitic Fish Canyon magma body. Experiments were performed using the natural andesitic bulk composition with and without added sulphur. The presence of sulphur slightly affects the phase equilibria by changing the phase proportions, stability fields of plagioclase, pyroxenes and ilmenite, and also affects the plagioclase composition. Phase equilibria and mineral composition data indicate that the magma may have contained 4·5 wt % water in the melt and that the pre-eruptive temperature was 875 ± 25°C. Assuming that the magma was in equilibrium with a fluid phase, the CO2 concentration of the melt is estimated to be in the range 2000–4000 ppm (at 400 MPa). Before eruption, the andesite had an oxidation state very close to, or slightly within, the co-stability field of anhydrite–pyrrhotite at NNO + 1·1. At these conditions, the sulphur content in the melt is 
500 ppm. Assuming open-system degassing resulting from continuing crystallization at depth, most of the CO2 dissolved in the andesitic melt should be released after the crystallization of <10 vol. % of the magma, corresponding to a cooling from 875 to 825–850°C. Thus, the fluids released owing to crystallization processes should be mainly composed of water at temperatures below 825°C. KEY WORDS: experimental study; andesite; volatile; Fish Canyon Tuff; Huerto Andesite
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL AND ANALYTICAL...RESULTS: PHASE EQUILIBRIA AND...DISCUSSIONCONCLUSIONS: IMPLICATIONS FOR...REFERENCES |
|---|
The importance of volatile constituents (H2O, CO2 and S) in magmatic processes is now well established. Volatiles influence crystal–melt phase relations and the order of crystallization of minerals from silicate melts, as well as melt dynamics, and consequently processes such as mixing, assimilation and differentiation. They may play an important role in subduction-related tectonic settings in the generation of large silicic magma bodies. Although partial melting of crustal materials and fractional crystallization of more mafic parent magmas are the two mechanisms commonly invoked to explain the generation of crystal-poor silicic magmas, reheating and partial remobilization of a crystal mush has also been considered as a possible mechanism for producing large silicic magma bodies (Sisson & Bacon, 1999
; Bachmann & Bergantz, 2003). Upward percolation of a hot fluid phase through upper-crustal magma mushes may facilitate the segregation of melt for the generation of voluminous rhyolite by their reheating and partial remelting. Possible examples of thermally rejuvenated and remobilized crystal-rich magmas include recent intermediate to silicic eruptions such as the Montserrat andesite (e.g. Couch et al., 2001) and Pinatubo dacite (Pallister et al., 1992), but also larger volcanic units such as the Kos Plateau Tuff (Allen, 2001) and the 5000 km3 Fish Canyon Tuff (Bachmann et al., 2002), one of the most voluminous pyroclastic units worldwide.
The Fish Canyon Tuff of the San Juan volcanic field in Colorado (USA) is a well-documented example of a voluminous, unzoned, phenocryst-rich pyroclastic deposit (Lipman et al., 1997) (Fig. 1). The water- and sulphur-rich lavas of the Huerto Andesite (hornblende- and anhydrite–pyrrhotite-bearing calc-alkaline lavas, Parat et al., 2005) erupted after the emplacement of the Fish Canyon Tuff. Bachmann & Bergantz (2003, 2004) suggested that the release of volatiles from a hotter, more mafic magma stored beneath the Fish Canyon Tuff magma would have contributed to the rejuvenation of a partially solidified batholith, causing the eruption of the Fish Canyon Tuff. The Huerto Andesite lavas possibly could, therefore, represent the degassing mafic magma source. Assuming this hypothesis to be correct, the main problem that needs to be resolved is whether or not the pre-eruptive conditions of the Huerto Andesite are compatible with the temperature and amount of volatiles required.
|
Phase equilibria experiments were conducted to constrain the conditions at which the phase assemblage of the Huerto Andesite could be reproduced and to constrain the composition of the fluid phase in equilibrium with the melt. We investigated (1) how changing volatiles (H2O, CO2 and S) affect the phase equilibria, crystalline phase proportions and compositions, (2) the solubility of H2O, CO2 and S in andesitic silicate melts, and (3) the partitioning of sulphur between silicate melt, minerals and fluid as a function of temperature and the fugacities of the volatile species.
Pre-eruptive conditions estimated for the Huerto Andesite and Fish Canyon magma system
The Huerto Andesite is a crystal-rich andesite (44 vol. % phenocrysts) with a hyalopilitic texture and phenocrysts of plagioclase (normally zoned from An60Or1 to An45Or3; 28 vol. %) and amphibole (11 vol. %), plus minor to sparse augite [Mg-number = 100 x Mg/(Mg + Fetotal) = 76], apatite, magnetite, ilmenite, pyrrhotite and anhydrite (Table 1) (DA4; Parat et al., 2005). The presence of euhedral, normally zoned amphibole phenocrysts (Mg-number = 66–55) without reaction rims suggests that they crystallized from an intermediate composition water-rich magma.
|
The association of the Huerto Andesite (28 Ma) with a large dacitic caldera indicates that the andesitic parent magma was stored beneath the silicic magma chamber in the upper crust (Bachmann et al., 2002
). The experimental study of Johnson & Rutherford (1989) and hornblende thermobarometry calculations of Bachmann et al. (2002) indicate a pressure of 240 MPa for the Fish Canyon magmatic system. In contrast, Bachmann & Bergantz (2003) estimated a saturation pressure of 460 MPa for the underlying mafic magma [at 900°C with 5 wt % H2O using the code VolatileCalc of Newman & Lowenstern (2002)]. Considering these estimated pressures and petrographic and isotopic evidence that the hornblende-bearing water-rich andesite magma ponded beneath the Fish Canyon magma body at a relatively deep crustal level (Parat et al., 2005), we have chosen to conduct all our experiments at a pressure of 400 MPa. The water content in the melt must have been relatively high, as hornblende is stable. Temperatures of coexisting pyroxenes and oxides in the andesite have been estimated using the geothermometers of Lindsley & Frost (1992) and Ghiorso & Sack (1991), respectively (Parat et al., 2005). The pre-eruptive temperatures are in the range 850–950°C.
Taking all the available information into account, the phase relationships of the Huerto Andesite magma were investigated by performing crystallization experiments, with and without sulphur, in the temperature range 850–950°C with different fluid composition (H2O–CO2). The presence of both anhydrite and pyrrhotite in the natural andesite constrains the oxygen fugacity at NNO to NNO + 1· 5, where NNO is nickel–nickel oxide buffer (expressed as log fO2) (Carroll & Rutherford, 1988).
|
EXPERIMENTAL AND ANALYTICAL METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL AND ANALYTICAL...RESULTS: PHASE EQUILIBRIA AND...DISCUSSIONCONCLUSIONS: IMPLICATIONS FOR...REFERENCES |
|---|
Preparation of charges
All experiments were performed on a natural andesitic sample (composition in Table 1). Dry glass was prepared by fusing 10 g of the crushed andesite in a Pt crucible at 1400°C and 1 atm for 3 h. After fusing, the glass was crushed and fused again at the same conditions to avoid heterogeneity. The resulting glass was finely crushed in an agate mortar (grain size <200 µm). Electron microprobe analysis showed that this glass is homogeneous within analytical uncertainty and that Fe and Na losses during fusion were minimal (compare X-ray fluorescence analysis of natural sample with microprobe analysis of dry glass; Table 1).
Charges were loaded into Au capsules (15 mm length, 2·8 mm internal diameter and 0·2 mm wall thickness). First, water as pure H2O and then CO2 as silver oxalate (Ag2C2O4, which decomposes to Ag and CO2 at low temperature) were loaded into the capsule. The rock powder was added in a second stage. The weight proportion of H2O + CO2[(H2O + CO2)/(H2O + CO2 + silicate), in wt %] was fixed to be around 0·1 in all experiments (see Holtz et al., 1992). The mole proportion of H2O, XH2Oin[= H2O/(H2O + CO2), molar], loaded in the capsule varied between 0·33 and unity. Sulphur was added as elemental sulphur with 1 wt % added to the rock powder. Additional experiments were conducted using anhydrite as the source of sulphur to determine the role of the sulphur in the starting material on the experimental results (see Clemente et al., 2004). The amount of added anhydrite was 4·25 wt % CaSO4, which represents an additional amount of 1 wt % S and 1·75 wt % CaO. We use the term ‘S-bearing’ for experiments with sulphur as elemental sulphur and ‘anhydrite-bearing’ for experiments with anhydrite as the source of sulphur.
