The beginning of melting of fertile and depleted peridotite at 1.5 GPa

https://doi.org/10.1016/S0012-821X(97)00162-3Get rights and content

Abstract

We have determined the solidus temperatures and liquid compositions for batch melts of fertile and depleted lherzolite in the 0–15% melting interval at 1.5 GPa. Because establishment of equilibrium is difficult at very low melt fractions we have used an iterative technique in which the liquids from sandwich-type experiments are synthesized and tested for multiple saturation at the same pressure, temperature conditions as those of the sandwich experiment. Only when all 4 mantle phases are stable and of correct composition is the result accepted as being at equilibrium. This technique permits accurate determination of solidus temperatures to ±10°C and provides the compositions of low-degree mantle melts. We find that at degrees of batch melting above 3%, liquids produced from a fertile peridotite (MORB-Pyrolite) contain ∼49% SiO2, have <4% Na2O and have Fe–Mg partitioning relationships with the solid phases which are typical of basalt. As the degree of melting drops to 0% in fertile peridotite (MORB-Pyrolite), SiO2 content of the melt increases to 53% and the Na2O content to 8%. The melts remain strongly nepheline-normative and Fe favours the melt relative to olivine even more strongly than normal. We ascribe this latter observation to Na–Fe3+ coupling in the melt at high Na2O content. These high-alkali and high-silica melts are not observed at the solidus of depleted peridotite (Tinaquillo Lherzolite). These experimental results provide the first direct tests of fractional melting models which invoke the successive extraction of low-degree partial melts. No melting model currently available provides a good description of the compositions of low-degree melts from a fertile peridotite at 1.5 GPa.

Introduction

During near-adiabatic upwelling, the suboceanic upper mantle partially melts by between 5% and 20% to produce MORB [1][2][3]. Given that silicate partial melts become interconnected along grain boundaries when they are present at ∼1% by volume [4] it now seems likely that this MORB-generation process more closely approximates perfect fractional melting than batch melting [1][5]. This is because the melts, once interconnected, should begin separating from the solid residue [5]. The compositions of such fractional melts must contain evidence of the conditions of pressure and temperature of generation and there is some indication that such information is available in MORB. Klein and Langmuir [2] demonstrated, for example, a correlation between MORB composition and depth of the ridge axis. They interpreted the relationships, semiquantitatively, in terms of the extent of (hypothetical) batch melting of the mantle and by the presence of clinopyroxene as a residual phase through the melting interval.

McKenzie and Bickle [3] made the first attempt to quantify the dependence of the extent of melting and melt composition on temperature using an empirical parameterization of experimental data at high degrees (>15%) of batch melting (e.g., [6]). Their quantitative approach can, in principle, be extended to the fractional melting process, but the current parameterization suffers from a lack of experimental data at low degrees of melting. Thus, although the model generates compositions which approximate MORB, there are significant errors, particularly in Si, Al, Mg and Ca contents. A similar observation applies to the melting model of Niu and Batiza [7] who estimated apparent bulk distribution coefficients for major elements during equilibrium partial melting of peridotite. Their model cannot, as will be demonstrated below, be used to calculate the compositions of low-degree partial melts of mantle peridotite. Experiments need to be performed on liquid and solid compositions appropriate to 0–5% melting in order to construct a valid model for fractional melting in which melt separates when present at ∼1% by volume.

In view of the need for data at low melt fractions, the aim of this study has been to determine solidus temperatures and low-degree melt compositions for fertile peridotite (close to MPY-90 as described in Falloon and Green, [8], see their table 1) and depleted peridotite (Tinaquillo Lherzolite, [6]) at 1.5 GPa. Of particular importance has been the use of several different methods to demonstrate equilibrium. As will be shown, the results conflict with most earlier melt models, particularly under the near-solidus conditions which are of greatest importance to fractional melting. The methodology and data presented here provide the means to treat polybaric fractional melting processes in a quantitative manner.

Section snippets

Experimental approaches to low-degree melting

The need for data at low degrees of partial melting has prompted several recent studies and the development of a new technique, the so-called “diamond-aggregate” method [9][10][11]. Kinzler and Grove [12] quantified melting relations using the traditional method of multiple saturation. In their experiments, small amounts of natural olivine and orthopyroxene were added to possible primary melt compositions and experiments performed at some temperature just below the liquidus. They then observed

Experimental and analytical techniques

All starting compositions (Table 1) were prepared by grinding together analytical grade oxides and carbonates under acetone. The mixes were decarbonated at 800°C in air and subsequently reduced in a CO/CO2 gas mixing furnace for 24 h at 1000°C and an oxygen fugacity (fO2) equivalent to two log units below the fayalite–magnetite–quartz buffer (FMQ−2). This treatment produces a fine-grained reactive mixture of olivine, pyroxene and incompletely reacted oxides.

The Fe3+/Fe2+ ratios of a selection

Approach to equilibrium

If the composition of the basalt part of the sandwich is far from equilibrium one of two things tends to happen. In the case of too high an SiO2 content, the melt reacts with olivine in the peridotite to produce an orthopyroxene-rich layer at the junction between melt and residual solids. If the basalt has insufficient SiO2, reaction between melt and pyroxene produces an olivine-enriched layer at the boundary between the two parts of the charge.

