A pristine low-Ti cumulate source for Chang’e 5 basalts revealed by Sr-Nd-Hf isotopes
Affiliations | Corresponding Author | Cite as | Funding information
Published by the European Association of Geochemistry
under Creative Commons License CC BY-NC-ND 4.0
Keywords: Chang’e-5 basalts, lunar magma ocean cumulate, lunar mantle source, Sr-Nd-Hf isotopes
- Share this article
-
Article views:80Cumulative count of HTML views and PDF downloads.
- Download Citation
- Rights & Permissions
top
Abstract
Figures
 Figure 1 207Pb/206Pb versus 204Pb/206Pb diagram for the basalt clast CE5C0000YJYX052. The red circles, grey triangles, blue squares represent lunar Pb isotopic data from phosphate grains, K-feldspar grains, K-rich glass pockets and Zr-rich mineral-bearing areas. The red circles are used to define the red isochron, and the blue line defined by the blue squares indicates terrestrial Pb contamination. The grey triangles represent those affected by terrestrial contamination. MSWD, mean squared weighted deviation. All the error crosses are at 2 sigma. A detailed description of the method can be found in the Supplementary Information. |  Figure 2 A comparison of radiogenic Sr–Nd–Hf isotopic compositions between CE5 basalt clasts and Apollo mare basalts. (a) 87Rb/86Sr values of mantle sources were calculated assuming a single-stage differentiated evolution from 4558 Ma with a lunar initial 87Sr/86Sr of 0.69903 following Elardo et al. (2014). (b-c) The εNd and εHf values of mantle sources were calculated assuming source formation at the time of LMO crystallisation (4.42 Ga) with a chondritic composition (Nyquist et al., 1995; Elardo et al., 2014). The grey bars represent in situ Sr and Nd isotopic compositions (Tian et al., 2021). Isotopic data: Elardo et al. (2014) and references therein, Sprung et al. (2013), Beard et al. (1998). |  Figure 3 (176Lu/177Hf)source and (147Sm/144Nd)source of lunar sample sources with Eu/Eu*, assuming LMO differentiation at 4.42 Ga. (a, c, e) Crystallisation mode of Snyder et al. (1992) with different initial LMO compositions. (b, d, f) Eu/Eu* values associated with their (147Sm/144Nd)source. Sequences 1 to 5 are presented in Figure S-7. The labels“Seq 1 + 2 & Seq 4” and “Seq 1 + 2 & Seq 5”indicate hybrid sources of early-forming Ol–Opx (Seq 1 + 2) and late-stage Cpx-bearing (Seq 4) or Cpx–Ilm-bearing (Seq 5) cumulates comprising various TIRL in the Ol–Opx cumulates. The black dotted curves represent mixtures of KREEP and Seq 3 cumulate. Data source: Sprung et al. (2013) and Unruh et al. (1984). |  Figure 4 (a) Modelling of partial melting and fractional crystallisation. Partial melts were calculated based on 86 PCS cumulate with 2 % TIRL. The yellow area represents 2 % to 4 % partial melts. The purple and orange areas represent 39 to ∼83 % and 71 to ∼93 % of the degree of fractional crystallisation to match the LREE and HREE for the calculated parental magma and whole rock, respectively. (b) REE patterns for minerals and glasses in CE5C0000YJYX052. The basaltic glass REE is calculated based on the highest Fe pigeonite (HF Pig) (McKay et al., 1991) or the felsic glass (Fg) (Shearer et al., 2001). The chondritic and KREEP data are from Anders and Grevesse (1989) and Warren (1989), respectively. |
| Figure 1 | Figure 2 | Figure 3 | Figure 4 |
top
Introduction
The Procellarum KREEP Terrane (PKT) is a unique geochemical crustal province on the Moon that features distinctive geochemical and thermal evolution. The PKT is characterised by enrichments in Th and other incompatible elements, as well as an extended volcanic history. Based on the Apollo and Luna samples, most previous studies have linked regional incompatible element enrichments to pre-existing KREEP components (Shearer et al., 2006
Shearer, C.K., Hess, P.C., Wieczorek, M.A., Pritchard, M.E., Parmentier, E.M., Borg, L.E., Longhi, J., Elkins-Tanton, L.T., Neal, C.R., Antonenko, I., Canup, R.M., Halliday, A.N., Grove, T.L., Hager, B.H., Lee, D.-C., Wiechert, U. (2006) Thermal and magmatic evolution of the Moon. Reviews in Mineralogy and Geochemistry 60, 365–518. https://doi.org/10.2138/rmg.2006.60.4
and references therein), which represent residues after extensive fractional crystallisations (>99 %) of the Lunar Magma Ocean (LMO) (Warren, 1989Warren, P.H. (1989) KREEP: major-element diversity, trace-element uniformity (almost). Workshop on Moon in transition: Apollo 14, KREEP, and evolved lunar rocks. 149–153.