Upon loading, the capsules were sealed by arc welding, surrounded with wet paper and frozen in liquid nitrogen to avoid water and CO2 loss, and then weighed to check for leaks. After completion of the experiments, the capsules were weighed again to ensure that there had been no loss of water and CO2 at high pressure and temperature. To evaluate the extent of iron and sulphur loss to the sample containers during the experiments, the gold capsules were analysed by electron microprobe. Results showed that iron and sulphur losses are small (<0·1 wt % FeO and <0·07 wt % SO2).
Experiments
All experiments were performed at 400 MPa in the temperature range 850–950°C (Table 2) in vertical, internally heated pressure vessels (University of Hannover) fitted with a Shaw-type membrane for the control of H2 fugacity (fH2) and a rapid-quench sample holder to avoid crystallization during quenching (Berndt et al., 2002). The pressure medium was a mixture of Ar + H2. The redox state of the experimental charges was controlled by directly reading PH2 and thus fH2, with a Shaw membrane connected to a transducer. The time needed to attain osmotic H2 equilibrium between the vessel and the H2-sensor membrane is typically about 30 h at 1150°C and 200 MPa (Berndt et al., 2002). Our experiments were conducted at lower temperature but the duration was longer (5 days). In six experimental series (each experimental series consists of several capsules at the same pressure, temperature and PH2, but with different XH2Oin) we recorded the PH2 and noted that it became constant at the end of the experiment, indicating that osmotic equilibrium was reached (PH2 recorded by the Shaw-membrane is close to that prevailing in the vessel). In two additional series, osmotic equilibrium between the vessel and the Shaw membrane system was not fully reached (runs 11, 12, 13, 19, 15, 16 and 20, Table 2). The prevailing oxygen fugacity (fO2) in these experiments was estimated using the mineral compositions of magnetite and ilmenite pairs and of pyroxene. The difference between the fO2 calculated from the PH2 (given by the Shaw-membrane) and the prevailing fO2 in the capsule is estimated to be 1 log unit.
|
Under water-saturated conditions, the proportion of H2 added to the pressure medium (Ar) was chosen so as to reach an fO2 corresponding to NNO + 1 [calculation of fO2 based on the equation of Schwab & Küstner (1981
)]. It should be noted that for water-undersaturated experiments, fH2O, and thus fO2, decreases with decreasing XH2Oin. The fO2 in the water-undersaturated experiments can be estimated using the relation log fO2 = log fO2 (at XH2Oin = 1) – 2 log XH2Oin (Whitney, 1972; Webster et al., 1987; Scaillet & Evans, 1999). We estimate that the overall error on the calculated fO2 is about 0·2–0·3 log units (Botcharnikov et al., 2005).
Electron microprobe analysis
Experimental products (minerals and glasses) were analysed using a Cameca SX100 microprobe (University of Hannover and Freiburg). All mineral phases were analysed using an accelerating voltage of 15 kV. The beam current was 6 nA for glass analyses, 15 nA for silicate mineral, oxide and apatite analyses, and 10 nA for anhydrite analyses. The beam was defocused to at least 20 µm for glass and focused to <2 µm for analysis of crystalline phases. All elements were analysed for 10 s, except sulphur (60 s) and fluorine (30 s). Standards used for calibration were Fe2O3, MgO, MnTiO3 (Mn and Ti), albite (Na), wollastonite (Si for mineral and glass and Ca for glass), apatite (P and F for glass and F, Ca and P for apatite), orthoclase (K), anhydrite (Ca and S for anhydrite and S for apatite and glass) and pyrite (S for pyrrhotite).
The standard deviations (2–18 analysis points) given in the tables and figures for mineral and glass analyses correspond to 1 S.D. of replicate analyses. Because of the small size of the crystals (<5 µm) in some experiments at low XH2Oin, it was not possible to obtain representative analyses of the composition of these minerals (in such cases, the analyses are not reported in the tables). However, phases have been qualitatively identified [by energy-dispersive spectrometry (EDS) + microprobe] and the characterization of the stability field of individual minerals was possible.
The water content of the experimental glasses was determined using a series of standards with a rhyolitic composition and different water contents (0–8 wt % H2O), following the ‘by-difference’ method described by Devine et al. (1995) and Koepke (1997). Using this technique, the uncertainty is in the range ±0·5–0·7 wt % H2O. The water content of three experiments at water-saturated conditions (XH2Oin = 1; no S added) determined ‘by-difference’ is 7·8, 8·2 and 8·4 at 850, 900 and 950°C, respectively. These values are within the water solubility range determined for andesitic melts at 400 MPa by other techniques (8 wt % H2O by Karl Fischer titration, Botcharnikov et al., 2006) and show the reliability of the ‘by-difference’ method using the analytical conditions described above.
To determine sulphur speciation in the melt, a peak search was performed for S-K
radiation for sulphur using a Large-PET crystal (2d = 0·875 nm), a beam current of 20 nA and a 20 µm spot. The mean peak position for S in the experimental glass (four spots) was compared with the mean position for four spots measured on standards of pyrite (S–) and anhydrite (S6+). The difference in peak position between anhydrite and pyrite, 
(S-K) = 1·3 eV, is in agreement with previous data on S peak searching [compilation by Matthews et al. (1999)].
|
RESULTS: PHASE EQUILIBRIA AND MINERAL AND MELT COMPOSITION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL AND ANALYTICAL...RESULTS: PHASE EQUILIBRIA AND...DISCUSSIONCONCLUSIONS: IMPLICATIONS FOR...REFERENCES |
|---|
Melt, minerals and bubbles are homogeneously distributed in the experimental products. The silicate minerals are generally similar in size (mostly smaller than 100 µm) and euhedral. Depending on the temperature and water content of the melt, the crystalline phases were plagioclase, amphibole, clinopyroxene, orthopyroxene, apatite, magnetite, ilmenite, anhydrite, and pyrrhotite. Quenched melt (glass) is observed in all run products, and in the following discussion we use the term ‘melt’ for the experimental glass. Liquidus conditions were approached in the CO2-free runs at 950°C, in which only anhydrite and pyrrhotite were stable in the S-bearing run and only magnetite was stable in S-free run (< 1 vol. % crystals, Table 2). Crystalline phases identified in the run products are given in Table 2. Figure 2 shows back-scattered electron images of the experimental products obtained at 850°C with different initial fluid composition (XH2Oin).
|
Phase relations
The phase stability fields for the andesite have been determined at 400 MPa in the temperature range 850–950°C and for 3–8 wt % water content in the melt (H2Omelt). Figure 3 shows the phase stabilities as a function of temperature and water content in the melt for S-bearing and S-free experiments (the melt water content is taken as the parameter reflecting XH2Oin).
Amphibole is the only silicate phase stable in CO2-free experiments at high temperature (875–900°C) and coexists with plagioclase at lower temperature (850°C). Amphibole, plagioclase and pyroxenes are stable together at T = 900–850°C and H2Omelt = 4·5–5 wt % in S-bearing runs and H2Omelt = 5–6 wt % in S-free runs. At H2Omelt <4 wt %, amphibole is not stable, whereas clinopyroxene and orthopyroxene are stable. Plagioclase crystallizes from a melt with H2Omelt lower than 4·5 wt % at 950°C and up to 7 wt % at low temperature (850°C, XH2Oin = 1). The presence of S influences the stability of plagioclase, clinopyroxene, orthopyroxene and ilmenite, but not the stability field of amphibole. There is a shift of the stability field of plagioclase to lower temperature in S-bearing ‘water-rich’ melt (H2Omelt >5 wt %). Clinopyroxene and orthopyroxene are stable minerals in S-free melt with H2Omelt <5·5 wt % and H2Omelt <4·5 wt % when S is present. Ilmenite crystallizes only in S-bearing experiments and coexists with pyrrhotite and magnetite. In the investigated temperature range, the water content does not strongly influence the stability field of pyroxenes (Fig. 3). This observation differs from previous experiments with andesitic compositions performed at higher temperature, where the pyroxene stability field strongly depends on water content (e.g. Martel et al., 1999; Botcharnikov et al., 2008). However, our experiments are restricted to a small temperature range (100°C) and the extension of the stability curves of these minerals toward higher temperatures should reveal that ‘orthopyroxene-in’ and ‘clinopyroxene-in’ curves are expected to be dependent on water content in the melt.
|
Apatite has been observed only in charges at 950°C. The curve showing the beginning of apatite crystallization is indicated in Fig. 3. Apatite has not been detected in most experiments with high crystal contents (low temperature or low H2Omelt; Table 2). Considering our knowledge of apatite solubility, this mineral should be present in such experiments. The low P2O5 content of coexisting melts in these experiments confirms that a P-bearing phase must have crystallized. Apatite may have formed very tiny phases, which could not be detected.
