Elsevier

Geochimica et Cosmochimica Acta

Volume 209, 15 July 2017, Pages 284-295
Geochimica et Cosmochimica Acta

Experimental calibration of a new oxybarometer for silicic magmas based on vanadium partitioning between magnetite and silicate melt

https://doi.org/10.1016/j.gca.2017.04.020Get rights and content

Abstract

Partition coefficients of vanadium between magnetite and rhyolitic silicate melt, DVmgt/melt, were experimentally determined as a function of oxygen fugacity (0.7–4.0 log units above the fayalite-magnetite-quartz buffer), temperature (800–1000 °C), melt alumina saturation index (ASI = 0.74–1.14), magnetite composition (0.2–14 wt% TiO2) and pressure (1–5 kbar; at H2O saturation). Experiments were performed by equilibrating small (≤20 µm), V-free magnetite grains in V-doped silicate melts (∼100 ppm V) and then analyzing both phases by LA-ICP-MS. Attainment of equilibrium was demonstrated by several reversal experiments. The results suggest that DVmgt/melt depends strongly on fO2, increasing by 1.5–1.7 log units from the MnO-Mn3O4 buffer to the Ni-NiO buffer, and to lesser (but still considerable) extents on melt alumina saturation index (ASI; increasing by 0.3–0.7 log units over 0.4 ASI units) and temperature (increasing by 0.3–0.7 log units over a 200 °C interval at a fixed fO2 buffer). Magnetite composition and melt water content seem to have negligible effects. The data were fitted by the following linear regression equation:

logDVmgt/melt=0.372610,000T+2.0465ASI-0.4773ΔFMQ-2.1214,

in which temperature is given in K, ASI refers to molar Al2O3/(CaO + Na2O + K2O) and ΔFMQ refers to the deviation of fO2 (in log units) from the fayalite-magnetite-quartz buffer. This equation reproduces all of our data within 0.3 log units, and 89% of them within 0.15 log units. The main advantages of this new oxybarometer over classical magnetite–ilmenite oxybarometry are (1) that it can be applied to rocks that do not contain ilmenite, and (2) that it is easier to apply to slowly-cooled rocks such as granites.

Introduction

Oxygen fugacity (fO2) is an important parameter in magmatic systems that affects the stability of mineral phases and fluid species (Lindsley and Frost, 1992, Keppler, 1993, Jugo et al., 2010). Furthermore, it affects mineral-melt and fluid-melt partition coefficients of many metals (Candela and Bouton, 1990, Taylor and Wall, 1992, Peiffert et al., 1994, Linnen et al., 1996, Jugo et al., 1999) and thus the mineralizing potential of intrusions (Ishihara, 1981, Lehmann, 1990, Blevin et al., 1996).

The most commonly applied method to constrain oxygen fugacity in intermediate to silicic rocks is Fe-Ti oxide thermobarometry (e.g. Buddington and Lindsley, 1964, Carmichael, 1967, Stormer, 1983, Andersen and Lindsley, 1988, Ghiorso and Sack, 1991, Lattard et al., 2005, Ghiorso and Evans, 2008). Other approaches are based on mineral reactions involving olivine, pyroxene, and/or sphene (Frost and Lindsley, 1992, Lindsley and Frost, 1992, Andersen et al., 1993, Xirouchakis et al., 2001), or biotite, K-feldspar and magnetite (Wones and Eugster, 1965, Wones, 1981). Most recently, oxygen fugacity has been estimated also from amphibole compositions (Ridolfi et al., 2009).

However, despite the various techniques listed above, reconstruction of magmatic fO2 in Si-rich igneous rocks remains a challenging task, particularly in the case of intrusive rocks. There are only few intrusive rocks that contain unaltered assemblages of the above mentioned minerals, because in most cases the minerals were either destroyed or reset at subsolidus conditions, such that fO2 estimation is either not possible anymore or leads to erroneous results. Furthermore, many samples contain only one Fe-Ti-oxide phase (magnetite or ilmenite), preventing application of the classical magnetite–ilmenite oxybarometry. The main goal of this study was to develop an oxybarometer that is based on phases which commonly occur as inclusions within quartz phenocrysts and thus were protected from subsolidus alteration. This oxybarometer can be applied to any silicic rock that contains magnetite, independent of whether or not the magma was saturated also in ilmenite.