). Furthermore, radiogenic heat from the KREEP has been proposed to induce extensive lunar mantle melting (Haskin et al., 2000Haskin, L.A., Gillis, J.J., Korotev, R.L., Jolliff, B.L. (2000) The materials of the lunar Procellarum KREEP Terrane: A synthesis of data from geomorphological mapping, remote sensing, and sample analyses. Journal of Geophysical Research: Planets 105, 20403–20415. https://doi.org/10.1029/1999je001128
). The Chang’e 5 (CE5) mission returned mare basalts with eruption ages of ∼2.0 Ga, indicating that the volcanic event in the PKT mare region lasted an additional 0.8 to 1.0 Gyr after the youngest lunar volcanic magmatism event (2.8 to 3.0 Ga; Shearer et al., 2023Shearer, C.K., Neal, C.R., Glotch, T.D., Prissel, T.C., Bell, A.S., Fernandes, V.A., Gaddis, L.R., Jolliff, B.L., Laneuville, M., Magna, T., Simon, J. (2023) Magmatic evolution II: A new view of post-differentiation magmatism. Reviews in Mineralogy and Geochemistry 89, 147–206. https://doi.org/10.2138/rmg.2023.89.04
) dated before the CE5 mission. However, recent analyses of CE5 mare basalt clasts have provided opposite conclusions about whether the mantle source was pristine or hybridised (mixing of LMO cumulates or melts at different cooling stages), which is critical for determining the mechanism of this young melting event (Tian et al., 2021Tian, H.-C., Wang, H., Chen, Y., Yang, W., Zhou, Q., Zhang, C., Lin, H.L., Huang, C., Wu, S.T., Jia, L.H., Xu, L., Zhang, D., Li, X.G., Chang, R., Yang, Y.H., Xie, L.W., Zhang, D.P., Zhang, G.L., Yang, S.H., Wu, F.Y. (2021) Non-KREEP origin for Chang’e-5 basalts in the Procellarum KREEP Terrane. Nature 600, 59–63. https://doi.org/10.1038/s41586-021-04119-5
; Su et al., 2022Su, B., Yuan, J., Chen, Y., Yang, W., Mitchell, R.N., Hui, H., Wang, H., Tian, H., Li, X.H., Wu, F.Y. (2022) Fusible mantle cumulates trigger young mare volcanism on the Cooling Moon. Science Advances 8, eabn2103. https://doi.org/10.1126/sciadv.abn2103
; Zong et al., 2022Zong, K., Wang, Z., Li, J., He, Q., Li, Y., Becker, H., Zhang, W., Hu, Z., He, T., Cao, K., She, Z., Wu, X., Xiao, L., Liu, Y. (2022) Bulk compositions of the Chang’E-5 lunar soil: Insights into chemical homogeneity, exotic addition, and origin of landing site basalts. Geochimica et Cosmochimica Acta 335, 284–296. https://doi.org/10.1016/j.gca.2022.06.037
; Jiang et al., 2023Jiang, Y., Kang, J., Liao, S., Elardo, S.M., Zong, K., Wang, S., Nie, C., Li, P., Yin, Z., Huang, F., Hsu, W. (2023) Fe and Mg Isotope Compositions Indicate a Hybrid Mantle Source for Young Chang’E 5 Mare Basalts. The Astrophysical Journal Letters 945, L26. https://doi.org/10.3847/2041-8213/acbd31