Magnetite is present in most S-bearing and in all S-free charges. Pyrrhotite and anhydrite are liquidus phases at 950°C. The crystallization of both sulphate and sulphide depends on the redox state of the experiment, in agreement with previous studies on anhydrite–pyrrhotite stability (Carroll & Rutherford, 1988). Anhydrite crystallizes in charges with an oxygen fugacity higher than NNO + 1, whereas pyrrhotite is present in all S-bearing charges (up to NNO + 1·2). In CO2-free and anhydrite-bearing experiments (number 9, Table 2), pyrrhotite is absent (and anhydrite is present) despite the relatively low oxidation state (NNO + 1·12). Under water-undersaturated conditions (H2Omelt 5 wt %) in anhydrite-bearing experiments, clinopyroxene replaces amphibole (in S-bearing experiments), probably as a result of a higher CaO content in the starting material (+1·75 wt % CaO).
Crystallinity
The phase compositions (SiO2, TiO2, Al2O3, FeO*, MnO, MgO, CaO, Na2O, K2O, and P2O5) were used to calculate modes for all charges, based on a least-squares fit of the starting composition using the approach of Stormer & Nicholls (1978). Calculated modes are given in Table 2 (wt %), together with the sum of the squares of the residuals to the fit (
r2). Fits were poor for several charges at 850 and 950°C, with r2 values >0·6, but most fits were satisfactory, with r2 <0·3.
The melt fraction decreases and the proportion of crystalline phases increases with decreasing XH2Oin and wt % H2Omelt at all investigated temperatures (Fig. 4, Table 2). The variation of melt fraction at 950°C is low, reflecting near-liquidus conditions. At lower temperatures, the melt fraction decreases by 50 wt % for a variation in water content of 4 wt % (Fig. 4a). As expected, the crystal fraction increases with decreasing temperature (at a given H2Omelt). The variation is c. 10 wt % for a temperature change of 50°C. For a given H2Omelt, the proportion of the melt in the S-bearing charges is up to 16 wt % higher than in S-free charges (at 900°C, 5 wt % H2Omelt). Among the stable silicate phases, only plagioclase shows a significant modal change related to the presence of sulphur (variation up to 18 wt % at 900°C, H2Omelt = 5 wt %; Fig. 4b). No differences are observed in amphibole and clinopyroxene modal proportions, whereas the amount of orthopyroxene is slightly lower in the S-bearing runs (3 wt % in S-bearing charge and 5 wt % in S-free charge at 900°C, H2Omelt = 5 wt %; Table 2). Magnetite is less abundant in the S-bearing charges in which pyrrhotite and ilmenite are stable phases. The amount of magnetite clearly increases with a decrease in temperature in S-free runs (e.g. from 2·6 to 3·4 wt % for T from 900 to 850°C, respectively, for H2Omelt = 5·8 wt %).
|
Mineral and melt compositions
Silicate phases
The anorthite (An) content in plagioclase increases strongly with an increase in H2Omelt (up to An75 for H2Omelt
8 wt % at 850°C) and temperature (e.g. from An44 to An65 for T from 875 to 950°C in S-bearing runs, respectively; Fig. 5, Table 3). The addition of sulphur in the system also results in an increase of the An content of plagioclase (up to 8 mol % at 900°C, H2O = 5 wt %; Fig. 5) that correlates with changes in crystallinity and melt composition. The orthoclase content increases significantly (from Or1 to Or6) with a decrease in water content (Table 3).
|
|
Experimental amphiboles are euhedral unzoned calcic amphiboles [tschermakite, (Ca + Na)B >1; NaB <0·5; Leake et al., 1997
] (Table 4). No significant variations in amphibole composition are observed as a function of temperature or H2Omelt. The total aluminium content in amphibole slightly increases with an increase of H2Omelt in S-free charges, whereas no significant changes are observed in S-bearing charges (Fig. 6). Mg-number is slightly higher in S-free runs and increases with temperature and oxygen fugacity (Fig. 7a).
|
|
|
The composition of pyroxenes is given in Table 5. Both augite and hypersthene are stable minerals over the range of experimental temperature (850–950°C). For augite, the wollastonite content (Wo) increases, the ferrosilite (Fs) content decreases and the enstatite (En) content is constant with an increase in oxygen fugacity and water content for a fixed temperature, whereas hypersthene shows an increase in En content, a decrease in Fs content and Wo content is constant. In S-free runs, at fixed oxygen fugacity, for both hypersthene and augite, the Mg-number decreases (and Fs increases) with decreasing temperature, whereas no significant changes are observed in S-bearing charges (Fig. 7b).
|
Accessory minerals
Particular attention was given to the presence of accessory mineral such apatite, ilmenite, magnetite, anhydrite and pyrrhotite because they permit characterization of the intensive parameters, the volatile content in the melt and the composition of the fluid phase in the natural system (e.g. Carroll & Rutherford, 1987
; Ghiorso & Sack, 1991; Parat et al., 2002; Scaillet & Pichavant, 2003). Because of the small size of the accessory minerals (<10 µm), analyses may have been contaminated by glass. To interpret the composition of the accessory minerals we have selected representative analyses (the analyses with the lowest silica content), given in Tables 6–9.
Apatites are hydroxy-fluor-apatites (1· 8 wt % F; XOH 0·6 and XF 0·4) and sulphur-rich in the S-bearing runs (up to 0·7 wt % SO3) (Table 6). The Na2O content in apatite from the S-free runs is below the detection limit, whereas apatite in S-bearing runs has up to 0·14 wt % Na2O, suggesting that in S-bearing runs both Na and S are involved in sulphur exchange reactions. The sulphur content in apatite decreases with a decrease in temperature and a decrease of sulphur content in the melt, in agreement with previous experimental data (Parat & Holtz, 2004, 2005). Sulphur substitutes into the tetrahedral site as the S6+ ion and several exchange reactions have been proposed based on natural apatite composition (e.g. Rouse & Dunn, 1982; Liu & Comodi, 1993). The good positive correlation between Si + Na and S + REE (rare earth element) contents observed for experimental apatites in S-bearing charges suggests that charge balance is maintained through the coupled substitution Si4+ + 2Na+ + 2S6+ + 4REE3+ = 4P5+ + 5Ca2+ in agreement with sulphur exchange in natural apatite from the Huerto Andesite (Parat et al., 2002). This sulphur exchange reaction differs from the one described in Parat & Holtz (2004, 2005) for experimental apatites coexisting with REE-free rhyolitic melt (simple substitution between Na and S), indicating that the melt trace element composition (e.g. REE content) probably influences the sulphur exchange reaction in apatite.
|
The ulvöspinel component (Xulvö) of magnetite in S-bearing and S-free charges and ilmenite component (Xilm) of ilmenite in S-bearing charges decrease with an increase in oxygen fugacity (Fig. 8; Tables 7 and 8). The MgO, MnO, and Al2O3 content of magnetite are identical in S-bearing and S-free charges. The MgO, and to a lesser extent Al2O3, content of magnetite increases with temperature (up to 2·8 and 4·6 wt % at 950°C, respectively). Magnetites in S-bearing runs have higher TiO2 and slightly lower Fe2O3 content than in S-free charges, resulting in higher amounts of the ulvöspinel component. Ilmenite is present only in S-bearing charges and has a relatively high ilmenite component (Ilm83–91) (Fig. 8b). Three charges have coexisting ilmenite and magnetite. Using the geothermometer formulated by Ghiorso & Sack (1991
), we calculated temperatures in agreement with experimental temperature within ±30°C and a lower oxygen fugacity by 0·3–1 log units, which may indicate a lower oxidation state or an influence of pyrrhotite crystallization on the determination of oxygen fugacity based on iron–titanium oxides.
