by | Sep 20, 2022 | mainpost, vol23

A partial melting control on the Zn isotope composition of basalts

J.M.D. Day1,2,

1Scripps Institution of Oceanography, San Diego, La Jolla, CA 92093-0244, USA
2Université Paris Cité, Institut de Physique du Globe de Paris, CNRS, 1 rue Jussieu, 75238 Paris, France

F. Moynier2,

2Université Paris Cité, Institut de Physique du Globe de Paris, CNRS, 1 rue Jussieu, 75238 Paris, France

O. Ishizuka3

3Institute of Earthquake and Volcano Geology, Geological Survey of Japan, AIST, Central 7, 1-1-1, Higashi, Tsukuba, Ibaraki 305-8567, Japan

Affiliations | Corresponding Author | Cite as | Funding information

Geochemical Perspectives Letters v23 | https://doi.org/10.7185/geochemlet.2230
Received 23 May 2022 | Accepted 23 August 2022 | Published 20 September 2022

Copyright © 2022 The Authors

Published by the European Association of Geochemistry
under Creative Commons License CC BY-NC-ND 4.0

Keywords: zinc isotopes, fractionation, partial melting, boninites, MORB, basalt


top

Abstract

Abstract | Introduction | Methods and Results | Discussion | Acknowledgements | References | Supplementary Information

Basaltic partial melts are produced in a range of tectonic settings, including fluid-assisted melting above subduction zones, decompression melting at ridges and thermally driven melting above mantle plumes. To examine the role of partial melting on Zn, isotope and abundance data are reported for modern large-degree partial melts of the mantle represented by 22 mid-ocean ridge basalts (MORB) from three ocean basins and the first data for boninites. Boninites have some of the lowest Zn abundances of all terrestrial basalts and Zn isotope compositions (δ66Zn = +0.21 ± 0.06 ‰), generally lighter than for MORB (δ66Zn = +0.28 ± 0.06 ‰). Accounting for partial melting, komatiites, boninites and MORB derive from mantle sources with δ66Zn of ∼0.16 ± 0.06 ‰. Lower-degree partial melts, such as alkali basalts, can have higher δ66Zn, with up to ∼0.4 ‰ variation possible from partial melting of distinct peridotite mantle sources. Partial melting of fertile lherzolitic and depleted harzburgitic mantle sources can generate significant Zn isotope variability and should be evaluated prior to ascribing crustal, enriched or lithological components to mantle reservoirs from Zn compositions of planetary basalts.

Figures and Tables

Figure 1 Bulk rock (a) Zn abundance and (b) isotopic composition as a function of MgO content for basaltic rocks. Data sources are given in Figure S-1, with Baffin Island bulk rock picrite data and the magmatic olivine compositions (assuming a possible range of olivine MgO contents) from McCoy-West et al. (2018). Martian shergottite and nakhlite data (shown as points and as a field) are from Paniello et al. (2012). New MORB data presented here (solid circles) are distinguished from published data (solid stars).

Figure 2 (a) Zinc isotopic composition and (b) Zn abundance versus estimated degree of partial melting for basaltic rocks (Table 2) versus models for partial melting of harzburgite, lherzolite and metasomatised peridotite, with dots conforming to 1 % increments of melting. Shown in upper panel are the estimated bulk silicate Earth (BSE) average (solid line) and standard deviation (shaded regions) values from Chen et al. (2013) and Sossi et al. (2018).

Table 1 Zinc isotope and abundance data for boninites and mid-ocean ridge basalts. Major element data and ages for boninites are from Ishizuka et al. (2011, 2014) and for MORB are from Le Roux (2000) and Deng et al. (2018). *Replicate analyses are reported for these samples.

Table 2 Zinc isotope, abundance data, melting type and extent for terrestrial basalts.

Figure 1 Figure 2 Table 1 Table 2

View all figures and tables





top

Introduction

Abstract | Introduction | Methods and Results | Discussion | Acknowledgements | References | Supplementary Information


Zinc stable isotopes have utility in the study of basaltic rocks, from the examination of contributions of potential mantle and crustal reservoirs with distinct 66Zn/64Zn ratios (e.g., Beunon et al., 2020

Beunon, H., Mattielli, N., Doucet, L.S., Moine, B., Debret, B. (2020) Mantle heterogeneity through Zn systematics in oceanic basalts: Evidence for a deep carbon cycling. Earth-Science Reviews 205, 103174. https://doi.org/10.1016/j.earscirev.2020.103174

), to identification of volatile element loss in planets (e.g., Paniello et al., 2012

Paniello, R.C., Day, J.M.D., Moynier, F. (2012) Zinc isotopic evidence for the origin of the Moon. Nature 490, 376–379. https://doi.org/10.1038/nature11507

). Zinc consists of five stable isotopes (64Zn [natural abundance 48.6], 66Zn [27.9], 67Zn [4.1], 68Zn [18.1], and 70Zn [0.6]) which are typically reported in per mil variations relative to a standard (δxZn = [((xZn/64Zn)sample/(xZn/64Zn)JMC-Lyon standard − 1) × 1000], where x typically refers to mass 66 or 68). Studies of basaltic magmatic differentiation series have demonstrated that Zn isotopes are not substantially modified (<0.1 ‰) during fractional crystallisation processes (Chen et al., 2013

Chen, H., Savage, P.S., Teng, F.-Z., Helz, R.T., Moynier, F. (2013) Zinc isotope fractionation during magmatic differentiation and the isotopic composition of the bulk Earth. Earth and Planetary Science Letters 369–370, 32–42. https://doi.org/10.1016/j.epsl.2013.02.037

), but that significant fractionation (≥0.1 ‰) in δ66Zn is likely to occur during mantle partial melting (Wang et al., 2017

Wang, Z.-Z., Liu, S.-A., Liu, J., Huang, J., Xiao, Y., Chu, Z.-Y., Zhao, X.-M., Tang, L. (2017) Zinc isotope fractionation during mantle melting and constraints on the Zn isotope composition of Earth’s upper mantle. Geochimica et Cosmochimica Acta 198, 151–167. https://doi.org/10.1016/j.gca.2016.11.014

). Partial melting processes can explain δ66Zn variations in arc lavas (Huang et al., 2018

Huang, J., Zhang, X.-C., Chen, S., Tang, L., Wörner, G., Yu, H., Huang, F. (2018) Zinc isotopic systematics of Kamchatka-Aleutian arc magmas controlled by mantle melting. Geochimica et Cosmochimica Acta 238, 85–101. https://doi.org/10.1016/j.gca.2018.07.012

), whereas it has been argued that δ66Zn variations in some mantle-derived basaltic rocks reflect contributions from distinct sources, including recycled carbonate (Beunon et al., 2020

Beunon, H., Mattielli, N., Doucet, L.S., Moine, B., Debret, B. (2020) Mantle heterogeneity through Zn systematics in oceanic basalts: Evidence for a deep carbon cycling. Earth-Science Reviews 205, 103174. https://doi.org/10.1016/j.earscirev.2020.103174

; Liu et al., 2022

Liu, S.A., Qu, Y.R., Wang, Z.Z., Li, M.L., Yang, C., Li, S.G. (2022) The fate of subducting carbon tracked by Mg and Zn isotopes: A review and new perspectives. Earth-Science Reviews 228, 104010. https://doi.org/10.1016/j.earscirev.2022.104010

).

Intraplate volcanic rocks, including ocean island basalts (OIB), were first used to estimate a bulk silicate Earth (BSE) Zn isotopic composition (δ66ZnBSE = +0.28 ± 0.05 ‰; all uncertainties 2 s.d.; Chen et al., 2013

Chen, H., Savage, P.S., Teng, F.-Z., Helz, R.T., Moynier, F. (2013) Zinc isotope fractionation during magmatic differentiation and the isotopic composition of the bulk Earth. Earth and Planetary Science Letters 369–370, 32–42. https://doi.org/10.1016/j.epsl.2013.02.037

), followed by direct analysis of fertile peridotites (+0.30 ± 0.03 ‰; Doucet et al., 2016

Doucet, L.S., Mattielli, N., Ionov, D.A., Debouge, W., Golovin, A.V. (2016) Zn isotopic heterogeneity in the mantle: A melting control? Earth and Planetary Science Letters 451, 232–240. https://doi.org/10.1016/j.epsl.2016.06.040

). Subsequently, peridotites and komatiites have been shown to have indistinguishable Zn isotopic compositions (+0.16 ± 0.06 ‰) and so were interpreted to represent a consistent δ66ZnBSE (Sossi et al., 2018

Sossi, P.A., Nebel, O., O’Neill, H.S.C., Moynier, F. (2018) Zinc isotope composition of the Earth and its behaviour during planetary accretion. Chemical Geology 477, 73–84. https://doi.org/10.1016/j.chemgeo.2017.12.006

), with slightly more elevated estimates also reported from modern and ancient mantle melts (+0.20 ±0.03 ‰; McCoy-West et al., 2018