). Thus, identifying the mantle source for the CE5 basalts is key to solving this issue.Radiogenic Sr-Nd-Hf isotopic compositions are sensitive to mantle sources that have undergone differentiation. In particular, Hf–Nd isotopic compositions can provide insight into the mantle source for mare basalts, because Hf–Nd isotopic covariations are sensitive to different mineral assemblages and source hybridisations (Beard et al., 1998
Beard, B.L., Taylor, L.A., Scherer, E.E., Johnson, C.M., Snyder, G.A. (1998) The source region and melting mineralogy of high-titanium and low-titanium lunar basalts deduced from Lu-Hf isotope data. Geochimica et Cosmochimica Acta 62, 525–544. https://doi.org/10.1016/S0016-7037(97)00373-6
; Sprung et al., 2013Sprung, P., Kleine, T., Scherer, E.E. (2013) Isotopic evidence for chondritic Lu/Hf and Sm/Nd of the Moon. Earth and Planetary Science Letters 380, 77–87. https://doi.org/10.1016/j.epsl.2013.08.018
). In situ Sr isotopic composition of plagioclase and Nd isotopic composition of merrillite suggests a depleted mantle source for the CE5 basalts (Tian et al., 2021Tian, H.-C., Wang, H., Chen, Y., Yang, W., Zhou, Q., Zhang, C., Lin, H.L., Huang, C., Wu, S.T., Jia, L.H., Xu, L., Zhang, D., Li, X.G., Chang, R., Yang, Y.H., Xie, L.W., Zhang, D.P., Zhang, G.L., Yang, S.H., Wu, F.Y. (2021) Non-KREEP origin for Chang’e-5 basalts in the Procellarum KREEP Terrane. Nature 600, 59–63. https://doi.org/10.1038/s41586-021-04119-5
). However, the Nd-Hf isotopic compositions of lunar materials can be altered by cosmic ray irradiation (Nyquist et al., 1995Nyquist, L.E., Wiesmann, H., Bansal, B., Shih, C.Y., Keith, J.E., Harper, C.L. (1995) 146Sm-142Nd formation interval for the lunar mantle. Geochimica et Cosmochimica Acta 59, 2817–2837. https://doi.org/10.1016/0016-7037(95)00175-y
; Sprung et al., 2010;Sprung, P., Kleine, T., Scherer, E.E. (2013) Isotopic evidence for chondritic Lu/Hf and Sm/Nd of the Moon. Earth and Planetary Science Letters 380, 77–87. https://doi.org/10.1016/j.epsl.2013.08.018
Carlson et al., 2014Carlson, R.W., Borg, L.E., Gaffney, A.M., Boyet, M. (2014) Rb-Sr, Sm-Nd and Lu-Hf isotope systematics of the lunar Mg-suite: the age of the lunar crust and its relation to the time of Moon formation. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 372, 20130246. https://doi.org/10.1098/rsta.2013.0246
). Thus, whole rock Sr–Nd–Hf isotopic compositions with correlations for neutron capture could provide a critical constraint on the source nature of the CE5 basalts.Here, to solve the above issues, we conduct an integrated study of whole rock Sr–Nd–Hf isotopic compositions (with corrections for thermal neutron capture effects), along with mineralogy, chemical composition, and SHRIMP Pb–Pb isotope dating on one porphyritic basalt clast from the CE5 mission (CE5C0000YJYX052, 72.1 mg, Fig. S-1).