For mineral–melt oxybarometry it is essential to focus on an element whose concentration varies as a function of oxygen fugacity in either the silicate melt or in a coexisting mineral phase, or to different extents in both. Further requirements are that the element of interest occurs in measurable amounts (i.e. above the detection limit of LA-ICP-MS measurements) in both phases, and that its concentration does not depend strongly on other factors such as mineral composition or melt composition. Previous studies have shown that vanadium partitioning between magnetite and silicate melt, DVmgt/melt, fulfills the above-mentioned requirements, and several calibrations have been developed (Irving, 1978, Horn et al., 1994, Canil, 1999, Canil, 2002, Toplis and Corgne, 2002, Righter et al., 2006a, Righter et al., 2006b, Mallmann and O'Neill, 2009). However, most of these studies focused on mafic to ultramafic systems at very high temperatures, which are not applicable to upper crustal rhyolitic magmas. The aim of the current study is to fill this gap by developing an experimental calibration of the vanadium magnetite–melt oxybarometer at P-T-x conditions that are relevant for silicic, upper crustal magmas. Such an oxybarometer would be useful not only for rocks lacking ilmenite, but also for slowly-cooled rocks such as granites because both silicate melt and to a lesser degree magnetite commonly occur as inclusions within quartz phenocrysts (e.g., Anderson et al., 2000, Audétat and Pettke, 2006, Audétat, 2015, Zhang and Audétat, 2017). If such inclusions are not intersected by later cracks and are analyzed as entities by laser-ablation ICP-MS (LA-ICP-MS), their original compositions – and thus DVmgt/melt partition coefficients and corresponding fO2 values – can be reconstructed.

Section snippets

Experimental methods

The following starting glasses were used in our experiments: (i) Synthetic haplogranite glasses with alumina saturation indices (ASI) of 0.7, 0.9 and 1.1, and (ii) natural obsidians from New Zealand, China and Armenia. The haplogranites were prepared from analytical grade SiO2, Al(OH)3, Na2CO3 and K2CO3. Their SiO2 content was fixed at the value corresponding to the 2 kbar haplogranite eutectic melt composition (Qz35Ab40Or25; Johannes and Holtz, 1996), and ASI was changed by varying Al2O3, Na2O

Analytical methods

The run products were analyzed by LA-ICP-MS using a system consisting of a GeolasPro 193 nm ArF Excimer Laser (Coherent, USA) attached to an Elan DRC-e (Perkin Elmer, Canada) quadrupole mass spectrometer. The ICP-MS was tuned to a ThO/Th rate of 0.05–0.1% and a Ca2+/Ca+ rate of 0.1–0.2% according to measurements on NIST SRM 610 glass (Jochum et al., 2011). The sample chamber was flushed with He gas at a rate of 0.4 l/min, to which 5 ml/min H2 gas was added on its way to the ICP-MS. The following

Results

Recovered samples consist of grains or clusters of magnetite embedded in optically transparent silicate glass (Fig. 2). The presence of large, isolated bubbles filled with aqueous fluid suggests that the experiments were saturated with an aqueous fluid phase, which is necessary to reach the fO2 imposed by the external buffer. Neutral pH was found in all samples, meaning that there was no CO2 (e.g. because of incomplete decarbonation of the glass), which could have changed the fO2 and the fluid

Discussion

In general, the DVmgt/melt values obtained in the present study are reproducible and vary regularly with temperature and ASI, suggesting that equilibrium was attained. Exceptions are three runs conducted at 800 °C at the Ni-NiO buffer (RA-V10c, RA-V15b,c and RA-V30b), which show an unusually large scatter (Fig. 4). In these experiments attainment of equilibrium appears to have been hindered by a combination of (1) low temperatures, (2) higher viscosity of the melt (e.g. Mysen and Toplis, 2007)