McCoy-West, A.J., Fitton, J.G., Pons, M.L., Inglis, E.C., Williams, H.M. (2018) The Fe and Zn isotope composition of deep mantle source regions: Insights from Baffin Island picrites. Geochimica et Cosmochimica Acta 238, 542–562. https://doi.org/10.1016/j.gca.2018.07.021

; Doucet et al., 2020

Doucet, L.S., Laurent, O., Ionov, D.A., Mattielli, N., Debaille, V., Debouge, W. (2020) Archean lithospheric differentiation: Insights from Fe and Zn isotopes. Geology 48, 1028–1032. https://doi.org/10.1130/G47647.1

), with a similar value established for depleted mid-ocean ridge basalt (MORB) mantle (δ66ZnDMM = +0.20 ± 0.05 ‰; Wang et al., 2017

Wang, Z.-Z., Liu, S.-A., Liu, J., Huang, J., Xiao, Y., Chu, Z.-Y., Zhao, X.-M., Tang, L. (2017) Zinc isotope fractionation during mantle melting and constraints on the Zn isotope composition of Earth’s upper mantle. Geochimica et Cosmochimica Acta 198, 151–167. https://doi.org/10.1016/j.gca.2016.11.014

). Observations of Zn isotope fractionation during partial melting (Wang et al., 2017

Wang, Z.-Z., Liu, S.-A., Liu, J., Huang, J., Xiao, Y., Chu, Z.-Y., Zhao, X.-M., Tang, L. (2017) Zinc isotope fractionation during mantle melting and constraints on the Zn isotope composition of Earth’s upper mantle. Geochimica et Cosmochimica Acta 198, 151–167. https://doi.org/10.1016/j.gca.2016.11.014

; Huang et al., 2018

Huang, J., Zhang, X.-C., Chen, S., Tang, L., Wörner, G., Yu, H., Huang, F. (2018) Zinc isotopic systematics of Kamchatka-Aleutian arc magmas controlled by mantle melting. Geochimica et Cosmochimica Acta 238, 85–101. https://doi.org/10.1016/j.gca.2018.07.012

) leave several outstanding questions prior to utilisation of Zn isotopes as tracers of distinct mantle and crustal reservoir contributions, and for estimating the BSE composition, particularly how Zn isotope variations differ in different melting regimes, such as during adiabatic decompression and during water-assisted mantle partial melting.

Limited data are available for Zn isotopes in MORB samples and data have yet to be reported for boninites. Boninites are potentially useful samples for understanding the behaviour of Zn during partial melting within the mantle. Unlike MORB, which are formed during adiabatic decompression melting, and hotspot volcanic rocks (OIB, komatiites) that are likely produced through thermal anomalies in the mantle, boninites are widely accepted to form from low pressure melting of previously depleted mantle sources by flux melting of water (e.g., Cameron et al., 1979

Cameron, W.E., Nisbet, E.G., Dietrich, V.J. (1979) Boninites, komatiites and ophiolitic basalts. Nature 280, 550–553. https://doi.org/10.1038/280550a0

; Crawford et al., 1989

Crawford, A.J., Falloon, T.J., Green, D.H. (1989) Classification, petrogenesis and tectonic setting of boninites. In: Crawford, A.J. (Ed.) Boninites and Related Rocks. Unwin Hyman, London, 1–49.

). These rocks should therefore have low Zn and relatively isotopically light δ66Zn due to their origin from refractory mantle similar to harzburgite (∼+0.16 ‰; Doucet et al., 2016

Doucet, L.S., Mattielli, N., Ionov, D.A., Debouge, W., Golovin, A.V. (2016) Zn isotopic heterogeneity in the mantle: A melting control? Earth and Planetary Science Letters 451, 232–240. https://doi.org/10.1016/j.epsl.2016.06.040

; Wang et al., 2017

Wang, Z.-Z., Liu, S.-A., Liu, J., Huang, J., Xiao, Y., Chu, Z.-Y., Zhao, X.-M., Tang, L. (2017) Zinc isotope fractionation during mantle melting and constraints on the Zn isotope composition of Earth’s upper mantle. Geochimica et Cosmochimica Acta 198, 151–167. https://doi.org/10.1016/j.gca.2016.11.014

). An alternative possibility exists that boninites might show Zn isotope variations due to variable additions of pelagic sedimentary components or altered igneous rocks that are known to have affected them from Sr-Nd-Hf-Pb isotope studies (e.g., Ishizuka et al., 2020

Ishizuka, O., Taylor, R.N., Umino, S., Kanayama, K. (2020) Geochemical Evolution of Arc and Slab Following Subduction Initiation: a Record from the Bonin Islands, Japan. Journal of Petrology 61, egaa050. https://doi.org/10.1093/petrology/egaa050

). The first boninite data are presented from Nakoudojima Island and the Bonin Ridge (Ishizuka et al., 2011

Ishizuka, O., Tani, K., Reagan, M.K., Kanayama, K., Umino, S., Harigane, Y., Sakamoto, I., Miyajima, Y., Yuasa, M., Dunkley, D.J. (2011) The timescales of subduction initiation and subsequent evolution of an oceanic island arc. Earth and Planetary Science Letters 306, 229–240. https://doi.org/10.1016/j.epsl.2011.04.006

, 2014

Ishizuka, O., Umino, S., Taylor, R.N., Kanayama, K. (2014) Evidence for Hydrothermal Activity in the Earliest Stages of Intraoceanic Arc Formation: Implications for Ophiolite-Hosted Hydrothermal Activity. Economic Geology 109, 2159–2178. https://doi.org/10.2113/econgeo.109.8.2159

), Japan, along with a new dataset for MORB from three ocean basins (Atlantic, Indian, Pacific) that greatly expands existing MORB Zn isotope data, to further examine Zn behaviour during partial melting processes.

top

Methods and Results

Abstract | Introduction | Methods and Results | Discussion | Acknowledgements | References | Supplementary Information


Methods for the separation and analysis of Zn abundances and isotopic composition are described in detail in the Supplementary Information. Five boninite samples from the Izu-Bonin Mariana Arc, Japan, were analysed for Zn isotope and elemental abundances (Table 1) for which bulk rock major- and trace-element data has been published previously (Ishizuka et al., 2011

Ishizuka, O., Tani, K., Reagan, M.K., Kanayama, K., Umino, S., Harigane, Y., Sakamoto, I., Miyajima, Y., Yuasa, M., Dunkley, D.J. (2011) The timescales of subduction initiation and subsequent evolution of an oceanic island arc. Earth and Planetary Science Letters 306, 229–240. https://doi.org/10.1016/j.epsl.2011.04.006

, 2014

Ishizuka, O., Umino, S., Taylor, R.N., Kanayama, K. (2014) Evidence for Hydrothermal Activity in the Earliest Stages of Intraoceanic Arc Formation: Implications for Ophiolite-Hosted Hydrothermal Activity. Economic Geology 109, 2159–2178. https://doi.org/10.2113/econgeo.109.8.2159

). These samples were erupted during the Eocene (44–48 Ma), span a range of MgO (4.4 to 15.2 wt. %) and SiO2 contents (55.2 to 62.4 wt. %), and have trace element and isotopic compositions consistent with partial melts originating from depleted mantle sources. Boninites have between 20 and 52 ppm Zn and δ66Zn from +0.17 ± 0.06 to +0.25 ± 0.06 ‰. Zinc isotope compositions are mass dependent and replicate analyses of KH07-2 D43 (Bonin Ridge) and 10100205 (Nakoudojima Island) provided identical results within uncertainties. Boninites have lower Zn abundances than MORB, OIB and most komatiites, and their average δ66Zn (+0.21 ± 0.06 ‰, 2 s.d.) is relatively homogeneous, similar to those measured for komatiites (Fig. S-1). For a given MgO content, boninites have the lowest Zn abundances and lie at the lowest values for δ66Zn of all basaltic rocks (Fig. 1).

Table 1 Zinc isotope and abundance data for boninites and mid-ocean ridge basalts. Major element data and ages for boninites are from Ishizuka et al. (2011

Ishizuka, O., Tani, K., Reagan, M.K., Kanayama, K., Umino, S., Harigane, Y., Sakamoto, I., Miyajima, Y., Yuasa, M., Dunkley, D.J. (2011) The timescales of subduction initiation and subsequent evolution of an oceanic island arc. Earth and Planetary Science Letters 306, 229–240. https://doi.org/10.1016/j.epsl.2011.04.006

, 2014)

Ishizuka, O., Umino, S., Taylor, R.N., Kanayama, K. (2014) Evidence for Hydrothermal Activity in the Earliest Stages of Intraoceanic Arc Formation: Implications for Ophiolite-Hosted Hydrothermal Activity. Economic Geology 109, 2159–2178. https://doi.org/10.2113/econgeo.109.8.2159

and for MORB are from Le Roux (2000)

Le Roux, P.J. (2000) The geochemistry of selected mid-ocean ridge basalts from the Southern mid-Atlantic ridge (40°-55°S). PhD Thesis, University of Cape Town. https://open.uct.ac.za/handle/11427/4207

and Deng et al. (2018)

Deng, Z., Moynier, F., Sossi, P.A., Chaussidon, M. (2018) Bridging the depleted MORB mantle and the continental crust using titanium isotopes. Geochemical Perspectives Letters 9, 11–15. https://doi.org/10.7185/geochemlet.1831

.