|
|
|
Pyrrhotite is stable in all S-bearing charges except at NNO + 1·1 in the water-saturated anhydrite-bearing run (number 9, Table 2). It occurs as hexagonal crystals (10–100 µm) but most charges also contain pyrrhotite globules that may represent either pyrrhotite or quenched immiscible sulphide liquid (Fe–O–S). NFeS (=Fe2+/S) of pyrrhotite (Table 9) increases with decreasing oxygen fugacity, whereas no variations are observed with changing temperature.
|
Melt composition; major elements
The composition of the experimental melts ranges from andesite to dacite (57–69 wt % SiO2) and varies according to temperature and water content. The composition of the melt at 950°C is relatively constant because the crystal proportion in these experiments is low (2 to
15 wt %, Fig. 9 and Table 10). The evolution paths are similar at 900, 875 and 850°C: CaO, MgO, FeO, Al2O3 and TiO2 contents decrease with a decrease in water content and temperature reflecting plagioclase, pyroxenes, amphibole and oxides crystallization, whereas K2O increases (a K-bearing phase is absent). Na2O is nearly constant and varies in the range 3·2–3·9 wt % over the range of experimental temperatures.
|
|
Changes in melt composition in the S-free and S-bearing experiments correlate with the observed changes in the phase proportions. S-bearing melts are enriched in Al2O3, CaO, TiO2, and to a lesser extent in FeO and MgO, and depleted in SiO2 and K2O. Al2O3 and CaO concentrations reflect a lower plagioclase proportion in the S-bearing charges than in the S-free charges. SiO2 and K2O concentrations are related to the crystal proportion, which is lower in the S-bearing charges. TiO2 in the melt varies as a function of temperature and is controlled by the amount of magnetite in all S-free runs and in S-bearing runs at 950°C. In contrast, in S-bearing charges at T <950°C, the variation of TiO2 content in the melt is controlled by the crystallization of ilmenite and magnetite. We note that the melt composition is identical in S-bearing and anhydrite-bearing experiments (even the CaO content), indicating that the addition of anhydrite as a source of sulphur does not change the melt composition, but rather only the phase assemblage and/or proportions.
Melt composition; volatile elements
The addition of sulphur as elemental S results in a decrease in XH2Oin in the capsule and thus in H2Omelt. This is confirmed by the water content determined in S-bearing melts: in experiments with CO2-free fluid phase (initial fluid phase composition), H2Omelt is up to 1·1 wt % lower in S-bearing charges than in S-free charges (e.g. at 900°C: 7·15 and 8·24 wt % in S-bearing and S-free melt, respectively; Table 2, Fig. 10). Such a decrease in H2Omelt is consistent with the observations of Scaillet & Macdonald (2006), who estimated that the XH2Oin in the equilibrium fluid decreases from 1 to 0·9 (or less than 0·9) with the addition of 1 wt % S (for experiments with approximately the same silicate/fluid ratio as in this study).
|
The sulphur content in melts coexisting with pyrrhotite is plotted in Fig. 11 as a function of fS2 (calculated from pyrrhotite composition). The sulphur content is close to that determined in previous studies of intermediate melt compositions at oxidizing conditions (Baker & Rutherford, 1996
; Clemente et al., 2004; Parat & Holtz, 2005; Scaillet & Macdonald, 2006). The highest value (2033 ppm) was obtained for the CO2-free run at 950°C. There is a general increase in the sulphur content of the melt with f S2, as expected. In CO2-free charges, the increase in sulphur content with fS2 is strong when compared with expected trends for basaltic or andesitic melts at a given temperature (Scaillet & Pichavant, 2003). This effect may be due to a positive dependence of sulphur solubility on temperature. The compositional change of the melts as a function of temperature (slightly more silicic at low temperature; 60·8 and 57·5 wt % SiO2 at 850 and 950°C, respectively) may also influence sulphur solubility, but this is probably not strong enough to explain such an increase between 850 and 950°C (126 and 2033 ppm, respectively). The S-K peak measured on experimental glass (900°C, run 1) is close to that measured on anhydrite [(S-K)anh = 0·97 eV] indicating that sulphate dominates over sulphide and that the molar fraction of sulphur as SO4 in the melt is 
(Matthews et al., 1999).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL AND ANALYTICAL...RESULTS: PHASE EQUILIBRIA AND...DISCUSSIONCONCLUSIONS: IMPLICATIONS FOR...REFERENCES |
|---|
Pre-eruptive conditions of the Huerto Andesite at 400 MPa
At 400 MPa, the experiments essentially reproduced the natural phase assemblage and proportion and mineral compositions of the Huerto Andesite. The natural phase assemblage corresponds to a very restricted field of T–H2O conditions (Fig. 3). In the S-bearing system, the co-crystallization of plagioclase, hornblende, clinopyroxene, apatite, ilmenite and magnetite requires a temperature below 900°C and a restricted H2Omelt of 4–5 wt % (XH2Oin
0·7). The amount of melt (groundmass) was also reproduced for this range of water contents in the temperature range 850–900°C (Fig. 4).
Additional constraints on pre-eruptive conditions are provided by a comparison of experimental and natural mineral compositions. From the plagioclase composition, we confirm the range of water contents and temperatures given above (Fig. 5). Natural plagioclase compositions (An60–40) were reproduced under water-undersaturated conditions (XH2Oin 0·7–0·8) with H2Omelt 4–5 wt % in the temperature range of 850–900°C (Fig. 5). However, the composition of plagioclase in the experiments in which the melt and plagioclase fractions were reproduced corresponds to a composition with a slightly lower anorthite content (An45–55) than the maximum observed in the cores of the natural plagioclase (An60) (Fig. 5). These An-rich plagioclase cores may have crystallized from a slightly more mafic melt (at higher temperature) and/or from a melt with a higher water content. The composition of most natural clinopyroxenes (CaO = 19–22 wt % and Mg-number = 65–77) was also reproduced in our experiments at a melt fraction of about 65 wt % (Fig. 7b) and temperatures of 850–900°C (assumed to be the pre-eruptive temperatures; see above); however, some natural clinopyroxene cores have higher CaO and MgO contents than the experimental clinopyroxenes. Similarly, most of the natural amphibole compositions were reproduced in experiments performed at relatively high oxygen fugacity (NNO + 0·8 to 1) at 400 MPa and 875–900°C (Mg-number = 55–66; Fig. 7a; Table 4). Some natural amphiboles have a lower SiO2/Al2O3 ratio (Al2O3 = 10·1–13·4 wt % and SiO2 = 41·5–44·3 wt %; Fig. 6) than the experimental amphibole, suggesting that they crystallized in equilibrium with a melt with a lower SiO2/Al2O3 ratio than the Huerto Andesite (Sisson & Grove, 1993; Pichavant et al., 2002). Both plagioclase and Fe–Mg minerals suggest that most of the mineral assemblage probably crystallized at 850–900°C in equilibrium with a melt with 4·5 wt % H2O; however, early crystallization may also have occurred from a more mafic, probably hotter (T >900°C), magma with a smaller amount of crystals. The occurrence of very S-rich apatites in the Huerto Andesite is another feature that might reveal an early crystallization stage in a hotter, more mafic S-rich magma than the andesite (basaltic magmas are often considered to be major sources of S and magmatic volatiles; e.g. Wallace, 2005).
The oxygen fugacity can be constrained from the composition of coexisting Fe–Ti oxides. The lower ulvöspinel and ilmenite contents (Xulvö = 0·20 and Xilm = 0·83, respectively) of the natural magnetite and ilmenite compared with the experimental magnetite and ilmenite from runs performed at 875–900°C and NNO + 1 (Xulvö = 0·27 and Xilm = 0·85, respectively; Table 6, Fig. 8) suggest more oxidized conditions than the experimental fO2 at 850–900°C, slightly higher than NNO + 1. This oxidation state is corroborated by the presence of anhydrite in the natural sample and in experimental charges at log fO2 > NNO + 0·9 (900°C) and probably by the high Mg-number of clinopyroxene, which may suggest a slightly more oxidized system in addition to crystallization from a more mafic melt.