All experiments were run for ≥48 h or greater to

Degree of melting (F) and solidus temperature

Since the amount of melt in our sandwich experiments is always greater than that which would be produced from the bulk peridotite, it is not possible to point count the proportions of phases in order to derive the fraction of melt. It is, however, possible to estimate the fraction of melt (F) which would be produced from the appropriate peridotite under the conditions of the experiment. The approach is only accurate if the melt composition does not change greatly during the experiment and for

Liquid compositions

Fig. 3 shows the melt compositions recalculated as molecular normative olivine (Mg,Fe)2SiO4, diopside Ca(Mg,Fe)Si2O6, quartz (SiO2) and jadeite plus calcium tschermakite (NaAlSi2O6+CaAl2SiO6) and projected away from diopside onto the base formed by the other 3 components [33]. It can clearly be seen that the lowest-temperature liquids plot towards the Jd+CaTs apex and that these liquids are Ne-normative for both MPY and Tinaquillo lherzolite bulk compositions. The data lie on an extension of

Discussion

There are many chemical and experimental data to support the hypothesis that MORB are polybaric, near-fractional melts of mantle peridotite. Trace-element concentrations in melt inclusions and in residual clinopyroxenes [1][15][38], isotopic ratios in MORB themselves [39][40][41], and experiments on melt/matrix rheology and melt transport properties [4][42][43][44] are all consistent with this analysis. A full description of the fractional melting process requires the compositions of very

Conclusions

We have developed a range of experimental techniques which enable the compositions of low-degree partial melts of peridotite to be precisely determined. It has been shown that the sandwich technique, with relatively large amounts of liquid, can reproduce the equilibrium compositions of low-degree partial melts provided the composition of the melt “pool” is adjusted to be very close to equilibrium at the beginning of the experiment. Product glasses from such experiments at 1.5 GPa have been

Supplementary data

Acknowledgements

The authors would like to thank Trevor Falloon for discussions concerning mantle melting. This manuscript benefitted greatly from a review by Ro Kinzler and an extremely thorough anonymous review. This research was supported by NERC grant GR3/08730. JDB is grateful to The Royal Society for research fellowship #516002. [FA]

References (45)

  • A.L. Jaques et al.

    Anhydrous melting of peridotite at 0–15 kb pressure and the genesis of tholeiitic basalts

    Contrib. Mineral. Petrol.

    (1980)
  • Y. Niu et al.

    An empirical method for calculating melt compositions produced beneath mid-ocean ridges: Application for axis and off-axis (seamount) melting

    J. Geophys. Res.

    (1991)
  • T.J. Falloon et al.

    Anhydrous partial melting of MORB pyrolite and other peridotite compositions at 10 kb: Implications for the origin of primitive MORB glasses

    Contrib. Mineral. Petrol.

    (1987)
  • K.T.M. Johnson et al.

    Segregation of high pressure partial melts from peridotite using aggregates of diamond: A new experimental approach

    Geophys. Res. Lett.

    (1992)
  • R.J. Kinzler et al.

    Primary magmas of mid-ocean ridge basalts, 1. Experiments and methods

    J. Geophys. Res.

    (1992)
  • R.J. Kinzler et al.

    Primary magmas of mid-ocean ridge basalts, 2. Applications

    J. Geophys. Res.

    (1992)
  • L.T. Bryndzia et al.

    Oxygen thermobarometry of abyssal spinel peridotites — The redox state and C–O–H volatile composition of the earth's sub-oceanic upper mantle

    Am. J. Sci.

    (1990)
  • K.T.M. Johnson et al.

    Melting in the oceanic upper mantle: An ion microprobe study of diopsides in abyssal peridotites

    J. Geophys. Res.

    (1990)
  • B.O. Mysen et al.

    Compositional variations of coexisting phases with degree of melting in the upper mantle

    Am. Mineral.

    (1977)
  • T.J. Falloon et al.

    Quest for low-degree partial melts

    Nature (London)

    (1996)
  • M.B. Baker et al.

    Compositions of near-solidus peridotite melts from experiments and thermodynamic calculations

    Nature (London)

    (1995)
  • E.M. Stolper

    A phase diagram for mid-ocean ridge basalts: Preliminary results and implications for petrogenesis

    Contrib. Mineral. Petrol.

    (1980)
  • Cited by (236)

    • The magnesium isotopic composition of the mantle

      2023, Geochimica et Cosmochimica Acta
    • Origin and tectonic implications of boninite dikes in the Shiquanhe ophiolite, western Bangong Suture, Tibet

      2021, Journal of Asian Earth Sciences
      Citation Excerpt :

      The mineral mode and the partition coefficients are from McDade et al. (2003), using the assumed temperature and pressure condition of 1315 degrees centigrade and 1.5 GPa. The partial melting reaction is from Robinson et al. (1998), using the assumed melting degree of 20–35%. The best-fit results indicate that the harzburgite melts at 20–35%, and then the 87.5% melt mixes with 12.5% molten GLOSS are within the field of the Shiquanhe boninite dike samples (Fig. 7B, Table S3).

    View all citing articles on Scopus
    View full text