Research ArticleGEOCHEMISTRY

The redox “filter” beneath magmatic orogens and the formation of continental crust

See allHide authors and affiliations

Science Advances  16 May 2018:
Vol. 4, no. 5, eaar4444
DOI: 10.1126/sciadv.aar4444
  • Fig. 1 Eu/Eu* in clinopyroxene and garnet in deep arc cumulates as a function of whole-rock Mg#.

    (A and B) Individual spot data measured by laser ablation ICP-MS. The error bars denoted in (A) and (B) are the long-term (~8 months) reproducibilities (2 SD) of measuring Eu/Eu* in basaltic glass standards BIR-1G and BCR-2G. (C) The mean Eu/Eu* in clinopyroxene and garnet for each rock sample. Error bars are 2 standard error of mean (2 SEM).

  • Fig. 2 Magma redox conditions calculated from garnet and clinopyroxene Eu anomalies.

    (A) Estimating fO2 of primitive magma. Garnet and clinopyroxene Eu/Eu*-fO2 plotted on the contours of Embedded Image. The probability distribution is for sample CC-ME1, the second most primitive garnet-bearing pyroxenite in this work. (B) Evolution of the bulk Eu partition coefficient relative to DEu* (bulk partition coefficient of Eu3+) during crystal fractionation, calculated using the observed Eu/Eu* of garnets and an incremental crystal fractionation model (see the Supplementary Materials and Methods). Dots with error bars (95% confidence interval) represent DEu/DEu* values estimated from Eu/Eu* of garnets in cumulates. Residual melt fraction estimated by converting whole-rock cumulate Mg# to the Mg# of a melt in equilibrium with the cumulate and inferring melt fraction from empirical relationships between melt Mg# and indices of melt fraction, such as K2O (see the Supplementary Materials and Methods). The uncertainties are from the measured garnet Eu/Eu* and cumulate-melt Mg# coupling and are propagated into the calculated DEu/DEu* and fO2 by a Monte Carlo resampling method (see the Supplementary Materials and Methods). Thick red line demarcates median values, and thin red line displays 95% confidence interval envelope. The total Eu partition coefficients in (B) were converted to fO2 in (C). Clinopyroxenes show the same Eu/Eu*-Mg# trend as garnet (Fig. 1), but we use garnet data to calculate the redox paths because the perfect incompatibility of Eu2+ in garnet simplifies Eq. 1 and reduces uncertainty.

  • Fig. 3 FeOT-MgO systematics in arc and mid ocean ridge (MOR) magmas.

    Data are presented as MgO binned (0.5 wt %) average and 2 SEM. Tholeiitic differentiation is represented by the mid-ocean ridge series (MOR). We divide arc igneous samples (mostly of Cenozoic ages) into three groups based on their modern arc crustal thickness: <25, 25 to 50, and >50 km. Arc magma compositions transition from tholeiitic to calc-alkaline as crust thickness increases. MOR data are from Keller et al. (58); arc data are from GeoRoc compilation and provided in data file S2.

  • Fig. 4 Garnet control on Fe depletion.

    (A) Percentage of total initial Fe content removed by various fractionating phases after 60% crystal fractionation. Results are from pMELTS simulation of hydrous basalt crystal fractionation (fractional crystallization) under 0.2 to 2.0 GPa, 4 wt % starting water content, and a constant fO2 of FMQ. (B) The extent of Fe depletion in arc magmas, as quantified by FeOT/MgO in moderately differentiated samples of 4 ± 1 wt % MgO, correlates with garnet fractionation index [Dy/Yb]N, where the subscript T denotes total Fe and N means C1 chondrite-normalized. Data are presented as FeOT/MgO binned (0.1) average values and 2 SEM.

  • Fig. 5 Modeled garnet effect on magma redox evolution with differentiation.

    (A) Fe depletion curves assuming 0 to 60% Fe removal by garnet fractionation. (B) Magma fO2 evolution curves as a function of the amount of Fe removed by garnet fractionation. Magma fO2 paths calculated from cumulate Eu/Eu* data (shaded area) are also shown for comparison. (C) Calculated Eu/Eu* in cumulate garnet based on the magma fO2 evolution curves in (B). Superimposed are the cumulate garnet Eu/Eu* data from this work.

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/4/5/eaar4444/DC1

    Supplementary Materials and Methods

    Supplementary Text

    fig. S1. Comparison between our measured Eu/Eu* in glass standards BCR-2G, BIR-1G, KL2-G, ML3B-G, and StHs6/80-G.

    fig. S2. Long-term (8 months) reproducibility in the analysis of glass standards BIR-1G and BCR-2G.

    fig. S3. Chondrite-normalized REE/FeO values for clinopyroxene.

    fig. S4. Chondrite-normalized REE/FeO values for garnet.

    fig. S5. Mean Eu/Eu* versus MgO in global arc lavas.

    fig. S6. Calculated logfO2 as a function of optical basicity assuming Eu/Eu* = 0.8 in garnet.

    fig. S7. Mineral Mg# versus whole-rock Mg# in the cumulates documented in this work.

    fig. S8. Melt-cumulate Mg# correlation simulated by pMELTS at 2 GPa, 4 wt % H2O, and various oxygen fugacities.

    fig. S9. Observed K2O-Mg# correlation in arc magmas, parameterized by an exponential function shown in the figure.

    fig. S10. Error propagation by Monte Carlo resampling.

    fig. S11. Sensitivity test of calculated DEu/DEu* and redox paths to the value of DEu* used in the crystal fractionation model.

    fig. S12. Calculated optical basicity (Λ) as a function of Mg# in arc magmas.

    data file S1. In-situ mineral composition data and cumulate whole-rock data.

    data file S2. Primitive arc magma compositions.

    data file S3. Global arc igneous rock compilation.

    References (5963)

  • Supplementary Materials

    This PDF file includes:

    • Supplementary Materials and Methods
    • Supplementary Text
    • fig. S1. Comparison between our measured Eu/Eu* in glass standards BCR-2G, BIR-1G, KL2-G, ML3B-G, and StHs6/80-G.
    • fig. S2. Long-term (8 months) reproducibility in the analysis of glass standards BIR-1G and BCR-2G.
    • fig. S3. Chondrite-normalized REE/FeO values for clinopyroxene.
    • fig. S4. Chondrite-normalized REE/FeO values for garnet.
    • fig. S5. Mean Eu/Eu* versus MgO in global arc lavas.
    • fig. S6. Calculated logfO2 as a function of optical basicity assuming Eu/Eu* = 0.8 in garnet.
    • fig. S7. Mineral Mg# versus whole-rock Mg# in the cumulates documented in this work.
    • fig. S8. Melt-cumulate Mg# correlation simulated by pMELTS at 2 GPa, 4 wt % H2O, and various oxygen fugacities.
    • fig. S9. Observed K2O-Mg# correlation in arc magmas, parameterized by an exponential function shown in the figure.
    • fig. S10. Error propagation by Monte Carlo resampling.
    • fig. S11. Sensitivity test of calculated DEu/DEu* and redox paths to the value of DEu* used in the crystal fractionation model.
    • fig. S12. Calculated optical basicity (Λ) as a function of Mg# in arc magmas.
    • References (60–64)

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    • data file S1 (Microsoft Excel format). In-situ mineral composition data and cumulate whole-rock data.
    • data file S2 (Microsoft Excel format). Primitive arc magma compositions.
    • data file S3 (Microsoft Excel format). Global arc igneous rock compilation.

    Files in this Data Supplement:

Navigate This Article