The concentration of other volatiles such as sulphur in the andesitic melt can be estimated from the measurement of S in the experimental melts or from the experimentally determined partition coefficient of S between apatite and melt. In the experimental melts, there was a dramatic decrease in the sulphur content of the melt with decreasing XH2Oin. For XH2Oin decreasing from unity to 0·8 at 900°C, the sulphur content in the experimental melt decreases from 1172 to 181 ppm S (run 1 and run 2, respectively; Table 2). We can compare these values with the sulphur content in a melt in equilibrium with sulphur-rich apatite. Using the average composition of the natural apatite from the Huerto Andesite (average 0·6 wt % SO3, Parat et al., 2002) and the experimentally calibrated partition coefficient of sulphur between apatite and melt (Kd = 4; Parat & Holtz, 2005), we estimated the sulphur content in the melt at 560 ppm (S). This value agrees with the sulphur content in the experimental melts in equilibrium with apatite: 260 and 1200 ppm S for 0·51 and 0·70 wt % SO3 in apatite, respectively. The estimated value of 500 ppm S in the Huerto andesitic melt is at least two times higher than that estimated from melt inclusions in anhydrite–pyrrhotite-bearing dacite and trachyandesites [e.g. S = 88 ppm for Pinatubo (Westrich & Gerlach, 1992) and 200 ppm for El Chichón magmas (Luhr, 1990)] but close to the value determined for anhydrite–pyrrhotite-bearing andesitic magmas from Nevado del Ruiz (up to 700 ppm S; Sigurdsson et al., 1990).
The presence of bubbles in our experimental charges does not allow us to measure directly the carbon solubility in glasses using IR spectroscopy. To estimate the carbon content in the melts, a multicomponent saturation model for dacite can be used. By using the H2O–CO2 saturation model of Papale et al. (2006), at a saturation pressure of 400 MPa, a water content in the melt of 4·5 wt % and T = 900°C, the CO2 solubility in the melt is 2000 ppm. This agrees with the experimentally determined CO2 solubility in andesitic melts in equilibrium with C–O–H fluids (Botcharnikov et al., 2006).
Fluid phase composition in equilibrium with the melt
The fluid phase composition in equilibrium with minerals and melt may be estimated using the thermodynamic properties of fluids. In the C–H–O–S system, the following species may be considered: H2O, H2, CO2, CO, CH4, SO2, and H2S. The fugacities of H2O and CO2 were estimated using the solubility models of Zhang (1999) and Behrens et al. (2004), respectively. For H2Omelt = 4·5 wt % and CO2 melt = 2000 ppm at 900°C and 400 MPa, we estimated fH2O = 1500 bar and fCO2 = 6500 bar. According to the equilibrium constants given by Helgeson et al. (1978), fCO and fCH4 are very low (< 0·1 bar), and are not considered further for mole proportion calculations.
The sulphur fugacity (fS2) was calculated using the composition of experimental pyrrhotite (Toulmin & Barton, 1964; Froese & Gunter, 1976): log fS2 = –3·8 to + 1·3 (fS2 = 0·0001–20 bar) for T = 850–950°C. The sulphur fugacity is well correlated with the oxygen fugacity and the sulphur content of the melt (Fig. 11), except for two runs (3 and 22), which may have been compromised because of the analysis of quenched immiscible sulphide. The sulphur fugacity calculated from natural pyrrhotites ranges from 0·025 to 3 bar (log fS2 = –1·6 to + 0·5) (Parat et al., 2002), which corresponds to a sulphur content of 250–1200 ppm in our experimental melts coexisting with pyrrhotite in the temperature range 850–900°C. For the Huerto Andesite, we chose the average value of 0·45 bar (calculated using the natural pyrrhotite composition) as a representative S2 fugacity. This value corresponds to an intermediate value between experiment 2 (H2Omelt = 4·9 wt %, S = 181 ppm) and experiment 1 (H2Omelt = 7·1 wt %, S = 1172 ppm) at 900°C. Then we used the thermodynamic data of Helgeson et al. (1978) to determine the SO2 and H2S fugacities. We obtained log fSO2 = 1·14 (fSO2 = 14 bar) and log fH2S = 1·72 (fH2S = 55 bar).
The composition of the fluid phase at 900°C and 400 MPa was estimated using the model proposed by Churakov & Gottschalk (2003a, 2003b) taking the thermodynamic properties of fluid mixtures into account. The mole fractions of H2O, H2, CO2, SO2 and H2S in the fluid phase were adjusted until matching the fugacities estimated above. The data, given in Table 11, show that the estimated fugacities were approached for CO2-rich fluids. CO2 is the dominant species with XCO2 = 0·62, followed by H2O with XH2O = 0·37; the amount of sulphur species are minor: XH2S = 0·005 and XSO2 = 0·001. The higher CO2 fugacity calculated for a water fugacity close to 1500 bar might suggest a slightly higher CO2 content in the melt (>2000 ppm). The estimated mole proportions agree with mass-balance calculations for H2O and CO2 in experimental charges with the mineral assemblage and a mineral proportion close to that of the natural andesite. For pyroxene-bearing charge at 900°C (experiment 3), XH2Ofluid = 0·63 [=H2Ofluid/(H2Ofluid + CO2-fluid), where H2Ofluid = H2Oinitial – (H2Omelt x wt % of melt), mol] and CO2-fluid = 0·37 (=CO2-initial, mol, because of the low solubility of CO2 in melt).
Effect of volatiles on modal mineralogy and phase compositions
Crystallization experiments performed under moderately oxidized conditions (NNO + 1) on an andesitic bulk composition confirm that dissolved H2O in silicate melts significantly influences the crystallization sequence (e.g. amphibole in water-rich and pyroxene in water-poor andesitic melts), the abundance and composition of minerals, and therefore the composition of the residual liquid. Plagioclase and clinopyroxene are the phases most sensitive to the water content in the melt and have high calcium contents (high An and Wo components, respectively) in equilibrium with a water-rich melt.
Carbon may be relatively abundant in the fluid phase in equilibrium with hydrous andesitic melts and, assuming that the Huerto andesitic magma was fluid-saturated, XCO2 may represent 60% of the fluid phase. However, carbon species (CO2, CH4, and CO) are only slightly soluble in silicate melts (<4500 ppm CO2, Botcharnikov et al., 2006) and carbonate minerals are not stable in andesitic melts. Thus, CO2 has no direct effect on the phase composition and proportion, but does reduce the water activity.
Of the volatiles present in the fluid phase, sulphur is a minor component (<1%) and is only slightly soluble in andesitic melts (< 2000 ppm). This study shows that the addition of sulphur at moderately oxidizing conditions (NNO + 1) may cause: (1) a decrease in the abundance of plagioclase and magnetite and an increase in melt fraction that result in changes in the composition of the residual liquid and in the composition of minerals, especially plagioclase; (2) a shift in the stability field of plagioclase to lower temperatures in S-bearing ‘water-rich’ melts and the appearance of clinopyroxene and orthopyroxene at lower H2O contents in S-bearing melts than in S-free melts; (3) the crystallization of pyrrhotite over the entire studied temperature and fO2 range (850–950°C and NNO + 0·5 to NNO + 1·2), of anhydrite at 900–950°C for log fO2 > NNO + 0·9, and of ilmenite at T <900°C.