SampleLocationRock TypeAge
(Ma)		Uncer.SiO2
(wt. %)		MgO
(wt. %)		Zn
(ppm)		δ66Zn
(‰)		±2σδ68Zn
(‰)		±2σn
10100202Nakoudojima IslandBoninite47.810.1755.613.445.30.220.020.400.033
10100215Nakoudojima IslandBoninite55.215.250.70.170.060.340.094
10100213Nakoudojima IslandBronzite Andesite57.17.119.90.250.060.450.104
10100205*Nakoudojima IslandBasaltic Dike58.26.749.80.200.070.400.054
49.30.180.040.370.044
KH07-2 D43*Bonin RidgeSubmarine boninite44.780.1662.44.451.80.210.050.430.104
46.00.240.050.460.074
SWIFT DR32-1-3gSouth West Indian RidgeMORB66.20.290.030.610.066
SWIFT DR04-2-3gSouth West Indian RidgeMORB6.20.260.060.510.134
SWIFT DR06-3-6gSouth West Indian RidgeMORB6.678.60.300.060.610.085
MD57 D2-8Central Indian RidgeMORB5.860.40.310.030.620.096
MD57 D7-5Central Indian RidgeMORB8.150.20.290.040.560.076
MD57 D9-4Central Indian RidgeMORB59.80.260.050.520.115
EW9309 10D-3gMid Atlantic Ridge, 40–50°SMORB50.77.956.50.310.030.580.046
EW9309 27D-1gMid Atlantic Ridge, 40–50°SMORB49.08.340.30.260.060.560.115
EW9309 3DS Atlantic (Discovery)MORB50.67.0 0.310.020.610.073
EW9309 20D*S Atlantic (Discovery)MORB50.86.9 0.290.090.590.233
 0.240.030.470.032
RD87 DR18-102North Atlantic (Dosso)MORB7.454.70.260.050.490.106
RD87 DR24North Atlantic (Dosso)MORB56.40.270.060.500.116
RD87 DR29-107North Atlantic (Dosso)MORB44.80.270.010.540.044
DIVA1 12-2Mid Atlantic Ridge, 37–38°NMORB49.60.350.030.720.046
DIVA1 13-3Mid Atlantic Ridge, 37–38°NMORB7.660.20.330.040.620.076
DIVA1 15-5Mid Atlantic Ridge, 37–38°NMORB5.975.90.270.060.520.115
PAC2 DR32-1gPacific-Antarctic RidgeMORB6.290.60.240.010.490.032
PAC2 DR37-3gPacific-Antarctic RidgeMORB75.30.240.060.480.094
PAC2 DR38-1gPacific-Antarctic RidgeMORB   7.670.10.270.020.550.064
SEARISE1 DR04East Pacific RiseMORB  50.06.892.80.250.030.480.045
SEARISE2 DR03East Pacific RiseMORB  51.27.362.00.280.060.540.096
CYP 12-34East Pacific RiseMORB  50.48.265.80.270.050.510.085

*Replicate analyses are reported for these samples.




Figure 1 Bulk rock (a) Zn abundance and (b) isotopic composition as a function of MgO content for basaltic rocks. Data sources are given in Figure S-1, with Baffin Island bulk rock picrite data and the magmatic olivine compositions (assuming a possible range of olivine MgO contents) from McCoy-West et al. (2018)

McCoy-West, A.J., Fitton, J.G., Pons, M.L., Inglis, E.C., Williams, H.M. (2018) The Fe and Zn isotope composition of deep mantle source regions: Insights from Baffin Island picrites. Geochimica et Cosmochimica Acta 238, 542–562. https://doi.org/10.1016/j.gca.2018.07.021

. Martian shergottite and nakhlite data (shown as points and as a field) are from Paniello et al. (2012)

Paniello, R.C., Day, J.M.D., Moynier, F. (2012) Zinc isotopic evidence for the origin of the Moon. Nature 490, 376–379. https://doi.org/10.1038/nature11507

. New MORB data presented here (solid circles) are distinguished from published data (solid stars).
Full size image


Zinc isotope data are reported for 22 mid-ocean ridge basalts spanning the Atlantic, Indian and Pacific Ocean basins in Table 1. Zinc isotope compositions fall along a mass-dependent slope and span a range in Zn abundances (40 to 93 ppm) yet a restricted range of δ66Zn (+0.28 ± 0.06 ‰; 2 s.d.), with no evidence for systematic variation between ocean basins. These results agree well with the limited range in δ66Zn reported for MORB from the Atlantic and Indian oceans (Fig. S-1; Wang et al., 2017

Wang, Z.-Z., Liu, S.-A., Liu, J., Huang, J., Xiao, Y., Chu, Z.-Y., Zhao, X.-M., Tang, L. (2017) Zinc isotope fractionation during mantle melting and constraints on the Zn isotope composition of Earth’s upper mantle. Geochimica et Cosmochimica Acta 198, 151–167. https://doi.org/10.1016/j.gca.2016.11.014

), although the Zn abundances span a wider range.

top

Discussion

Abstract | Introduction | Methods and Results | Discussion | Acknowledgements | References | Supplementary Information


Boninite magmatism and the cause of low Zn and δ66Zn. Partial melting occurs within shallow forearc regions during release of volatiles from the subducting slab into the mantle wedge and drives boninite magmatism (e.g., Ishizuka et al., 2011

Ishizuka, O., Tani, K., Reagan, M.K., Kanayama, K., Umino, S., Harigane, Y., Sakamoto, I., Miyajima, Y., Yuasa, M., Dunkley, D.J. (2011) The timescales of subduction initiation and subsequent evolution of an oceanic island arc. Earth and Planetary Science Letters 306, 229–240. https://doi.org/10.1016/j.epsl.2011.04.006

, 2014

Ishizuka, O., Umino, S., Taylor, R.N., Kanayama, K. (2014) Evidence for Hydrothermal Activity in the Earliest Stages of Intraoceanic Arc Formation: Implications for Ophiolite-Hosted Hydrothermal Activity. Economic Geology 109, 2159–2178. https://doi.org/10.2113/econgeo.109.8.2159

; Shervais et al., 2021

Shervais, J.W., Reagan, M.K., Godard, M., Prytulak, J., Ryan, J.G., et al. (2021) Magmatic Response to Subduction Initiation, Part II: Boninites and Related Rocks of the Izu-Bonin Arc From IODP Expedition 352. Geochemistry, Geophysics, Geosystems 22, e2020GC009093. https://doi.org/10.1029/2020GC009093

). Such melting produces high MgO (>8 wt. %), SiO2 (52–63 wt. %), and low TiO2 (<0.5 wt. %) melts distinct from komatiites, MORB, OIB or any other type of basaltic rock (e.g., Crawford et al., 1989

Crawford, A.J., Falloon, T.J., Green, D.H. (1989) Classification, petrogenesis and tectonic setting of boninites. In: Crawford, A.J. (Ed.) Boninites and Related Rocks. Unwin Hyman, London, 1–49.

), and which can be reproduced in melting experiments mimicking hydrous melting conditions of refractory harzburgite-like lithologies (Van der Laan et al., 1989

Van der Laan, S.R., Flower, M.J.F., Koster van Groos, A.F. (1989) Experimental evidence for the origin of boninites: near-liquidus phase relations to 7.5 kbar. In: Crawford, A.J. (Ed.) Boninites and Related Rocks. Unwin Hyman, London, 112–147.

; Falloon and Danyushevsky, 2000

Falloon, T.J., Danyushevsky, L.V. (2000) Melting of Refractory Mantle at 1.5, 2 and 2.5 GPa under Anhydrous and H2O-undersaturated Conditions: Implications for the Petrogenesis of High-Ca Boninites and the Influence of Subduction Components on Mantle Melting. Journal of Petrology 41, 257–283. https://doi.org/10.1093/petrology/41.2.257

). Forearc mantle is dominated by highly melt-depleted (residues after >20 % partial melting) harzburgite mantle formed by both modern and ancient melt events and is characterised by low Zn contents (Day and Brown, 2021

Day, J.M.D., Brown, D.B. (2021) Ancient Melt-Depletion in Fresh to Strongly Serpentinized Tonga Trench Peridotites. Journal of Petrology 62, egab088. https://doi.org/10.1093/petrology/egab088

), which can only be further melted by additions of fluids or excess temperatures. The fluid-assisted melting of such shallow depleted residues within the melting region of the nascent Izu-Bonin-Mariana subduction system in the Eocene is widely accepted as the cause of boninite magmatism at that time (Ishizuka et al., 2011

Ishizuka, O., Tani, K., Reagan, M.K., Kanayama, K., Umino, S., Harigane, Y., Sakamoto, I., Miyajima, Y., Yuasa, M., Dunkley, D.J. (2011) The timescales of subduction initiation and subsequent evolution of an oceanic island arc. Earth and Planetary Science Letters 306, 229–240. https://doi.org/10.1016/j.epsl.2011.04.006

; Shervais et al., 2021

Shervais, J.W., Reagan, M.K., Godard, M., Prytulak, J., Ryan, J.G., et al. (2021) Magmatic Response to Subduction Initiation, Part II: Boninites and Related Rocks of the Izu-Bonin Arc From IODP Expedition 352. Geochemistry, Geophysics, Geosystems 22, e2020GC009093. https://doi.org/10.1029/2020GC009093

).

