Granite geochemistry is not diagnostic of the role of water in the source

https://doi.org/10.1016/j.epsl.2021.116927Get rights and content

Highlights

  • Experiments, models and nanogranitoids are used concurrently and comparatively.

  • Water-fluxed melting does not produce a specific geochemical signature in melt.

  • Granite compositions are produced at low- to medium-P, whatever the fluid regime is.

  • The abundance of peraluminous granites does not imply a dry nature of orogenic crust.

Abstract

The diverse fluid regimes during melting of the metasedimentary crust have been often discriminated on the basis of the composition of anatectic granitoids, with granites indicating fluid-absent melting conditions and trondhjemitic compositions suggesting the addition of external water in the source region. The lack of abundant metasedimentary-derived trondhjemites in the geological record is supposed to prove the minor role of water-fluxed melting in the crustal maturation. In terms of trace elements, instead, Rb, Sr and Ba contents and their ratios have been commonly used to discriminate dehydration vs. water-fluxed melting scenarios. Here I show that reconciling results of melting experiments, thermodynamic modeling and nanogranitoid study brings out a different picture. Equilibrium thermodynamics cannot properly reproduce melt compositions of the selected benchmark experiments, with the latter having trondhjemitic compositions mainly for the metastable behavior of muscovite during laboratory runs. The formation of sufficient volumes of extractable trondhjemitic melts is related to high pressure melting conditions (≥8 kbar at 700 °C and ≥11 kbar at 750 °C) or K-poor bulk rock compositions, rather than to the only presence of water. At low- to medium-pressure, crustal melts are granites in composition, whatever the fluid regime is. It is inferred that the abundance of anatectic peraluminous granites (compared to metasedimentary-derived trondhjemites) does not imply a dry nature of the orogenic crust. Likewise, the use of LILE (Rb, Sr and Ba) signatures may lead to erroneous conclusions on the fluid regime of the deep continental crust.

Introduction

Since the time continental crust has existed, partial melting under variable pressure, temperature and fluid (PTXH2O) conditions has promoted its maturation (Sawyer et al., 2011). Fluid-absent melting has been the paradigm for more than 40 years (Brown, 2013; Thompson and Algor, 1977). However, regional-scale shear zones seem to be ideal loci for the infiltration of external water triggering melting (Carvalho et al., 2016; Sawyer, 2010; Weinberg and Hasalová, 2015a), and multiple and protracted water-fluxed melting events seem to occur in crustal roots of orogenic belts (Rubatto et al., 2009), of magmatic arcs (Collins et al., 2016, 2020) and in continental back-arcs (Wolfram et al., 2019). In particular, in the last years there has been renewed interest in trying to understand the real extent of water-fluxed melting and its role in crustal maturation (Collins et al., 2016; Weinberg and Hasalová, 2015a). This has added fuel to the fire, heating up the ongoing debate (Clemens and Stevens, 2015; Weinberg and Hasalová, 2015b), because water-fluxed melting may deeply impact on thermal structure, fertility and rheology of the orogenic crust (Weinberg and Hasalová, 2015a).

A major, as yet unresolved question is if the geochemical signature of water-fluxed melting is real or not. The results of experiments of Patiño Douce and Harris (1998) led to assume that water-fluxed melting of the metasedimentary crust produces trondhjemites, in contrast to peraluminous granites which, instead, are the products of fluid-absent reactions (e.g., Jiang and Zhu, 2017; Johnston et al., 2015; Nabelek, 2020; Wang et al., 2012; Yang et al., 2019; Zeng et al., 2005). On the other hand, the application of quantitative phase petrology (i.e. thermodynamic modeling or phase equilibria modeling) has raised doubts about the real impact of water-fluxed melting on melt composition, in particular at low-pressure (García-Arias et al., 2015; Schwindinger et al., 2019; Sola et al., 2017). However, the same tool has recently provided different results in the study of Mayne et al., 2020, supporting the inferences of Patiño Douce and Harris (1998). The fact that metasedimentary-derived trondhjemites are less abundant in the geological record compared to anatectic granites is largely used to minimize the role of water-fluxed melting, supporting the view of a dry continental crust (Clemens et al., 2020; Mayne et al., 2020) where water in crustal melts derives solely from the breakdown of hydrous minerals. In terms of trace elements, Harris and Inger (1992) and Inger and Harris (1993) proposed the use of LILE (Large Ion Lithophile Element; Rb, Sr and Ba) contents of anatectic granitoids to discriminate between dehydration and water-fluxed melting. Despite this approach has been largely adopted in last decades (e.g., Ferreira et al., 2020; Gao et al., 2017; Wang et al., 2012; Zeng et al., 2005), recent studies have questioned its reliability (Aikman et al., 2012; Schwindinger et al., 2019).