top
Methods and Results
The detailed methods are presented in the Supplementary Information. Four crushed chips (18.5 mg) were prepared for mineralogical and chronological investigations. These four sections display subophitic mineral assemblages, with similar modal mineral abundances of pyroxene and plagioclase (Table S-1). However, the ilmenite abundances in individual thin sections largely varied from 2.05 vol. % to 15.44 vol. % using EPMA (Electron probe microanalyser) at the University of Science and Technology of China (USTC), China, which corresponds to 3.58 wt. % to 16.26 wt. % TiO2 in the whole rock (Table S-2). Pyroxenes in the four sections had similar compositions and were compositionally zoned with Mg-rich cores and Fe-rich rims (Figs. S-2, 3, Table S-4). The plagioclase grains were either homogeneous or weakly zoned (An77-88; Fig. S-2, Table S-5). All four sections were enriched in mesostasis consisting of fayalite (Fa95-99), K-feldspar (Or88-94), troilite with metallic iron blebs, phosphate, baddeleyite, and Si-K-rich and Fe-rich glasses (Fig. S-2, Tables S-5–11). The dissolved whole rock powder (∼30 mg) gave 3.64 wt. % TiO2 using ICP-OES at USTC (Table S-3). Moreover, this whole-rock powder exhibited KREEP-like features with incompatible element enrichment (e.g., K, Th and REEs determined by ICP-MS at USTC; Table S-3), in which light REE concentrations were significantly greater than those in the Apollo and Luna mare basalts (Fig. S-4c).
Figure 1 207Pb/206Pb versus 204Pb/206Pb diagram for the basalt clast CE5C0000YJYX052. The red circles, grey triangles, blue squares represent lunar Pb isotopic data from phosphate grains, K-feldspar grains, K-rich glass pockets and Zr-rich mineral-bearing areas. The red circles are used to define the red isochron, and the blue line defined by the blue squares indicates terrestrial Pb contamination. The grey triangles represent those affected by terrestrial contamination. MSWD, mean squared weighted deviation. All the error crosses are at 2 sigma. A detailed description of the method can be found in the Supplementary Information.
The lead isotopic ratios for phosphate, K-feldspar, K-rich glass and Zr-rich mineral-bearing areas in four sections, determined using SHRIMP, are presented in Table S-12. The data yield an age of 1962 ± 90 Ma (MSWD = 1.6, n = 9; Fig. 1), in accordance with 1963 ± 57 Ma (Che et al., 2021
Che, X., Nemchin, A., Liu, D., Long, T., Wang, C., Norman, M.D., Joy, K.H., Tartese, R., Head, J., Jolliff, B., Snape, J.F., Neal, C.R., Whitehouse, M.J., Crow, C., Benedix, G., Jourdan, F., Yang, Z., Yang, C., Liu, J., Xie, S., Bao, Z., Fan, R., Li, D., Li, Z., Webb, S.G. (2021) Age and composition of young basalts on the Moon, measured from samples returned by Chang’e-5. Science 374, 887–890. https://doi.org/10.1126/science.abl7957
) and 2030 ± 4 Ma (Li et al., 2021Li, Q.-L., Zhou, Q., Liu, Y., Xiao, Z., Lin, Y., Li, J.H., Ma, H.X., Tang, G.Q., Guo, S., Yuan, J.Y., Li, J., Wu, F.Y., Ouyang, Z., Li, C., Li, X.H. (2021) Two-billion-year-old volcanism on the Moon from Chang’e-5 basalts. Nature 600, 54–58. https://doi.org/10.1038/s41586-021-04100-2