Conclusions and future perspectives

The experimental data presented in this study allowed identification of the main factors controlling vanadium partitioning between magnetite and melt in silicic magmas. DVmgt/melt is most strongly affected by fO2, changing by 1.5–1.7 log units between the Ni-NiO and MnO-Mn3O4 buffers. This result thus confirms earlier studies noting a strong dependence of vanadium partitioning on oxygen fugacity. Two other major parameters are temperature and melt composition (ASI), whereas magnetite

Acknowledgements

This research was supported by the German Science Foundation grant AU 314/5-1. We are thankful to Matteo Masotta, Hans Keppler, and Dan Frost for their useful comments and help in interpreting our data, and to Stefan Übelhack for machining tools for the experiments. Antony Burnham, John C. Ayers and two anonymous reviewers are thanked for their thorough and insightful reviews. Róbert Arató is also grateful to Miklós Arató for his help with the linear regression, as well as to Haihao Guo, Ananya

References (55)

  • P.J. Jugo et al.

    Sulfur K-edge XANES analysis of natural and synthetic basaltic glasses: implications for S speciation and S content as function of oxygen fugacity

    Geochim. Cosmochim. Acta

    (2010)
  • R.L. Linnen et al.

    The combined effects of fO2 and melt composition on SnO2 solubility and tin diffusivity in haplogranitic melts

    Geochim. Cosmochim. Acta

    (1996)
  • R.L. Nielsen et al.

    Major- and trace-element magnetite–melt equilibria

    Chem. Geol.

    (1994)
  • C. Peiffert et al.

    Uranium in granitic magmas: Part 1. Experimental determination of uranium solubility and fluid-melt partition coefficients in the uranium oxide-haplogranite-H2O-Na2CO3 system at 720–770 °C, 2 kbar

    Geochim. Cosmochim. Acta

    (1994)
  • K. Righter et al.

    Partitioning of Ni, Co and V between spinel-structured oxides and silicate melts: importance of spinel composition

    Chem. Geol.

    (2006)
  • M. Wilke et al.

    The oxidation state of iron in silicic melt at 500 MPa water pressure

    Chem. Geol.

    (2002)
  • Z. Zajacz et al.

    Determination of fluid/melt partition coefficients by LA-ICPMS analysis of co-existing fluid and silicate melt inclusions: controls on element partitioning

    Geochim. Cosmochim. Acta

    (2008)
  • D.J. Andersen et al.

    Internally consistent solution models for Fe-Mg-Mn-Ti oxides: Fe-Ti oxides

    Am. Miner.

    (1988)
  • A.T. Anderson et al.

    Evolution of Bishop Tuff rhyolitic magma based on melt and magnetite inclusions and zoned phenocrysts

    J. Petrol.

    (2000)
  • A. Audétat

    Compositional evolution and formation conditions of magmas and fluids related to porphyry Mo mineralization at Climax, Colorado

    J. Petrol.

    (2015)
  • A. Audétat et al.

    Evolution of a porphyry-Cu mineralized magma system at Santa Rita, New Mexico (USA)

    J. Petrol.

    (2006)
  • P.L. Blevin et al.

    Intrusive metallogenic provinces in eastern Australia based on granite source and composition

    Geol. Soc. Am. Spec. Pap.

    (1996)
  • A. Buddington et al.

    Iron-titanium oxide minerals and synthetic equivalents

    J. Petrol.

    (1964)
  • P.A. Candela et al.

    The influence of oxygen fugacity on tungsten and molybdenum partitioning between silicate melts and ilmenite

    Econ. Geol.

    (1990)
  • I.S.E. Carmichael

    The iron-titanium oxides of salic volcanic rocks and their associated ferromagnesian silicates

    Contrib. Mineral. Petrol.

    (1967)
  • M. Dickenson et al.

    The structural role and homogeneous redox equilibria of iron in peraluminous, metaluminous and peralkaline silicate melts

    Contrib. Mineral. Petrol.

    (1986)
  • B.R. Frost

    Introduction to oxygen fugacity and its petrologic importance

    Rev. Mineral. Geochem.