Here, I address these issues with a new approach that combines the three tools allowing the investigation of crustal melts in their source region (melting experiments, quantitative phase petrology and nanogranitoids). This study has three main objectives: (1) to evaluate the reliability of some experimental data; (2) to model the possible effect of pressure and bulk composition on the formation of melt compositions; and (3) to interrogate the current database of nanogranitoids (crystallized melt inclusions from deep crustal rocks) which provide us with the pristine composition of natural crustal melts produced under variable PTXH2O conditions (Bartoli et al., 2016; Carvalho et al., 2019; Cesare et al., 2015). These objectives are being synthesized in this manuscript to address the larger question on the use and misuse of geochemical records to decipher fluid regime during crustal melting. For the comparative study between experiments and phase equilibria modeling, I focus specifically on the benchmark work of Patiño Douce and Harris (1998), having led to a link in the literature between fluid regime and melt chemistry and representing, therefore, a landmark study for many crustal petrologists (see above). It will be shown that water-fluxed crustal melting does not produce anatectic melts characterized by a specific geochemical signature and that the nature of the fluid regime during melting processes cannot be recovered considering the geochemistry of granitoid rocks. In this paper I use water-fluxed melting for anatectic systems characterized by the ingress of external water, opposed to water-present melting specifically used for melting occurring in the proximity of the wet solidus, where only a internally-derived, free aqueous fluid phase is present. Water refers to a supercritical H2O phase.

Section snippets

Phase equilibria modeling

In this study the more recent re-parameterized activity–composition (ax) models for some minerals and melt presented by White et al. (2014a, 2014b) and the dataset of Holland and Powell (2011) were utilized. Phase equilibria modeling was performed in the ten-component MnO–Na2O–CaO–K2O–FeO–MgO–Al2O3–SiO2–H2O–TiO2 (MnNCKFMASHT) system, using Perple_X software (Connolly, 2009). Ferric iron was not considered because i) it was not quantified in the selected bulk rocks, and ii) magnetite and

Melting experiments vs. equilibrium thermodynamics

Fig. 1a-e show the calculated PT phase diagram sections obtained for the bulk compositions used by Patiño Douce and Harris (1998). Calculated phase relations are comparable with those of typical metapelites undergoing partial melting. The upper amphibolite-facies (≤750–775 °C) assemblages are generally characterized by the presence of one or two micas along with quartz, garnet, aluminosilicates, plagioclase and a Ti-bearing oxide. Under granulite-facies conditions, mineral assemblages commonly

The role of metastable muscovite

For the aim of this study, the most important compositional discrepancy to take into account is that shown by K2O (Fig. 2), which in turn affects the normative albite/orthoclase (Ab/Or) of melts (Fig. 3). Although electron microprobe analyses of hydrous felsic glasses may be affected by alkali migration (Morgan and London, 1996), this cannot be a significant cause of the observed discrepancy of K2O content (Fig. 2), because analytical strategies were adopted to minimize its effect (Patiño Douce

The role of pressure and bulk composition

The comparative study reported above demonstrates that equilibrium thermodynamic calculations cannot properly reproduce the trondhjemitic composition of experimental melts of Patiño Douce and Harris (1998), and that the presence or absence of muscovite seems to be the controlling factor in whether the melts are trondhjemitic or granitic. Models and experiments agree about the formation of trondhjemites solely at 700 °C and 10 kbar, (Fig. 3), when muscovite is present in both experiment and

Model predictions: why are they so different?

The results presented in this study are consistent with those of García-Arias et al. (2015), Sola et al. (2017) and Schwindinger et al. (2019), but differ significantly from those of Mayne et al. (2020). The latter would suggest that the fluid state of the system has a stronger control on melt composition with respect to the PT path and that melt compositions calculated under H2O in excess conditions have only few counterparts in the natural rock record.

However, very important differences do

Where does nature lie in all this?

Because experiments and models are a significant simplification of nature (Bartoli and Carvalho, 2021; White et al., 2011; Section 4), the constraints they provide need to be carefully tested against natural occurrences. Here the nanogranitoid database is interrogated for the first time to evaluate the potential impact of water-fluxed melting of fertile lithologies on melt composition. Nanogranitoid inclusions represent former droplets of natural anatectic melt trapped in peritectic minerals of

Implications

The results of this study question the common assumption that water-fluxed melting of the metasedimentary crust produces only trondhjemites. Rather, high-pressure conditions and protolith composition play an important role in the formation of these rocks and water-fluxed melting may produce peraluminous K2O-rich granitoids too (see also Sola et al., 2017 and Schwindinger et al., 2019). Similarly, it is not unambiguously possible to distinguish between fluid-present and fluid-absent melting from

CRediT authorship contribution statement

Omar Bartoli: conceptualization, methodology, visualization, writing.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was funded by University of Padua (grant BART_SID19_01 to O. Bartoli). I thank B. Cesare, S. Poli, G. Stevens, R.W. White and R.F. Weinberg for critically reading an earlier version. The manuscript benefitted from reviews by W.J. Collins and an anonymous reviewer.

References (71)

  • M. García-Arias et al.

    Water-present melting in the middle crust: the case of the Ollo de Sapo gneiss in the Iberian Massif (Spain)

    Chem. Geol.