Doucet et al. (2016)

Doucet, L.S., Mattielli, N., Ionov, D.A., Debouge, W., Golovin, A.V. (2016) Zn isotopic heterogeneity in the mantle: A melting control? Earth and Planetary Science Letters 451, 232–240. https://doi.org/10.1016/j.epsl.2016.06.040

and Wang et al. (2017)

Wang, Z.-Z., Liu, S.-A., Liu, J., Huang, J., Xiao, Y., Chu, Z.-Y., Zhao, X.-M., Tang, L. (2017) Zinc isotope fractionation during mantle melting and constraints on the Zn isotope composition of Earth’s upper mantle. Geochimica et Cosmochimica Acta 198, 151–167. https://doi.org/10.1016/j.gca.2016.11.014

presented data for refractory harzburgites from a variety of continental lithospheric mantle localities for δ66Zn, showing that they have generally similar Zn abundances to forearc harzburgites (∼40 ppm) and lower δ66Zn (∼0.16 ‰) than lherzolites (∼0.28 ‰; Fig. 1). Furthermore, analyses of minerals from spinel lherzolites indicate that olivine and orthopyroxene have lower δ66Zn than clinopyroxene, or the main carrier of Zn, spinel, which is characterised by heavy δ66Zn (Wang et al., 2017

Wang, Z.-Z., Liu, S.-A., Liu, J., Huang, J., Xiao, Y., Chu, Z.-Y., Zhao, X.-M., Tang, L. (2017) Zinc isotope fractionation during mantle melting and constraints on the Zn isotope composition of Earth’s upper mantle. Geochimica et Cosmochimica Acta 198, 151–167. https://doi.org/10.1016/j.gca.2016.11.014

). From these constraints, it is likely that boninite Zn systematics are a direct consequence of flux melting of low-Zn depleted harzburgites at low pressure.

Two potential issues are whether Zn can be affected by subducted pelagic sedimentary components or altered igneous rocks, and the possibility of mineral accumulation processes modifying Zn isotope systematics of boninites. Nakoudojima boninites have 87Sr/86Sri (0.7041–0.7049), 143Nd/144Ndi (0.5127–0.5129) and Pb isotope systematics consistent with variable contamination from pelagic sediment (Ishizuka et al., 2020

Ishizuka, O., Taylor, R.N., Umino, S., Kanayama, K. (2020) Geochemical Evolution of Arc and Slab Following Subduction Initiation: a Record from the Bonin Islands, Japan. Journal of Petrology 61, egaa050. https://doi.org/10.1093/petrology/egaa050

). Drill core samples from the Bonin Ridge (e.g., KH07-2 D43) have more MORB-like 87Sr/86Sri of 0.7031–0.7037 and 143Nd/144Ndi of 0.5131–0.5132 (Ishizuka et al., 2011

Ishizuka, O., Tani, K., Reagan, M.K., Kanayama, K., Umino, S., Harigane, Y., Sakamoto, I., Miyajima, Y., Yuasa, M., Dunkley, D.J. (2011) The timescales of subduction initiation and subsequent evolution of an oceanic island arc. Earth and Planetary Science Letters 306, 229–240. https://doi.org/10.1016/j.epsl.2011.04.006

). Despite the limited number of boninite samples studied, the large isotopic variations in Sr-Nd-Pb, yet limited range in Zn isotope ratios, suggest Zn in boninites is likely to be relatively unaffected by subducting sediment or altered igneous rock components. An alternative source for low δ66Zn is accumulation of olivine with isotopically light Zn (McCoy-West et al., 2018

McCoy-West, A.J., Fitton, J.G., Pons, M.L., Inglis, E.C., Williams, H.M. (2018) The Fe and Zn isotope composition of deep mantle source regions: Insights from Baffin Island picrites. Geochimica et Cosmochimica Acta 238, 542–562. https://doi.org/10.1016/j.gca.2018.07.021

; Sossi et al., 2018

Sossi, P.A., Nebel, O., O’Neill, H.S.C., Moynier, F. (2018) Zinc isotope composition of the Earth and its behaviour during planetary accretion. Chemical Geology 477, 73–84. https://doi.org/10.1016/j.chemgeo.2017.12.006

). This is not the case for boninites examined here, which do not contain significant accumulative olivine. Compared with komatiites, OIB, alkali basalts, basanites, and nephelinites, boninites plot at the lowest Zn and δ66Zn for a given MgO of any terrestrial basaltic partial melt (Fig. 1). New data for MORB extend the range of Zn to abundances that can be as low as in boninites. This may indicate the presence of harzburgite source components in some MORB.

Mantle partial melting controls on Zn isotope fractionation.Due to their genesis in an unusual tectonic setting, boninites represent some of the largest-degree partial melts of all modern mantle-derived melts (Supplementary Information, Table 2). There are relationships between δ66Zn and Zn and the extent of partial melting in basaltic rocks (Fig. 2). With decreasing extents of partial melting, both δ66Zn and Zn concentrations tend to increase. An increase in incompatible (D < 1) Zn is expected at lower extents of partial melting, but the cause(s) of δ66Zn increases are less clear. For example, this relationship could indicate obfuscation of components enriched in Zn with high δ66Zn, such as recycled carbonates, at higher degrees of partial melting (Beunon et al., 2020

Beunon, H., Mattielli, N., Doucet, L.S., Moine, B., Debret, B. (2020) Mantle heterogeneity through Zn systematics in oceanic basalts: Evidence for a deep carbon cycling. Earth-Science Reviews 205, 103174. https://doi.org/10.1016/j.earscirev.2020.103174

). Alternatively, such variations may result from melting variably depleted or fertile mantle peridotite without the obvious presence of recycled components, noting the difficulty in melting refractory mantle sources at low degrees of partial melting. To examine these possibilities, a family of non-modal partial melting models were constructed for Zn using previous modelling compilations (see Table S-1). The first assumes a fertile mantle garnet lherzolite composition (green lines) similar to that considered previously (e.g., Sossi et al., 2018

Sossi, P.A., Nebel, O., O’Neill, H.S.C., Moynier, F. (2018) Zinc isotope composition of the Earth and its behaviour during planetary accretion. Chemical Geology 477, 73–84. https://doi.org/10.1016/j.chemgeo.2017.12.006

); the second is a refractory mantle spinel harzburgite (red lines), and the third is a metasomatised mantle composition (blue lines) (Fig. 2).

Table 2 Zinc isotope, abundance data, melting type and extent for terrestrial basalts.
Rock TypeMelting ProcessPartial
Melting
Extent		Estimate
(%)		Locationδ66Zn
(‰)		±2σAv. Zn
(ppm)		Range Zn
(ppm)		nData Sources
KomatiiteHigh- to low-P high-T Melting<30–50 %40 ± 10Plume?0.160.066622–13918 Sossi et al. (2018)

Sossi, P.A., Nebel, O., O’Neill, H.S.C., Moynier, F. (2018) Zinc isotope composition of the Earth and its behaviour during planetary accretion. Chemical Geology 477, 73–84. https://doi.org/10.1016/j.chemgeo.2017.12.006

; Herzberg (1992)

Herzberg, C. (1992) Depth and degree of melting of komatiites. Journal of Geophysical Research: Solid Earth 97, 4521–4540. https://doi.org/10.1029/91JB03066

BoniniteLow-P Flux Melting12–23 %18 ± 8Forearc0.210.064520–525This study; Shervais et al. (2021)

Shervais, J.W., Reagan, M.K., Godard, M., Prytulak, J., Ryan, J.G., et al. (2021) Magmatic Response to Subduction Initiation, Part II: Boninites and Related Rocks of the Izu-Bonin Arc From IODP Expedition 352. Geochemistry, Geophysics, Geosystems 22, e2020GC009093. https://doi.org/10.1029/2020GC009093

MORBAdiabatic Decompression∼10 %10 ± 4Ridge0.280.066440–9328 This study; Wang et al. (2017)

Wang, Z.-Z., Liu, S.-A., Liu, J., Huang, J., Xiao, Y., Chu, Z.-Y., Zhao, X.-M., Tang, L. (2017) Zinc isotope fractionation during mantle melting and constraints on the Zn isotope composition of Earth’s upper mantle. Geochimica et Cosmochimica Acta 198, 151–167. https://doi.org/10.1016/j.gca.2016.11.014

OIBHigh-T Melting2–10 %5 ± 4Plume0.300.1012081–16557 Chen et al. (2013)

Chen, H., Savage, P.S., Teng, F.-Z., Helz, R.T., Moynier, F. (2013) Zinc isotope fractionation during magmatic differentiation and the isotopic composition of the bulk Earth. Earth and Planetary Science Letters 369–370, 32–42. https://doi.org/10.1016/j.epsl.2013.02.037

