TY - JOUR

T1 - Calcium isotopic fractionation in mantle peridotites by melting and metasomatism and Ca isotope composition of the Bulk Silicate Earth

AU - Kang, Jin Ting

AU - Ionov, Dmitri A.

AU - Liu, Fang

AU - Zhang, Chen Lei

AU - Golovin, Alexander V.

AU - Qin, Li Ping

AU - Zhang, Zhao Feng

AU - Huang, Fang

N1 - Publisher Copyright: © 2017 Elsevier B.V.

PY - 2017/9/15

Y1 - 2017/9/15

N2 - To better constrain the Ca isotopic composition of the Bulk Silicate Earth (BSE) and explore the Ca isotope fractionation in the mantle, we determined the Ca isotopic composition of 28 peridotite xenoliths from Mongolia, southern Siberia and the Siberian craton. The samples are divided in three chemical groups: (1) fertile, unmetasomatized lherzolites (3.7–4.7 wt.% Al2O3); (2) moderately melt-depleted peridotites (1.3–3.0 wt.% Al2O3) with no or very limited metasomatism (LREE-depleted cpx); (3) strongly metasomatized peridotites (LREE-enriched cpx and bulk rock) further divided in subgroups 3a (harzburgites, 0.1–1.0% Al2O3) and 3b (fertile lherzolites, 3.9–4.3% Al2O3). In Group 1, δ44/40Ca of fertile spinel and garnet peridotites, which experienced little or no melting and metasomatism, show a limited variation from 0.90 to 0.99‰ (relative to SRM 915a) and an average of 0.94 ± 0.05‰ (2SD, n=14), which defines the Ca isotopic composition of the BSE. In Group 2, the δ44/40Ca is the highest for three rocks with the lowest Al2O3, i.e. the greatest melt extraction degrees (average 1.06±0.04‰, i.e. ∼0.1‰ heavier than the BSE estimate). Simple modeling of modal melting shows that partial melting of the BSE with 103ln⁡αperidotite-melt ranging from 0.10 to 0.25 can explain the Group 2 data. By contrast, δ44/40Ca in eight out of nine metasomatized Group 3 peridotites are lower than the BSE estimate. The Group 3a harzburgites show the greatest δ44/40Ca variation range (0.25–0.96‰), with δ44/40Ca positively correlated with CaO and negatively correlated with Ce/Eu. Chemical evidence suggests that the residual, melt-depleted, low-Ca protoliths of the Group 3a harzburgites were metasomatized, likely by carbonate-rich melts/fluids. We argue that such fluids may have low (≤0.25‰) δ44/40Ca either because they contain recycled crustal components or because Ca isotopes, similar to trace elements and their ratios, may be fractionated by kinetic and/or chromatographic effects of melt percolation in the mantle. The δ44/40Ca in Group 3b lherzolites (0.83–0.89‰) are lower than in the BSE as well, but the effects of metasomatism on δ44/40Ca are smaller, possibly because of the high Ca contents in their protoliths and/or smaller δ44/40Ca differences between the protoliths and metasomatic agents. The BSE estimates based on fertile peridotites in this study fall in the δ44/40Ca ranges for oceanic and continental basalts, various meteorites (achondrites; carbonaceous, ordinary and enstatite chondrites), Mars, and the Moon. These results provide benchmarks for the application of Ca isotopes to planet formation, mantle evolution, and crustal recycling.

AB - To better constrain the Ca isotopic composition of the Bulk Silicate Earth (BSE) and explore the Ca isotope fractionation in the mantle, we determined the Ca isotopic composition of 28 peridotite xenoliths from Mongolia, southern Siberia and the Siberian craton. The samples are divided in three chemical groups: (1) fertile, unmetasomatized lherzolites (3.7–4.7 wt.% Al2O3); (2) moderately melt-depleted peridotites (1.3–3.0 wt.% Al2O3) with no or very limited metasomatism (LREE-depleted cpx); (3) strongly metasomatized peridotites (LREE-enriched cpx and bulk rock) further divided in subgroups 3a (harzburgites, 0.1–1.0% Al2O3) and 3b (fertile lherzolites, 3.9–4.3% Al2O3). In Group 1, δ44/40Ca of fertile spinel and garnet peridotites, which experienced little or no melting and metasomatism, show a limited variation from 0.90 to 0.99‰ (relative to SRM 915a) and an average of 0.94 ± 0.05‰ (2SD, n=14), which defines the Ca isotopic composition of the BSE. In Group 2, the δ44/40Ca is the highest for three rocks with the lowest Al2O3, i.e. the greatest melt extraction degrees (average 1.06±0.04‰, i.e. ∼0.1‰ heavier than the BSE estimate). Simple modeling of modal melting shows that partial melting of the BSE with 103ln⁡αperidotite-melt ranging from 0.10 to 0.25 can explain the Group 2 data. By contrast, δ44/40Ca in eight out of nine metasomatized Group 3 peridotites are lower than the BSE estimate. The Group 3a harzburgites show the greatest δ44/40Ca variation range (0.25–0.96‰), with δ44/40Ca positively correlated with CaO and negatively correlated with Ce/Eu. Chemical evidence suggests that the residual, melt-depleted, low-Ca protoliths of the Group 3a harzburgites were metasomatized, likely by carbonate-rich melts/fluids. We argue that such fluids may have low (≤0.25‰) δ44/40Ca either because they contain recycled crustal components or because Ca isotopes, similar to trace elements and their ratios, may be fractionated by kinetic and/or chromatographic effects of melt percolation in the mantle. The δ44/40Ca in Group 3b lherzolites (0.83–0.89‰) are lower than in the BSE as well, but the effects of metasomatism on δ44/40Ca are smaller, possibly because of the high Ca contents in their protoliths and/or smaller δ44/40Ca differences between the protoliths and metasomatic agents. The BSE estimates based on fertile peridotites in this study fall in the δ44/40Ca ranges for oceanic and continental basalts, various meteorites (achondrites; carbonaceous, ordinary and enstatite chondrites), Mars, and the Moon. These results provide benchmarks for the application of Ca isotopes to planet formation, mantle evolution, and crustal recycling.

KW - Bulk Silicate Earth

KW - Ca isotopes

KW - lithospheric mantle

KW - metasomatism

KW - partial melting

KW - peridotite xenolith

KW - ROCK

KW - LITHOSPHERIC MANTLE

KW - CENTRAL SIBERIA

KW - XENOLITHS

KW - UDACHNAYA KIMBERLITE

KW - ABUNDANCES

KW - ORTHO-PYROXENE

KW - TRACE-ELEMENT COMPOSITIONS

KW - CHEMICAL-COMPOSITION

KW - GEOCHEMISTRY

UR - http://www.scopus.com/inward/record.url?scp=85030448759&partnerID=8YFLogxK

U2 - 10.1016/j.epsl.2017.05.035

DO - 10.1016/j.epsl.2017.05.035

M3 - Article

AN - SCOPUS:85030448759

VL - 474

SP - 128

EP - 137

JO - Earth and Planetary Science Letters

JF - Earth and Planetary Science Letters

SN - 0012-821X

ER -