The low abundance of plagioclase and magnetite in the presence of sulphur may be due to either (1) a large reduction in the Ca and Fe content of the liquid caused by the presence of S and/or (2) the formation of stable Ca–S and/or Fe–S complexes in the melt, which inhibit the crystallization of plagioclase and magnetite. The systematic low proportions of plagioclase in S-bearing charges without and with stable anhydrite (<1·2 wt %) are correlated with a high CaO content in the melt and a high anorthite content of the plagioclase (the anorthite content of plagioclase correlates with the Ca/Na of the melt) and cannot result from anhydrite crystallization. The most probable explanation could be the formation of Ca–SO4 complexes in the melt that decrease the activity of Ca (e.g. Parat & Holtz, 2004). Pyrrhotite, magnetite and ilmenite are stable phases in S-bearing charges and their proportion is nearly constant over the range of experimental temperatures and water activities. The lower amount of magnetite in the S-bearing charges than in the S-free charges may be directly related to the crystallization of pyrrhotite (up to 3·3 wt %, representing 2 wt % of FeO removed from the melt), whereas the low and constant amount of ilmenite (<0·6 wt %) suggests that ilmenite saturation may be closely related to melt TiO2 content (Toplis & Carroll, 1995). The identical abundance of Fe-silicates in charges with and without sulphur and the absence of significant changes in Fe–Mg-silicate composition and FeO concentration in the melt with addition of sulphur suggest that the distribution of Fe between silicates and melt in the S-bearing experiments may result from mass-balance constraints and pyrrhotite–magnetite or pyrrhotite–magnetite–ilmenite exchange reactions (Whitney, 1984).
There are only a few experimental studies that provide information on the influence of sulphur on phase equilibria. Scaillet & Evans (1999) and Costa et al. (2004) investigated the phase relations of dacitic bulk-rock compositions (Pinatubo dacite and Tatara San Pedro dacite, respectively) in a CO2-free system (fluid phase composed of H–O–S-bearing species) and under oxidized conditions (>NNO + 1·2). They observed similar changes in melt and mineral proportion to those reported in this study. At log fO2 > NNO + 2·5, they reported the stabilization of biotite or/and gedrite instead of hornblende, a decrease in the plagioclase proportion and an increase in magnetite proportion. Scaillet & Evans (1999) suggested that the crystallization of ferromagnesian amphibole instead of calcic amphibole may be related to the sulphur content of the system and the resulting uptake of Ca by anhydrite, whereas Costa et al. (2004) favoured the formation of Ca–SO4 complexes in the melt. The absence of analytical methods to determine clearly the sulphur speciation in melts does not permit us to constrain quantitatively the role of S in S-bearing magmatic systems. Our data indicate a strong affinity of S with Ca in the melt structure and show that S is an important volatile to take into account for understanding the stability and composition of mineral phases as well as the liquid line of descent in magmatic systems.
|
CONCLUSIONS: IMPLICATIONS FOR THE REJUVENATION MODEL OF THE FISH CANYON MAGMA CHAMBER |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL AND ANALYTICAL...RESULTS: PHASE EQUILIBRIA AND...DISCUSSIONCONCLUSIONS: IMPLICATIONS FOR...REFERENCES |
|---|
Nearly all the major features of the hornblende-bearing rocks of the Huerto Andesite erupted in the San Juan volcanic field can be reproduced by our S-bearing phase equilibrium experiments. Phase diagrams and mineral compositions indicate that the magma may have been fluid-saturated with 4·5 wt % of water in the melt and that the pre-eruptive temperature was 875 ± 25°C (assuming a pressure of 400 MPa). Before eruption, the andesite had an oxidation state very close to, or slightly within, the co-stability field of anhydrite–pyrrhotite at NNO + 1·1.
The crystal content of the Huerto Andesite (around 45 vol. %) and the early crystallization of An-rich plagioclase together with Mg–Ca-rich clinopyroxene indicates that the temperature of the mafic parent magma at the liquidus may have been significantly higher than 875 ± 25°C, at around 950°C (Fig. 3). One main question is whether or not the Huerto Andesite was saturated with respect to fluids during crystallization and prior to eruption. The occurrence of very S-rich apatites in the Huerto Andesite is a feature that might reveal an early crystallization stage in a hotter, more mafic and S-rich magma than the andesite; however, the complex zoning in S concentration (Parat et al., 2002) may also indicate that fluid-saturated conditions prevailed during apatite crystallization (e.g. strong variations of fS2 being caused by episodic replenishment and degassing events at depth). As soon as fluid saturation was reached, H2O and CO2 may have been released associated with further crystallization. The solubility of CO2 in andesitic melts is low when compared with that of H2O (Botcharnikov et al., 2006) and CO2 is preferentially incorporated into the fluid phase. The distribution of CO2 and H2O between fluid and melt is such that the first fluids that are degassed should incorporate most of the CO2 (Dixon & Stolper, 1995). Thus, if the equilibrium fluid after 45% crystallization contains 60 mol % CO2, as suggested by our calculations, the amount of fluid released by cooling from 950–925°C to the pre-eruptive temperature (875 ± 25°C) must have been small and the initial water content of the andesitic melt prior to the onset of degassing must have been only slightly higher than 4·5 wt % H2O. Assuming this water concentration in the residual melt and considering that the crystal content is around 45 vol. % in the Huerto Andesite, our data indicate that the primary melts leading to the formation of magmas such as the Huerto Andesite probably contained 2·5 wt % H2O. This value can be used to estimate the possible amount of fluids that can be released by andesitic magmas crystallizing at depth. The release of such fluids is considered to have promoted the rejuvenation of silicic magmas in the upper crust, leading to the eruption of >5000 km3 of the dacitic Fish Canyon magma (Bachmann et al., 2002).
Knowing the water concentration of the andesitic melt, assuming a pre-eruptive temperature of 875 ± 25°C and assuming that the magma is saturated with respect to a fluid phase at this temperature, the available experimental data on andesitic systems (Botcharnikov et al., 2006) and the model of Papale et al. (2006) can be used to predict the CO2 concentration of the andesitic melts (see above). This concentration is in the range 2000–4000 ppm, and the molar fraction of H2O in the fluid [XH2O = H2O/(H2O + CO2)] after 100% degassing (or 100% crystallization, if the water incorporated into amphibole is neglected) would be >0·95. However, considering the gas sparging model of Bachman & Bergantz (2003), degassing with continuing crystallization at depth needs to be modelled by an open-system process, leading to the release of small amounts of CO2-rich fluids in the very first stages and high amounts of H2O-rich fluids during later crystallization stages (Dixon & Stolper, 1995). Most of the CO2 should be released after 5–10% crystallization. Thus, assuming that the Huerto Andesite is representative of the mafic magma that degassed at depth, the composition of the fluids that infiltrated the Fish Canyon batholith was water-rich (XH2O >0·95) at temperatures as high as 800–825°C (slightly below the pre-eruptive temperatures). Fluids exsolved at lower temperatures would be even more water-rich.
Assuming that the reactivation of the Fish Canyon magma chamber was initiated by the percolation of hot gas (gas sparging) released from a cooling 3000 km3 andesitic magma (Bachmann & Bergantz, 2003) similar to that of the Huerto Andesite (875°C, 4·5 wt % H2O; 45 vol. % crystals), it can be calculated that more than 250 km3 of water would be released as a result of the crystallization of the andesite (we assumed a density of 2500 kg/m3 and of 600 kg/m3 for the andesitic melt and the fluid, respectively). Bachmann & Bergantz (2003) calculated that the percolation of only 30 km3 of fluid (at 800°C) may be necessary to remobilize a crystal mush (with <70 vol. % crystals) that has reached its rheological locking point. Following the model of Bachmann and Bergantz (2003), the temperature of the 7500 km3 Fish Canyon magma may increase by 40°C (accounting for a cooling of the andesite magma over 150 kyr) as a result of gas sparging and heat transfer from the andesitic magma to the rhyolitic magma. Considering that the residual melts in the rhyolitic magma chamber contain mainly water (most CO2 dissolved in the rhyolitic melts may already have escaped in early crystallization stages) and that the rhyolitic magma was saturated with respect to quartz, alkali feldspar and Ab-rich plagioclase (Bachmann et al., 2002), such a temperature increase would result in an increase of the melt fraction by 10 wt % [based on data of Holtz et al. (2001) for the haplogranitic system]. This value is significantly lower than the dissolution of 20% estimated by Bachman & Bergantz (2003). As mentioned by Bachmann & Bergantz (2003), other mechanisms need to be taken into account to explain the reactivation of voluminous silicic mushes such as the Fish Canyon magma. Bachmann & Bergantz (2003) neglected the mass exchange between silicate melt and the H2O-rich fluid. However, the increase of the bulk water content of a magma in equilibrium with quartz and two feldspars is expected to cause a significant increase of the melt fraction (at isothermal conditions). If 75 km3 of water (30% of the total water) released from the crystallization of 3000 km3 of the Huerto Andesite reacts with the 7500 km3 Fish Canyon rhyolitic magma, the bulk water content of the magma would increase by 0·2 wt % H2O, which would in turn lead to an increase of the melt fraction by 5 wt % [at 760°C in the pressure range 200–300 MPa, applying the data obtained for the haplogranitic system; Holtz et al. (2001)]. In conclusion, one additional mechanism which may account in part for the reactivation of silicic mushes is a melting reaction caused by the supply of external fluids, composed mainly of water.