; Wang et al. (2017)

Wang, Z.-Z., Liu, S.-A., Liu, J., Huang, J., Xiao, Y., Chu, Z.-Y., Zhao, X.-M., Tang, L. (2017) Zinc isotope fractionation during mantle melting and constraints on the Zn isotope composition of Earth’s upper mantle. Geochimica et Cosmochimica Acta 198, 151–167. https://doi.org/10.1016/j.gca.2016.11.014

Alkali BasaltLow-degree melt<4 %4 ± 2Various0.340.069691–10618 Wang et al. (2018)

Wang, Z.-Z., Liu, S.-A., Chen, L.-H., Li, S.-G., Zeng, G. (2018) Compositional transition in natural alkaline lavas through silica-undersaturated melt–lithosphere interaction. Geology 46, 771–774. https://doi.org/10.1130/G45145.1

BasaniteLow-degree melt<3 %3 ± 2Various0.420.0812295–16320 Wang et al. (2018)

Wang, Z.-Z., Liu, S.-A., Chen, L.-H., Li, S.-G., Zeng, G. (2018) Compositional transition in natural alkaline lavas through silica-undersaturated melt–lithosphere interaction. Geology 46, 771–774. https://doi.org/10.1130/G45145.1

NepheliniteLow-degree melt<2 %2 ± 1Various0.450.06168149–18419 Wang et al. (2018)

Wang, Z.-Z., Liu, S.-A., Chen, L.-H., Li, S.-G., Zeng, G. (2018) Compositional transition in natural alkaline lavas through silica-undersaturated melt–lithosphere interaction. Geology 46, 771–774. https://doi.org/10.1130/G45145.1




Figure 2 (a) Zinc isotopic composition and (b) Zn abundance versus estimated degree of partial melting for basaltic rocks (Table 2) versus models for partial melting of harzburgite, lherzolite and metasomatised peridotite, with dots conforming to 1 % increments of melting. Shown in upper panel are the estimated bulk silicate Earth (BSE) average (solid line) and standard deviation (shaded regions) values from Chen et al. (2013)

Chen, H., Savage, P.S., Teng, F.-Z., Helz, R.T., Moynier, F. (2013) Zinc isotope fractionation during magmatic differentiation and the isotopic composition of the bulk Earth. Earth and Planetary Science Letters 369–370, 32–42. https://doi.org/10.1016/j.epsl.2013.02.037

and Sossi et al. (2018)

Sossi, P.A., Nebel, O., O’Neill, H.S.C., Moynier, F. (2018) Zinc isotope composition of the Earth and its behaviour during planetary accretion. Chemical Geology 477, 73–84. https://doi.org/10.1016/j.chemgeo.2017.12.006

.
Full size image


Of interest for boninite petrogenesis is the refractory mantle harzburgite model. Assuming a starting composition similar to harzburgites presented by Doucet et al. (2016)

Doucet, L.S., Mattielli, N., Ionov, D.A., Debouge, W., Golovin, A.V. (2016) Zn isotopic heterogeneity in the mantle: A melting control? Earth and Planetary Science Letters 451, 232–240. https://doi.org/10.1016/j.epsl.2016.06.040

this model yields Zn and δ66Zn values within uncertainty of measured boninite compositions and expected degrees of partial melting (Fig. 2). The model supports the low Zn and δ66Zn of boninites reflecting partial melting of refractory forearc mantle peridotites. MORB lie slightly above the partial melting estimates for spinel harzburgite, yet are also consistent with generation from a relatively refractory mantle source, as would be expected for melts of DMM. Conversely, the fertile lherzolite mantle composition of Doucet et al. (2016)

Doucet, L.S., Mattielli, N., Ionov, D.A., Debouge, W., Golovin, A.V. (2016) Zn isotopic heterogeneity in the mantle: A melting control? Earth and Planetary Science Letters 451, 232–240. https://doi.org/10.1016/j.epsl.2016.06.040

can reproduce the Zn abundances observed in OIB and komatiites, but not the δ66Zn of komatiites, where a low δ66Zn source, more akin to refractory harzburgite, better matches komatiite data. OIB span a range of compositions that could be explained by mixtures of both refractory and fertile peridotite sources. Conversely, low-degree partial melts like alkali basalts, basanites and nephelinites from eastern China (Wang et al., 2018

Wang, Z.-Z., Liu, S.-A., Chen, L.-H., Li, S.-G., Zeng, G. (2018) Compositional transition in natural alkaline lavas through silica-undersaturated melt–lithosphere interaction. Geology 46, 771–774. https://doi.org/10.1130/G45145.1

) require fertile melt compositions, but their high Zn abundances also permit contributions from hybridised metasomatised peridotite mantle sources.

Mantle-derived basaltic rocks can span ∼0.4 ‰ variation in δ66Zn. In the absence of other evidence, such as O-Ca-Sr-Nd-Os-Pb isotopic variability reflecting distinct mantle or crustal components, much of the variation in Zn isotopes can be attributed to different extents of partial melting of variably fertile and refractory peridotite mantle sources. Studies using Zn isotopes to examine potential carbonate or enriched recycled mantle sources in OIB and related rocks should consider the potential for partial melting control on Zn isotope fractionation. Relatively low degree partial melts are likely to sample more extreme end member compositions for Zn. The highest degree partial melts examined are MORB, boninites and komatiites, and these necessarily sample the largest regions of mantle and so are most likely to express mantle composition. Based on data trends and models, these magmatic rocks appear to sample mantle sources akin to harzburgite mantle for δ66Zn. In this sense, high degree partial melts from ancient (komatiites) and modern (boninites, MORB) magmatism conform to the value determined by Sossi et al. (2018)

Sossi, P.A., Nebel, O., O’Neill, H.S.C., Moynier, F. (2018) Zinc isotope composition of the Earth and its behaviour during planetary accretion. Chemical Geology 477, 73–84. https://doi.org/10.1016/j.chemgeo.2017.12.006

of +0.16 ± 0.06 ‰, consistent with BSE being perhaps up to +0.2 ‰ (e.g., McCoy-West et al., 2018

McCoy-West, A.J., Fitton, J.G., Pons, M.L., Inglis, E.C., Williams, H.M. (2018) The Fe and Zn isotope composition of deep mantle source regions: Insights from Baffin Island picrites. Geochimica et Cosmochimica Acta 238, 542–562. https://doi.org/10.1016/j.gca.2018.07.021

; Doucet et al., 2020

Doucet, L.S., Laurent, O., Ionov, D.A., Mattielli, N., Debaille, V., Debouge, W. (2020) Archean lithospheric differentiation: Insights from Fe and Zn isotopes. Geology 48, 1028–1032. https://doi.org/10.1130/G47647.1

).

It has been shown that basaltic rocks from the Moon have δ66Zn values ∼1 ‰ heavier and Zn abundances >20 times less than terrestrial basalts, interpreted to reflect volatile loss during lunar formation (Paniello et al., 2012

Paniello, R.C., Day, J.M.D., Moynier, F. (2012) Zinc isotopic evidence for the origin of the Moon. Nature 490, 376–379. https://doi.org/10.1038/nature11507

). Such large differences between basalts from the Earth and Moon cannot be reconciled by differences in extents of partial melting. Paniello et al. (2012)

Paniello, R.C., Day, J.M.D., Moynier, F. (2012) Zinc isotopic evidence for the origin of the Moon. Nature 490, 376–379. https://doi.org/10.1038/nature11507

also presented Zn isotope data for martian meteorites (Fig. 1). Shergottites are relatively high-degree basaltic partial melts of depleted and enriched mantle sources in Mars, while nakhlites have been considered akin to rejuvenated lava, which are formed by lower degrees of partial melting, meaning the closest analogues to these samples are plume-derived melts on Earth (Day et al., 2018

Day, J.M.D., Tait, K.T., Udry, A., Moynier, F., Liu, Y., Neal, C.R. (2018) Martian magmatism from plume metasomatized mantle. Nature Communications 9, 4799. https://doi.org/10.1038/s41467-018-07191-0

). Martian meteorites have low Zn contents and δ66Zn in the upper range of terrestrial plume-derived lavas (OIB, komatiites) for a given MgO content. Such results are consistent with lower Zn contents in bulk silicate Mars and may suggest a slightly heavier bulk silicate Mars value for δ66Zn compared to Earth. This conclusion needs to be confirmed by further analyses of martian meteorites.

top

Acknowledgements

Abstract | Introduction | Methods and Results | Discussion | Acknowledgements | References | Supplementary Information


Reviews by L. Doucet and an anonymous referee are appreciated. An Institut de Physique du Globe de Paris Visiting Professor position (JMDD), ERC grant agreement No. 101001282 (METAL), the UnivEarthS Labex program (numbers: ANR-10-LABX-0023 and ANR-11-IDEX-0005-02), the IPGP multidisciplinary program PARI, and the Region île-de-France DIM ACAV + and SESAME Grants no. 12015908 supported this work.