    (1991)
  • Cited by (20)

    • Magmatic-hydrothermal evolution of the El Laco iron deposit revealed by trace element geochemistry and high-resolution chemical mapping of magnetite assemblages

      2022, Geochimica et Cosmochimica Acta
      Citation Excerpt :

      The Ga vs. V data of magnetite in Fig. 9c is consistent with previous observations that Ga and V concentrations are typically higher in igneous than in hydrothermal magnetite (Nadoll et al., 2014; Salazar et al., 2020; Palma et al., 2020). Experimental and empirical evidence indicate that redox-sensitive elements such as V are preferentially incorporated into magnetite at high temperature and more reducing conditions (Toplis and Corgne, 2002; Nadoll et al., 2014; Arató and Audétat, 2017; Sievwright et al., 2017). The partitioning behavior of Ga between magnetite and silicate melt, on the other hand, is relatively insensitive to oxygen fugacity changes (D(Ga) 0.93–1.36; Sievwright et al., 2017); therefore, high concentrations of Ga in igneous magnetite suggests a temperature control (Fig. 6; 9c), and consistent with recent empirical (Liu et al., 2015; He et al., 2016; Polivchuk, 2017) and experimental studies (D(Ga) > 2.49; Horn et al., 1994).

    • Geothermometry using minor and trace elements in igneous and hydrothermal magnetite

      2020, Chemical Geology
      Citation Excerpt :

      A subset of 19 experiments from the Evans et al. (2006) study on a synthetic Mn-rich simple system (MnMgAlTiFeO) were removed from any consideration for Mn in Mag, as they contain this element at order of magnitude concentrations above those observed in the other natural rocks. A majority of the data derive from the compilation of Arato and Audetat (2017a, 2017b, 2017c), who applied the Mg/Mn criterion of Bacon and Hirschmann (1988) in experiments that contained Ilm to determine if Mag represented equilibrium values. We added other more recent data.

    • Generation of high-Mg diorites and associated iron mineralization within an intracontinental setting: Insights from ore-barren and ore-bearing intrusions in the eastern North China Craton

      2019, Gondwana Research
      Citation Excerpt :

      Because oxide mineralogy and partitioning data among mineral phases can provide a more detailed assessment of magma oxidation state than do by whole-rock chemical analyses (Frost, 1991; Meinert et al., 2005), we use the elements sensitive to oxygen fugacity, such as V and Ti in magnetite, to further constrain the oxidation states of the host magmas. Vanadium has a range of possible oxidation states (such as V3+, V4+, V5+, Toplis and Corgne, 2002; Bordage et al., 2011), the partitioning of which between magnetite and melt strongly depends on oxygen fugacity, and also on melt alumina saturation index (ASI) as well as temperature (T) to some extent (Arató and Audétat, 2017). According to the method of Arató and Audétat (2017) (logDVmgt/melt = 0.3726 × 10,000 / T(K) + 2.0465 × ASI-0.4773 × ∆FMQ-2.1214), the calculated ∆FMQ in the Zibo high-Mg diorites increases slightly from the most mafic ore-barren gabbroic diorite (2.91–3.33, averaging at 3.14) to the less mafic moderately mineralized monzodiorite (3.24–3.41, averaging at 3.35), but decreases obviously to the most felsic strongly mineralized quartz monzonite (2.70–3.0, averaging at 2.86).

    • Discovery of modern (post-1850 CE) lavas in south-central British Columbia, Canada: Origin from coal fires or intraplate volcanism?

      2018, Lithos
      Citation Excerpt :

      The V contents in till from British Columbia show a positive correlation with Fe, and negative correlation with Si (Canil and Lacourse, 2011), suggesting magnetite is the primary host for V. If fusion occurred at high fO2, near or above the HM buffer as shown at Tranquille, then magnetite, having DNbmt/liq is > 1 and DVmt/liq < 1 at these conditions (Arató and Audétat, 2017; Nielsen and Beard, 2000) would sequester Nb, and release more V to the melt, compared to the till precursor. This would produce the observed Nb depletion and V enrichment in the lava relative to till (Fig. 6).

    View all citing articles on Scopus
    View full text