|
ACKNOWLEDGEMENTS |
|---|
We would like to thank M. Freise for experimental help and O. Diedrich for technical assistance during sample preparations. Discussions with R. Botcharnikov and constructive reviews by M. Streck, M. Pichavant and M. Rutherford are greatly appreciated and helped to improve the manuscript. We thank M. Wilson for editorial management. Research was supported by a Marie Curie Individual Fellowship (HPMF-CT-2001-01508) and a German Science Foundation project (DFG; Ho 1337/17).
|
FOOTNOTES |
|---|
**Present address: CODES-ARC - University of Tasmania, Hobart, TAS 7001, Australia
*Corresponding author. E-mail: Fleurice.Parat{at}minpet.uni-freiburg.de
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL AND ANALYTICAL...RESULTS: PHASE EQUILIBRIA AND...DISCUSSIONCONCLUSIONS: IMPLICATIONS FOR...REFERENCES |
|---|
Allen SR. Reconstruction of a major caldera-forming eruption from pyroclastic deposit characteristics: Kos Plateau Tuff, eastern Aegean Sea. Journal of Volcanology and Geothermal Research (2001) 105:141–162.[CrossRef][Web of Science]
Bachmann O, Dungan MA, Lipman PW. The Fish Canyon magma body, San Juan volcanic field, Colorado: Rejuvenation and eruption of an upper-crustal batholith. Journal of Petrology (2002) 43(8):1469–1503.
Bachmann O, Bergantz GW. Rejuvenation of the Fish Canyon magma body: a window into the evolution of large-volume silicic magma systems. Geology (2003) 31(9):789–792.
Bachmann O, Bergantz GW. On the origin of crystal-poor rhyolites: extracted from batholithic crystal mushes. Journal of Petrology (2004) 45(8):1565–1582.
Baker L, Rutherford MJ. Sulphur diffusion in rhyolite melts. Contributions to Mineralogy and Petrology (1996) 123:335–344.[CrossRef][Web of Science]
Behrens H, Ohlhorst S, Holtz F, Champenois M. CO2 solubility in dacitic melts equilibrated with H2O–CO2 fluids: implications for modeling the solubility of CO2 in silicic melts. Geochimica et Cosmochimica Acta (2004) 68(22):4687–4703.[CrossRef][Web of Science]
Berndt J, Liebske C, Holtz F, Freise M, Nowak M, Ziegenbein D, Hurkuck W, Koepke J. A combined rapid-quench and H2-membrane setup for internally heated pressure vessels: Description and application for water solubility in basaltic melts. American Mineralogist (2002) 87:1717–1726.
Botcharnikov R, Behrens H, Holtz F. Solubility and speciation of C–O–H fluids in andesitic melt at T = 1100–1300°C and P = 200 and 500 MPa. Chemical Geology (2006) 229:125–143.[CrossRef][Web of Science]
Botcharnikov R, Holtz F, Almeev R, Sato H, Behrens H. Storage conditions and evolution of andesitic magma prior to the 1991–95 eruption of Unzen volcano: Constraints from natural samples and phase equilibria experiments. Journal of Volcanology and Geothermal Research (2008) (in press).
Botcharnikov RE, Koepke J, Holtz F, McCammon C, Wilke M. The effect of water activity on the oxidation and structural state of Fe in a ferrobasaltic melt. Geochimica et Cosmochimica Acta (2005) 69(21):5071–5085.[CrossRef][Web of Science]
Carroll MR, Rutherford MJ. The stability of igneous anhydrite: experimental results and implications for sulphur behavior in the 1982 El Chichon trachyandesite and other evolved magmas. Journal of Petrology (1987) 28:781–801.
Carroll MR, Rutherford MJ. Sulphur speciation in hydrous experimental glasses of varying oxidation states: results from measured wavelength shifts of sulphur X-rays. American Mineralogist (1988) 73:845–849.[Abstract]
Churakov SV, Gottschalk M. Perturbation theory based equation state. I Pure fluid. Geochimica and Cosmochimica Acta (2003a) 67(13):2397–2414.[CrossRef]
Churakov SV, Gottschalk M. Perturbation theory based equation state. II. Fluid mixtures. Geochimica et Cosmochimica Acta (2003b) 67(13):2415–2425.[CrossRef][Web of Science]
Clemente B, Scaillet B, Pichavant M. The solubility of sulphur in hydrous rhyolitic melts. Journal of Petrology (2004) 45:2171–2196.
Costa F, Scaillet B, Pichavant M. Petrological and experimental constraints on the pre-eruption conditions of holocene dacite from the Volcán San Pedro (36°S, Chilean Andes) and the importance of sulphur in silicic subduction-related magmas. Journal of Petrology (2004) 45(4):855–881.
Couch S, Sparks RSJ, Carroll MR. Mineral disequilibrium in lavas explained by convective self-mixing in open magma chambers. Nature (2001) 411(6841):1037–1039.[CrossRef]
Devine JD, Gardner JE, Brack HP, Layne GD, Rutherford MJ. Comparison of microanalytical methods for estimating H2O contents of silicic volcanic glasses. American Mineralogist (1995) 80:319–328.[Abstract]
Dixon JE, Stolper E. An experimental study of water and carbon dioxide solubilities in mid-ocean ridge basaltic liquids. Part II: Applications to degassing. Journal of Petrology (1995) 36:1633–1646.
Ebadi A, Johannes W. Beginning of melting and composition of first melts in the system Qz–Ab–Or–H2O–CO2. Contributions to Mineralogy and Petrology (1991) 106:286–295.[CrossRef][Web of Science]
Froese E, Gunter AE. A note on the pyrrhotite–sulphur vapor equilibrium. Economic Geology (1976) 71:1589–1594.
Ghiorso MS, Sack RO. Fe–Ti oxide geothermometry: thermodynamic formulation and the estimation of intensive variables in silicic magmas. Contributions to Mineralogy and Petrology (1991) 108:485–510.[CrossRef][Web of Science]
Helgeson HC, Delany JM, Nesbitt HW, Bird DK. Summary and critique of the thermodynamic properties of minerals. American Journal of Science (1978) 278-A:1–229.
Holloway JR, Pan V, Gudmundsson G. High-pressure fluid-absent melting experiments in the presence of graphite; oxygen fugacity, ferric/ferrous ratio and dissolved CO2. European Journal of Mineralogy (1992) 4(1):105–114.
Holtz F, Pichavant M, Barbey P, Johannes W. Effects of H2O on liquidus phase-relations in the haplogranite system at 2 and 5 kbar. American Mineralogist (1992) 77(11–12):1223–1241.[Abstract]
Holtz F, Becker A, Freise M, Johannes W. The water-undersaturated and dry Qz–Ab–Or system revisited. Experimental results at very low water activities and geological implications. Contributions to Mineralogy and Petrology (2001) 141(3):347–357.[Web of Science]
Johnson MC, Rutherford MJ. Experimentally determined conditions in the Fish Canyon Tuff, Colorado, magma chamber. Journal of Petrology (1989) 30(3):711–737.
Koepke J. Analyse von wasserhaltigen silikatischen Glasern mit der Mikrosonde: Wassergehalte und Alkalienverluste. Berichte der Deutschen Mineralogischen Gesellschaft. Beiheft in European Journal of Mineralogy (1997) 9:200.