Editor: Anat Shahar

top

References

Abstract | Introduction | Methods and Results | Discussion | Acknowledgements | References | Supplementary Information

Beunon, H., Mattielli, N., Doucet, L.S., Moine, B., Debret, B. (2020) Mantle heterogeneity through Zn systematics in oceanic basalts: Evidence for a deep carbon cycling. Earth-Science Reviews 205, 103174. https://doi.org/10.1016/j.earscirev.2020.103174
Show in context

Zinc stable isotopes have utility in the study of basaltic rocks, from the examination of contributions of potential mantle and crustal reservoirs with distinct 66Zn/64Zn ratios (e.g., Beunon et al., 2020), to identification of volatile element loss in planets (e.g., Paniello et al., 2012).
View in article
Partial melting processes can explain δ66Zn variations in arc lavas (Huang et al., 2018), whereas it has been argued that δ66Zn variations in some mantle-derived basaltic rocks reflect contributions from distinct sources, including recycled carbonate (Beunon et al., 2020; Liu et al., 2022).
View in article
For example, this relationship could indicate obfuscation of components enriched in Zn with high δ66Zn, such as recycled carbonates, at higher degrees of partial melting (Beunon et al., 2020).
View in article


Cameron, W.E., Nisbet, E.G., Dietrich, V.J. (1979) Boninites, komatiites and ophiolitic basalts. Nature 280, 550–553. https://doi.org/10.1038/280550a0
Show in context

Unlike MORB, which are formed during adiabatic decompression melting, and hotspot volcanic rocks (OIB, komatiites) that are likely produced through thermal anomalies in the mantle, boninites are widely accepted to form from low pressure melting of previously depleted mantle sources by flux melting of water (e.g., Cameron et al., 1979; Crawford et al., 1989).
View in article


Chen, H., Savage, P.S., Teng, F.-Z., Helz, R.T., Moynier, F. (2013) Zinc isotope fractionation during magmatic differentiation and the isotopic composition of the bulk Earth. Earth and Planetary Science Letters 369–370, 32–42. https://doi.org/10.1016/j.epsl.2013.02.037
Show in context

Studies of basaltic magmatic differentiation series have demonstrated that Zn isotopes are not substantially modified (<0.1 ‰) during fractional crystallisation processes (Chen et al., 2013), but that significant fractionation (≥0.1 ‰) in δ66Zn is likely to occur during mantle partial melting (Wang et al., 2017).
View in article
Intraplate volcanic rocks, including ocean island basalts (OIB), were first used to estimate a bulk silicate Earth (BSE) Zn isotopic composition (δ66ZnBSE = +0.28 ± 0.05 ‰; all uncertainties 2 s.d.; Chen et al., 2013), followed by direct analysis of fertile peridotites (+0.30 ± 0.03 ‰; Doucet et al., 2016).
View in article
Chen et al. (2013); Wang et al. (2017).
View in article
Shown in upper panel are the estimated bulk silicate Earth (BSE) average (solid line) and standard deviation (shaded regions) values from Chen et al. (2013) and Sossi et al. (2018).
View in article


Crawford, A.J., Falloon, T.J., Green, D.H. (1989) Classification, petrogenesis and tectonic setting of boninites. In: Crawford, A.J. (Ed.) Boninites and Related Rocks. Unwin Hyman, London, 1–49.
Show in context

Unlike MORB, which are formed during adiabatic decompression melting, and hotspot volcanic rocks (OIB, komatiites) that are likely produced through thermal anomalies in the mantle, boninites are widely accepted to form from low pressure melting of previously depleted mantle sources by flux melting of water (e.g., Cameron et al., 1979; Crawford et al., 1989).
View in article
Such melting produces high MgO (>8 wt. %), SiO2 (52–63 wt. %), and low TiO2 (<0.5 wt. %) melts distinct from komatiites, MORB, OIB or any other type of basaltic rock (e.g., Crawford et al., 1989), and which can be reproduced in melting experiments mimicking hydrous melting conditions of refractory harzburgite-like lithologies (Van der Laan et al., 1989; Falloon and Danyushevsky, 2000).
View in article


Day, J.M.D., Brown, D.B. (2021) Ancient Melt-Depletion in Fresh to Strongly Serpentinized Tonga Trench Peridotites. Journal of Petrology 62, egab088. https://doi.org/10.1093/petrology/egab088
Show in context

Forearc mantle is dominated by highly melt-depleted (residues after >20 % partial melting) harzburgite mantle formed by both modern and ancient melt events and is characterised by low Zn contents (Day and Brown, 2021), which can only be further melted by additions of fluids or excess temperatures.
View in article


Day, J.M.D., Tait, K.T., Udry, A., Moynier, F., Liu, Y., Neal, C.R. (2018) Martian magmatism from plume metasomatized mantle. Nature Communications 9, 4799. https://doi.org/10.1038/s41467-018-07191-0
Show in context

Shergottites are relatively high-degree basaltic partial melts of depleted and enriched mantle sources in Mars, while nakhlites have been considered akin to rejuvenated lava, which are formed by lower degrees of partial melting, meaning the closest analogues to these samples are plume-derived melts on Earth (Day et al., 2018).
View in article


Deng, Z., Moynier, F., Sossi, P.A., Chaussidon, M. (2018) Bridging the depleted MORB mantle and the continental crust using titanium isotopes. Geochemical Perspectives Letters 9, 11–15. https://doi.org/10.7185/geochemlet.1831
Show in context

Major element data and ages for boninites are from Ishizuka et al. (2011, 2014) and for MORB are from Le Roux (2000) and Deng et al. (2018).
View in article


Doucet, L.S., Mattielli, N., Ionov, D.A., Debouge, W., Golovin, A.V. (2016) Zn isotopic heterogeneity in the mantle: A melting control? Earth and Planetary Science Letters 451, 232–240. https://doi.org/10.1016/j.epsl.2016.06.040
Show in context

Intraplate volcanic rocks, including ocean island basalts (OIB), were first used to estimate a bulk silicate Earth (BSE) Zn isotopic composition (δ66ZnBSE = +0.28 ± 0.05 ‰; all uncertainties 2 s.d.; Chen et al., 2013), followed by direct analysis of fertile peridotites (+0.30 ± 0.03 ‰; Doucet et al., 2016).
View in article
These rocks should therefore have low Zn and relatively isotopically light δ66Zn due to their origin from refractory mantle similar to harzburgite (∼+0.16 ‰; Doucet et al., 2016; Wang et al., 2017).
View in article
Doucet et al. (2016) and Wang et al. (2017) presented data for refractory harzburgites from a variety of continental lithospheric mantle localities for δ66Zn, showing that they have generally similar Zn abundances to forearc harzburgites (∼40 ppm) and lower δ66Zn (∼0.16 ‰) than lherzolites (∼0.28 ‰; Fig. 1).
View in article
Of interest for boninite petrogenesis is the refractory mantle harzburgite model. Assuming a starting composition similar to harzburgites presented by Doucet et al. (2016) this model yields Zn and δ66Zn values within uncertainty of measured boninite compositions and expected degrees of partial melting (Fig. 2).
View in article
Conversely, the fertile lherzolite mantle composition of Doucet et al. (2016) can reproduce the Zn abundances observed in OIB and komatiites, but not the δ66Zn of komatiites, where a low δ66Zn source, more akin to refractory harzburgite, better matches komatiite data.
View in article


Doucet, L.S., Laurent, O., Ionov, D.A., Mattielli, N., Debaille, V., Debouge, W. (2020) Archean lithospheric differentiation: Insights from Fe and Zn isotopes. Geology 48, 1028–1032. https://doi.org/10.1130/G47647.1
Show in context

Subsequently, peridotites and komatiites have been shown to have indistinguishable Zn isotopic compositions (+0.16 ± 0.06 ‰) and so were interpreted to represent a consistent δ66ZnBSE (Sossi et al., 2018), with slightly more elevated estimates also reported from modern and ancient mantle melts (+0.20 ±0.03 ‰; McCoy-West et al., 2018; Doucet et al., 2020), with a similar value established for depleted mid-ocean ridge basalt (MORB) mantle (δ66ZnDMM = +0.20 ± 0.05 ‰; Wang et al., 2017).
View in article
In this sense, high degree partial melts from ancient (komatiites) and modern (boninites, MORB) magmatism conform to the value determined by Sossi et al. (2018) of +0.16 ± 0.06 ‰, consistent with BSE being perhaps up to +0.2 ‰ (e.g., McCoy-West et al., 2018; Doucet et al., 2020).
View in article


Falloon, T.J., Danyushevsky, L.V. (2000) Melting of Refractory Mantle at 1.5, 2 and 2.5 GPa under Anhydrous and H2O-undersaturated Conditions: Implications for the Petrogenesis of High-Ca Boninites and the Influence of Subduction Components on Mantle Melting. Journal of Petrology 41, 257–283. https://doi.org/10.1093/petrology/41.2.257
Show in context

Such melting produces high MgO (>8 wt. %), SiO2 (52–63 wt. %), and low TiO2 (<0.5 wt. %) melts distinct from komatiites, MORB, OIB or any other type of basaltic rock (e.g., Crawford et al., 1989), and which can be reproduced in melting experiments mimicking hydrous melting conditions of refractory harzburgite-like lithologies (Van der Laan et al., 1989; Falloon and Danyushevsky, 2000).
View in article