    (2015)
  • S. Guillot et al.

    Geochemical constraints on the bimodal origin of High Himalayan leucogranites

    Lithos

    (1995)
  • S. Iaccarino et al.

    Pressure–temperature–time–deformation path of kyanite-bearing migmatitic paragneiss in the Kali Gandaki valley (Central Nepal): investigation of Late Eocene-Early Oligocene melting processes

    Lithos

    (2015)
  • P. Le Fort et al.

    Crustal generation of the Himalayan leucogranites

    Tectonophysics

    (1987)
  • E.W. Sawyer

    Migmatites formed by water-fluxed partial melting of a leucogranodiorite protolith: microstructures in the residual rocks and source of the fluid

    Lithos

    (2010)
  • C. Siégel et al.

    Use and abuse of zircon-based thermometers: a critical review and a recommended approach to identify antecrystic zircons

    Earth-Sci. Rev.

    (2018)
  • R. Weinberg et al.

    Water-fluxed melting of the continental crust: a review

    Lithos

    (2015)
  • R. Weinberg et al.

    Water-fluxed melting of the continental crust: a review

    Lithos

    (2015)
  • L. Yang et al.

    Is Himalayan leucogranite a product by in situ partial melting of the Greater Himalayan Crystalline? A comparative study of leucosome and leucogranite from Nyalam, southern Tibet

    Lithos

    (2019)
  • A. Acosta-Vigil et al.

    Mechanisms of crustal anatexis: a geochemical study of partially melted metapelitic enclaves and host dacite, SE Spain

    J. Petrol.

    (2010)
  • O. Bartoli et al.

    Recovering the composition of melt and the fluid regime at the onset of crustal anatexis and S-type granite formation

    Geology

    (2013)
  • O. Bartoli et al.

    Phase equilibria constraints on melting of stromatic migmatites from Ronda (S. Spain): insights on the formation of peritectic garnet

    J. Metamorph. Geol.

    (2013)
  • O. Bartoli et al.

    Granitoid magmas preserved as melt inclusions in high-grade metamorphic rocks

    Am. Mineral.

    (2016)
  • O. Bartoli et al.

    Geochemistry of Eocene-Early Oligocene low-temperature crustal melts from Greater Himalayan Sequence (Nepal): a nanogranitoid perspective

    Contrib. Mineral. Petrol.

    (2019)
  • A.J. Brearley et al.

    Effects of H2O on the disequilibrium breakdown of muscovite+quartz

    J. Petrol.

    (1990)
  • M. Brown

    Granite: from genesis to emplacement

    Bull. Geol. Soc. Am.

    (2013)
  • I.S. Buick et al.

    The role of water retention in the anatexis of metapelites in the Bushveld complex aureole, South Africa: an experimental study

    J. Petrol.

    (2004)
  • R. Carosi et al.

    Eocene partial melting recorded in peritectic garnets from kyanite-gneiss, Greater Himalayan Sequence central Nepal

    Geol. Soc. (Lond.) Spec. Publ.

    (2015)
  • B.B. Carvalho et al.

    Crustal reworking in a shear zone: transformation of metagranite to migmatite

    J. Metamorph. Geol.

    (2016)
  • B.B. Carvalho et al.

    Anatexis and fluid regime of the deep continental crust: new clues from melt and fluid inclusions in metapelitic migmatites from Ivrea Zone (NW Italy)

    J. Metamorph. Geol.

    (2019)
  • A. Castro et al.

    Origin of peraluminous granites and granodiorites, Iberian massif, Spain. An experimental test of granite petrogenesis

    Contrib. Mineral. Petrol.

    (1999)
  • J. Chen et al.

    Melting conditions in the modern Tibetan crust since the Miocene

    Nat. Commun.

    (2018)
  • W.J. Collins et al.

    Water-fluxed crustal melting produces Cordilleran batholiths

    Geology

    (2016)
  • W.J. Collins et al.

    Critical role of water in the formation of continental crust

    Nat. Geosci.

    (2020)
  • J.A.D. Connolly

    The geodynamic equation of state: what and how

    Geochem. Geophys. Geosyst.

    (2009)
  • Cited by (19)

    • Barium isotope evidence for the generation of peralkaline granites from a fluid-metasomatized crustal source

      2022, Chemical Geology
      Citation Excerpt :

      Rubidium, Sr, and Ba concentrations, oxygen isotopes, and liquid temperatures of granitic magma have been widely used to infer water-fluxed melting (Weinberg and Hasalová, 2015 and references therein; Volante et al., 2020). However, Sola et al. (2017) suggested that the main change caused by the presence of aqueous fluids during anatexis is an increase in melt fraction, with minor effects on melt elemental composition (Schwindinger et al., 2019; Bartoli, 2021). Because oxygen is a major element, oxygen isotopes are not sensitive to the addition of external fluids into the source regions (Weinberg and Hasalová, 2015), whereas Ba isotopes are.

    View all citing articles on Scopus
    View full text