Leake BE, Woolley AR, Arps CES, et al. Nomenclature of amphiboles: report of the subcommittee on amphiboles of the International Mineralogical Association, commission on new minerals and mineral names. American Mineralogist (1997) 82:1019–1037.[Abstract]
Lindsley DH, Frost BR. Equilibria among Fe–Ti oxides, pyroxenes, olivine, and quartz: Part I. Theory. American Mineralogist (1992) 77:987–1003.[Abstract]
Lipman PW, Dungan MA, Bachmann O. Comagmatic granophyric granite in the Fish Canyon Tuff, Colorado: Implications for magma-chamber processes during a large ash-flow eruption. Geology (1997) 25(10):915–918.
Lipman PW. Geologic map of the Central San Juan caldera cluster, Southern Colorado. US Geological Survey, Geologic Investigations Series. (2006) I-2799.
Liu Y, Comodi P. Some aspects of the crystal-chemistry of apatites. Mineralogical Magazine (1993) 57:709–719.[Web of Science]
Luhr JF. Experimental phase relations of water- and sulfur-saturated arc magmas and the 1982 eruptions of El Chichón volcano. Journal of Petrology (1990) 31:1071–1114.
Martel C, Pichavant M, Holtz F, Scaillet B, Bourdier JL, Traineau H. Effects of f(O2) and H2O on andesite phase relations between 2 and 4 kbar. Journal of Geophysical Research—Solid Earth (1999) 104(B12):29453–29470.[CrossRef]
Matthews SJ, Moncrieff DHS, Carroll MR. Empirical calibration of the sulphur valence oxygen barometer from natural and experimental glasses; method and applications. Mineralogical Magazine (1999) 63(3):421–431.[Abstract]
Morimoto N, Fabries J, Ferguson AK, Ginzburg IV, Ross M, Seifert FA, Zussman J, Aoki K, Gottardi G. Nomenclature of pyroxenes. American Mineralogist (1988) 73(9–10):1123–1133.[Abstract]
Newman S, Lowenstern JB. VolatileCalc: a silicate melt–H2O–CO2 solution model written in Visual Basic for Excel. Computers and Geosciences (2002) 28(5):597–604.[CrossRef]
Pallister JS, Hoblitt RP, Reyes AG. A basalt trigger for the 1991 eruptions of Pinatubo volcano? Nature (1992) 356:426–428.[CrossRef]
Papale P, Moretti R, Barbato D. The compositional dependence of the saturation surface of H2O + CO2 fluids in silicate melts. Chemical Geology (2006) 229:78–95.[CrossRef][Web of Science]
Parat F, Dungan MA, Streck MJ. Anhydrite, pyrrhotite and sulphur-rich apatite: tracing the sulphur evolution of an Oligocene andesite (Eagle Mountain, Colorado, U.S.A.). Lithos (2002) 64:63–75.[CrossRef][Web of Science]
Parat F, Holtz F. Sulphur partitioning between apatite and melt and effect of sulphur on apatite solubility at oxidizing conditions. Contributions to Mineralogy and Petrology (2004) 147:201–212.[CrossRef][Web of Science]
Parat F, Dungan MA, Lipman PW. Contemporaneous trachyandesitic and calc-alkaline volcanism of the Huerto Andesite, San Juan Volcanic Field, Colorado, U.S.A. Journal of Petrology (2005) 46(5):859–891.
Parat F, Holtz F. Sulphur partition coefficient between apatite and rhyolite: the role of bulk S content. Contributions to Mineralogy and Petrology (2005) 150(6):643–651.[CrossRef][Web of Science]
Pichavant M, Martel C, Bourdier J.-L, Scaillet B. Physical conditions, structure, and dynamics of a zoned magma chamber: Mount Pelee (Martinique, Lesser Antilles Arc). Journal of Geophysical Research—Solid Earth (2002) 107(B5). 2093, doi:10.1029/2001JB000315.
Rouse RC, Dunn PJ. A contribution to the crystal chemistry of ellestadite and the silicate sulfate apatites. American Mineralogist (1982) 67(1–2):90–96.[Abstract]
Scaillet B, Evans BW. The 15 June 1991 eruption of Mount Pinatubo; I, Phase equilibria and pre-eruption P–T–fO2–fH2O conditions of the dacite magmas. Journal of Petrology (1999) 40(3):381–411.[CrossRef][Web of Science]
Scaillet B, Pichavant M. Experimental constraints on volatile abundances in arc magmas and their implications for degassing processes. In: Oppenheimer, C. Pyle, D. & Barclay, J. (eds). Volcanic Degassing. Geological Society, London, Special Publications (2003) 213:23–52.[CrossRef]
Scaillet B, Macdonald R. Experimental and thermodynamic constraints on the sulphur yield of peralkaline and metaluminous silicic flood eruptions. Journal of Petrology (2006) 47:1413–1437.
Schwab RG, Küstner D. The equilibrium fugacities of important oxygen buffers in technology and petrology. Neues Jahrbuch für Mineralogie (1981) 140:112–142.
Sigurdsson H, Carey S, Palais JM, Devine J. Pre-eruption compositional gradients and mixing of andesite and dacite erupted from Nevado del Ruiz volcano, Colombia in 1985. Journal of Volcanology and Geothermal Research (1990) 41:127–151.[CrossRef][Web of Science]
Sisson TW, Grove TL. Experimental investigations of the role of H2O in calc-alkaline differentiation and subduction zone magmatism. Contributions to Mineralogy and Petrology (1993) 113:143–166.[CrossRef][Web of Science]
Sisson TW, Bacon CR. Gas-driven filter pressing in magmas. Geology (1999) 27:613–616.
Stormer JC, Nicholls J. XLFRAC: A program for the interactive testing of magmatic differentiation models. Computers and Geosciences (1978) 4:143–159.[CrossRef]
Stormer JC. The effects of recalculation on estimates of temperature and oxygen fugacity from analyses of multicomponent iron–titanium oxides. American Mineralogist (1983) 68:586–594.[Abstract]
Toplis MJ, Carroll MR. An experimental study of the influence of oxygen fugacity on Fe–Ti oxide stability, phase relations, and mineral–melt equilibria in ferro-basaltic systems. Journal of Petrology (1995) 36(5):1137–1170.
Toulmin P, Barton PB. A thermodynamic study of pyrite and pyrrhotite. Geochimica et Cosmochimica Acta (1964) 28:641–671.[CrossRef][Web of Science]
Wallace PJ. Volatiles in subduction zone magmas: concentrations and fluxes based on melt inclusion and volcanic gas data. Journal of Volcanology and Geothermal Research (2005) 140:217–240.[CrossRef][Web of Science]
Webster JD, Holloway JR, Hervig RL. Phase equilibria of a Be, U and F-enriched vitrophyre from Spor Mountain, Utah. Geochimica et Cosmochimica Acta (1987) 51(3):389–402.[CrossRef][Web of Science]
Westrich HR, Gerlach TM. Magmatic gas source for the stratospheric SO2 cloud from the June 15, 1991, eruption of Mount Pinatubo. Geology (1992) 20(10):867–870.
Whitney JA. The effect of reduced H2O fugacity on the buffering of oxygen fugacity in hydrothermal experiments. American Mineralogist (1972) 57(11–12):1902–1908.[Web of Science]
Whitney JA. Fugacities of sulphurous gases in pyrrhotite-bearing silicic magmas. American Mineralogist (1984) 69(1–2):68–78.[Abstract]
Zhang Y. H2O in rhyolitic glasses and melts: measurement, speciation, solubility and diffusion. Reviews of Geophysics (1999) 37(4):493–516.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. D. Deering Cannibalization of an amphibole-rich andesitic progenitor induced by caldera-collapse during the Matahina eruption: Evidence from amphibole compositions American Mineralogist, August 1, 2009; 94(8-9): 1162 - 1174. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Blundy and K. Cashman Petrologic Reconstruction of Magmatic System Variables and Processes Reviews in Mineralogy and Geochemistry, January 1, 2008; 69(1): 179 - 239. [Full Text] [PDF]
|
|
|
|
|
|
O. Bachmann and G. W. Bergantz Deciphering Magma Chamber Dynamics from Styles of Compositional Zoning in Large Silicic Ash Flow Sheets Reviews in Mineralogy and Geochemistry, January 1, 2008; 69(1): 651 - 674. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||