Herzberg, C. (1992) Depth and degree of melting of komatiites. Journal of Geophysical Research: Solid Earth 97, 4521–4540. https://doi.org/10.1029/91JB03066
Show in context

Sossi et al. (2018); Herzberg (1992).
View in article


Huang, J., Zhang, X.-C., Chen, S., Tang, L., Wörner, G., Yu, H., Huang, F. (2018) Zinc isotopic systematics of Kamchatka-Aleutian arc magmas controlled by mantle melting. Geochimica et Cosmochimica Acta 238, 85–101. https://doi.org/10.1016/j.gca.2018.07.012
Show in context

Partial melting processes can explain δ66Zn variations in arc lavas (Huang et al., 2018), whereas it has been argued that δ66Zn variations in some mantle-derived basaltic rocks reflect contributions from distinct sources, including recycled carbonate (Beunon et al., 2020; Liu et al., 2022).
View in article
Observations of Zn isotope fractionation during partial melting (Wang et al., 2017; Huang et al., 2018) leave several outstanding questions prior to utilisation of Zn isotopes as tracers of distinct mantle and crustal reservoir contributions, and for estimating the BSE composition, particularly how Zn isotope variations differ in different melting regimes, such as during adiabatic decompression and during water-assisted mantle partial melting.
View in article


Ishizuka, O., Tani, K., Reagan, M.K., Kanayama, K., Umino, S., Harigane, Y., Sakamoto, I., Miyajima, Y., Yuasa, M., Dunkley, D.J. (2011) The timescales of subduction initiation and subsequent evolution of an oceanic island arc. Earth and Planetary Science Letters 306, 229–240. https://doi.org/10.1016/j.epsl.2011.04.006
Show in context

The first boninite data are presented from Nakoudojima Island and the Bonin Ridge (Ishizuka et al., 2011, 2014), Japan, along with a new dataset for MORB from three ocean basins (Atlantic, Indian, Pacific) that greatly expands existing MORB Zn isotope data, to further examine Zn behaviour during partial melting processes.
View in article
Five boninite samples from the Izu-Bonin Mariana Arc, Japan, were analysed for Zn isotope and elemental abundances (Table 1) for which bulk rock major- and trace-element data has been published previously (Ishizuka et al., 2011, 2014).
View in article
Major element data and ages for boninites are from Ishizuka et al. (2011, 2014) and for MORB are from Le Roux (2000) and Deng et al. (2018).
View in article
Partial melting occurs within shallow forearc regions during release of volatiles from the subducting slab into the mantle wedge and drives boninite magmatism (e.g., Ishizuka et al., 2011, 2014; Shervais et al., 2021).
View in article
The fluid-assisted melting of such shallow depleted residues within the melting region of the nascent Izu-Bonin-Mariana subduction system in the Eocene is widely accepted as the cause of boninite magmatism at that time (Ishizuka et al., 2011; Shervais et al., 2021).
View in article
Drill core samples from the Bonin Ridge (e.g., KH07-2 D43) have more MORB-like 87Sr/86Sri of 0.7031–0.7037 and 143Nd/144Ndi of 0.5131–0.5132 (Ishizuka et al., 2011).
View in article


Ishizuka, O., Umino, S., Taylor, R.N., Kanayama, K. (2014) Evidence for Hydrothermal Activity in the Earliest Stages of Intraoceanic Arc Formation: Implications for Ophiolite-Hosted Hydrothermal Activity. Economic Geology 109, 2159–2178. https://doi.org/10.2113/econgeo.109.8.2159
Show in context

The first boninite data are presented from Nakoudojima Island and the Bonin Ridge (Ishizuka et al., 2011, 2014), Japan, along with a new dataset for MORB from three ocean basins (Atlantic, Indian, Pacific) that greatly expands existing MORB Zn isotope data, to further examine Zn behaviour during partial melting processes.
View in article
Five boninite samples from the Izu-Bonin Mariana Arc, Japan, were analysed for Zn isotope and elemental abundances (Table 1) for which bulk rock major- and trace-element data has been published previously (Ishizuka et al., 2011, 2014).
View in article
Major element data and ages for boninites are from Ishizuka et al. (2011, 2014) and for MORB are from Le Roux (2000) and Deng et al. (2018).
View in article
Partial melting occurs within shallow forearc regions during release of volatiles from the subducting slab into the mantle wedge and drives boninite magmatism (e.g., Ishizuka et al., 2011, 2014; Shervais et al., 2021).
View in article


Ishizuka, O., Taylor, R.N., Umino, S., Kanayama, K. (2020) Geochemical Evolution of Arc and Slab Following Subduction Initiation: a Record from the Bonin Islands, Japan. Journal of Petrology 61, egaa050. https://doi.org/10.1093/petrology/egaa050
Show in context

An alternative possibility exists that boninites might show Zn isotope variations due to variable additions of pelagic sedimentary components or altered igneous rocks that are known to have affected them from Sr-Nd-Hf-Pb isotope studies (e.g., Ishizuka et al., 2020).
View in article
Nakoudojima boninites have 87Sr/86Sri (0.7041–0.7049), 143Nd/144Ndi (0.5127–0.5129) and Pb isotope systematics consistent with variable contamination from pelagic sediment (Ishizuka et al., 2020).
View in article


Le Roux, P.J. (2000) The geochemistry of selected mid-ocean ridge basalts from the Southern mid-Atlantic ridge (40°-55°S). PhD Thesis, University of Cape Town. https://open.uct.ac.za/handle/11427/4207
Show in context

Major element data and ages for boninites are from Ishizuka et al. (2011, 2014) and for MORB are from Le Roux (2000) and Deng et al. (2018).
View in article


Liu, S.A., Qu, Y.R., Wang, Z.Z., Li, M.L., Yang, C., Li, S.G. (2022) The fate of subducting carbon tracked by Mg and Zn isotopes: A review and new perspectives. Earth-Science Reviews 228, 104010. https://doi.org/10.1016/j.earscirev.2022.104010
Show in context

Partial melting processes can explain δ66Zn variations in arc lavas (Huang et al., 2018), whereas it has been argued that δ66Zn variations in some mantle-derived basaltic rocks reflect contributions from distinct sources, including recycled carbonate (Beunon et al., 2020; Liu et al., 2022).
View in article


McCoy-West, A.J., Fitton, J.G., Pons, M.L., Inglis, E.C., Williams, H.M. (2018) The Fe and Zn isotope composition of deep mantle source regions: Insights from Baffin Island picrites. Geochimica et Cosmochimica Acta 238, 542–562. https://doi.org/10.1016/j.gca.2018.07.021
Show in context

Subsequently, peridotites and komatiites have been shown to have indistinguishable Zn isotopic compositions (+0.16 ± 0.06 ‰) and so were interpreted to represent a consistent δ66ZnBSE (Sossi et al., 2018), with slightly more elevated estimates also reported from modern and ancient mantle melts (+0.20 ±0.03 ‰; McCoy-West et al., 2018; Doucet et al., 2020), with a similar value established for depleted mid-ocean ridge basalt (MORB) mantle (δ66ZnDMM = +0.20 ± 0.05 ‰; Wang et al., 2017).
View in article
Data sources are given in Figure S-1, with Baffin Island bulk rock picrite data and the magmatic olivine compositions (assuming a possible range of olivine MgO contents) from McCoy-West et al. (2018).
View in article
An alternative source for low δ66Zn is accumulation of olivine with isotopically light Zn (McCoy-West et al., 2018; Sossi et al., 2018).
View in article
In this sense, high degree partial melts from ancient (komatiites) and modern (boninites, MORB) magmatism conform to the value determined by Sossi et al. (2018) of +0.16 ± 0.06 ‰, consistent with BSE being perhaps up to +0.2 ‰ (e.g., McCoy-West et al., 2018; Doucet et al., 2020).
View in article


Paniello, R.C., Day, J.M.D., Moynier, F. (2012) Zinc isotopic evidence for the origin of the Moon. Nature 490, 376–379. https://doi.org/10.1038/nature11507
Show in context

Zinc stable isotopes have utility in the study of basaltic rocks, from the examination of contributions of potential mantle and crustal reservoirs with distinct 66Zn/64Zn ratios (e.g., Beunon et al., 2020), to identification of volatile element loss in planets (e.g., Paniello et al., 2012).
View in article
Martian shergottite and nakhlite data (shown as points and as a field) are from Paniello et al. (2012).
View in article
It has been shown that basaltic rocks from the Moon have δ66Zn values ∼1 ‰ heavier and Zn abundances >20 times less than terrestrial basalts, interpreted to reflect volatile loss during lunar formation (Paniello et al., 2012).
View in article
Paniello et al. (2012) also presented Zn isotope data for martian meteorites (Fig. 1).
View in article


Shervais, J.W., Reagan, M.K., Godard, M., Prytulak, J., Ryan, J.G., et al. (2021) Magmatic Response to Subduction Initiation, Part II: Boninites and Related Rocks of the Izu-Bonin Arc From IODP Expedition 352. Geochemistry, Geophysics, Geosystems 22, e2020GC009093. https://doi.org/10.1029/2020GC009093
Show in context

Partial melting occurs within shallow forearc regions during release of volatiles from the subducting slab into the mantle wedge and drives boninite magmatism (e.g., Ishizuka et al., 2011, 2014; Shervais et al., 2021).
View in article
The fluid-assisted melting of such shallow depleted residues within the melting region of the nascent Izu-Bonin-Mariana subduction system in the Eocene is widely accepted as the cause of boninite magmatism at that time (Ishizuka et al., 2011; Shervais et al., 2021).
View in article
Shervais et al. (2021).
View in article


Sossi, P.A., Nebel, O., O’Neill, H.S.C., Moynier, F. (2018) Zinc isotope composition of the Earth and its behaviour during planetary accretion. Chemical Geology 477, 73–84. https://doi.org/10.1016/j.chemgeo.2017.12.006
Show in context

Subsequently, peridotites and komatiites have been shown to have indistinguishable Zn isotopic compositions (+0.16 ± 0.06 ‰) and so were interpreted to represent a consistent δ66ZnBSE (Sossi et al., 2018), with slightly more elevated estimates also reported from modern and ancient mantle melts (+0.20 ±0.03 ‰; McCoy-West et al., 2018; Doucet et al., 2020), with a similar value established for depleted mid-ocean ridge basalt (MORB) mantle (δ66ZnDMM = +0.20 ± 0.05 ‰; Wang et al., 2017).
View in article
An alternative source for low δ66Zn is accumulation of olivine with isotopically light Zn (McCoy-West et al., 2018; Sossi et al., 2018).
View in article
The first assumes a fertile mantle garnet lherzolite composition (green lines) similar to that considered previously (e.g., Sossi et al., 2018); the second is a refractory mantle spinel harzburgite (red lines), and the third is a metasomatised mantle composition (blue lines) (Fig. 2).
View in article
Sossi et al. (2018); Herzberg (1992).
View in article
Shown in upper panel are the estimated bulk silicate Earth (BSE) average (solid line) and standard deviation (shaded regions) values from Chen et al. (2013) and Sossi et al. (2018).
View in article
In this sense, high degree partial melts from ancient (komatiites) and modern (boninites, MORB) magmatism conform to the value determined by Sossi et al. (2018) of +0.16 ± 0.06 ‰, consistent with BSE being perhaps up to +0.2 ‰ (e.g., McCoy-West et al., 2018; Doucet et al., 2020).
View in article


Van der Laan, S.R., Flower, M.J.F., Koster van Groos, A.F. (1989) Experimental evidence for the origin of boninites: near-liquidus phase relations to 7.5 kbar. In: Crawford, A.J. (Ed.) Boninites and Related Rocks. Unwin Hyman, London, 112–147.
Show in context

Such melting produces high MgO (>8 wt. %), SiO2 (52–63 wt. %), and low TiO2 (<0.5 wt. %) melts distinct from komatiites, MORB, OIB or any other type of basaltic rock (e.g., Crawford et al., 1989), and which can be reproduced in melting experiments mimicking hydrous melting conditions of refractory harzburgite-like lithologies (Van der Laan et al., 1989; Falloon and Danyushevsky, 2000).
View in article


Wang, Z.-Z., Liu, S.-A., Liu, J., Huang, J., Xiao, Y., Chu, Z.-Y., Zhao, X.-M., Tang, L. (2017) Zinc isotope fractionation during mantle melting and constraints on the Zn isotope composition of Earth’s upper mantle. Geochimica et Cosmochimica Acta 198, 151–167. https://doi.org/10.1016/j.gca.2016.11.014
Show in context

Studies of basaltic magmatic differentiation series have demonstrated that Zn isotopes are not substantially modified (<0.1 ‰) during fractional crystallisation processes (Chen et al., 2013), but that significant fractionation (≥0.1 ‰) in δ66Zn is likely to occur during mantle partial melting (Wang et al., 2017).
View in article
Subsequently, peridotites and komatiites have been shown to have indistinguishable Zn isotopic compositions (+0.16 ± 0.06 ‰) and so were interpreted to represent a consistent δ66ZnBSE (Sossi et al., 2018), with slightly more elevated estimates also reported from modern and ancient mantle melts (+0.20 ±0.03 ‰; McCoy-West et al., 2018; Doucet et al., 2020), with a similar value established for depleted mid-ocean ridge basalt (MORB) mantle (δ66ZnDMM = +0.20 ± 0.05 ‰; Wang et al., 2017).
View in article
Observations of Zn isotope fractionation during partial melting (Wang et al., 2017; Huang et al., 2018) leave several outstanding questions prior to utilisation of Zn isotopes as tracers of distinct mantle and crustal reservoir contributions, and for estimating the BSE composition, particularly how Zn isotope variations differ in different melting regimes, such as during adiabatic decompression and during water-assisted mantle partial melting.
View in article
These rocks should therefore have low Zn and relatively isotopically light δ66Zn due to their origin from refractory mantle similar to harzburgite (∼+0.16 ‰; Doucet et al., 2016; Wang et al., 2017).
View in article
These results agree well with the limited range in δ66Zn reported for MORB from the Atlantic and Indian oceans (Fig. S-1; Wang et al., 2017), although the Zn abundances span a wider range.
View in article
Doucet et al. (2016) and Wang et al. (2017) presented data for refractory harzburgites from a variety of continental lithospheric mantle localities for δ66Zn, showing that they have generally similar Zn abundances to forearc harzburgites (∼40 ppm) and lower δ66Zn (∼0.16 ‰) than lherzolites (∼0.28 ‰; Fig. 1).
View in article
Furthermore, analyses of minerals from spinel lherzolites indicate that olivine and orthopyroxene have lower δ66Zn than clinopyroxene, or the main carrier of Zn, spinel, which is characterised by heavy δ66Zn (Wang et al., 2017).
View in article
This study; Wang et al. (2017).
View in article
Chen et al. (2013); Wang et al. (2017).
View in article


Wang, Z.-Z., Liu, S.-A., Chen, L.-H., Li, S.-G., Zeng, G. (2018) Compositional transition in natural alkaline lavas through silica-undersaturated melt–lithosphere interaction. Geology 46, 771–774. https://doi.org/10.1130/G45145.1
Show in context

Conversely, low-degree partial melts like alkali basalts, basanites and nephelinites from eastern China (Wang et al., 2018) require fertile melt compositions, but their high Zn abundances also permit contributions from hybridised metasomatised peridotite mantle sources.
View in article
Wang et al. (2018).
View in article
Wang et al. (2018).
View in article
Wang et al. (2018).
View in article



top

Supplementary Information

Abstract | Introduction | Methods and Results | Discussion | Acknowledgements | References | Supplementary Information


The Supplementary Information includes:
  • Methods
  • Partial Melting Estimates
  • Partial Melting Model
  • Table S1
  • Figure S1
  • Supplementary References
    •

    Download the Supplementary Information (PDF).
    top

    Figures



    Figure 1 Bulk rock (a) Zn abundance and (b) isotopic composition as a function of MgO content for basaltic rocks. Data sources are given in Figure S-1, with Baffin Island bulk rock picrite data and the magmatic olivine compositions (assuming a possible range of olivine MgO contents) from McCoy-West et al. (2018)

    McCoy-West, A.J., Fitton, J.G., Pons, M.L., Inglis, E.C., Williams, H.M. (2018) The Fe and Zn isotope composition of deep mantle source regions: Insights from Baffin Island picrites. Geochimica et Cosmochimica Acta 238, 542–562. https://doi.org/10.1016/j.gca.2018.07.021

    . Martian shergottite and nakhlite data (shown as points and as a field) are from Paniello et al. (2012)

    Paniello, R.C., Day, J.M.D., Moynier, F. (2012) Zinc isotopic evidence for the origin of the Moon. Nature 490, 376–379. https://doi.org/10.1038/nature11507

    . New MORB data presented here (solid circles) are distinguished from published data (solid stars).
    Back to article


    Figure 2 (a) Zinc isotopic composition and (b) Zn abundance versus estimated degree of partial melting for basaltic rocks (Table 2) versus models for partial melting of harzburgite, lherzolite and metasomatised peridotite, with dots conforming to 1 % increments of melting. Shown in upper panel are the estimated bulk silicate Earth (BSE) average (solid line) and standard deviation (shaded regions) values from Chen et al. (2013)

    Chen, H., Savage, P.S., Teng, F.-Z., Helz, R.T., Moynier, F. (2013) Zinc isotope fractionation during magmatic differentiation and the isotopic composition of the bulk Earth. Earth and Planetary Science Letters 369–370, 32–42. https://doi.org/10.1016/j.epsl.2013.02.037

    and Sossi et al. (2018)

    Sossi, P.A., Nebel, O., O’Neill, H.S.C., Moynier, F. (2018) Zinc isotope composition of the Earth and its behaviour during planetary accretion. Chemical Geology 477, 73–84. https://doi.org/10.1016/j.chemgeo.2017.12.006

    .
    Back to article