) for other CE5 clasts. Extensive thermal neutron capture effects were recorded by ε180Hf = −2.01 ± 0.32 and ε149Sm = −15.4 ± 2.1, as determined by solution MC-ICP-MS and TIMS, respectively (Table S-13). After corrections for such effects, the whole rock data exhibit initial εNd(1.96 Ga) = 7.49 ± 0.31, εHf(1.96 Ga) = 42.36 ± 0.54, and unaffected (87Sr/86Sr)1.96 Ga = 0.699666 ± 0.000008, respectively (Table S-13).top
An Apollo-Like Pristine Low-Ti Cumulate Source Revealed by Nd-Hf Isotopic Compositions
Considering the possibility of different mantle sources, we first compared CE5C0000YJYX052 with other CE5 samples, all of which were collected in a limited region. Several lines of evidence suggest a common source. First, CE5 basalt clasts give indistinguishable crystallisation ages of ∼2.0 Ga. Second, its whole rock Sr and Nd isotopic compositions are comparable to in situ analysed data within errors (Fig. 2; Tian et al., 2021
Tian, H.-C., Wang, H., Chen, Y., Yang, W., Zhou, Q., Zhang, C., Lin, H.L., Huang, C., Wu, S.T., Jia, L.H., Xu, L., Zhang, D., Li, X.G., Chang, R., Yang, Y.H., Xie, L.W., Zhang, D.P., Zhang, G.L., Yang, S.H., Wu, F.Y. (2021) Non-KREEP origin for Chang’e-5 basalts in the Procellarum KREEP Terrane. Nature 600, 59–63. https://doi.org/10.1038/s41586-021-04119-5
). Third, major minerals from CE5C0000YJYX052 and other CE5 clasts are compositionally similar (Fig. S-3). Lastly, the major element compositions of all CE5 basalt clasts are correlated with the Mg# values according to the low-Ti basaltic magma evolution (Fig. S-4). Thus, we propose that the investigated basalt has a common source to others; its whole rock Sr–Nd–Hf isotopic compositions could be a representation of such a young magma source.Figure 2 A comparison of radiogenic Sr–Nd–Hf isotopic compositions between CE5 basalt clasts and Apollo mare basalts. (a) 87Rb/86Sr values of mantle sources were calculated assuming a single-stage differentiated evolution from 4558 Ma with a lunar initial 87Sr/86Sr of 0.69903 following Elardo et al. (2014)
Elardo, S.M., Shearer, C.K., Fagan, A.L., Borg, L.E., Gaffney, A.M., Burger, P.V., Neal, C.R., Fernandes, V.A., McCubbin, F.M. (2014) The origin of young mare basalts inferred from lunar meteorites Northwest Africa 4734, 032, and LaPaz Icefield 02205. Meteoritics and Planetary Science 49, 261–291. https://doi.org/10.1111/maps.12239
. (b-c) The εNd and εHf values of mantle sources were calculated assuming source formation at the time of LMO crystallisation (4.42 Ga) with a chondritic composition (Nyquist et al., 1995Nyquist, L.E., Wiesmann, H., Bansal, B., Shih, C.Y., Keith, J.E., Harper, C.L. (1995) 146Sm-142Nd formation interval for the lunar mantle. Geochimica et Cosmochimica Acta 59, 2817–2837. https://doi.org/10.1016/0016-7037(95)00175-y
; Elardo et al., 2014Elardo, S.M., Shearer, C.K., Fagan, A.L., Borg, L.E., Gaffney, A.M., Burger, P.V., Neal, C.R., Fernandes, V.A., McCubbin, F.M. (2014) The origin of young mare basalts inferred from lunar meteorites Northwest Africa 4734, 032, and LaPaz Icefield 02205. Meteoritics and Planetary Science 49, 261–291. https://doi.org/10.1111/maps.12239
). The grey bars represent in situ Sr and Nd isotopic compositions (Tian et al., 2021Tian, H.-C., Wang, H., Chen, Y., Yang, W., Zhou, Q., Zhang, C., Lin, H.L., Huang, C., Wu, S.T., Jia, L.H., Xu, L., Zhang, D., Li, X.G., Chang, R., Yang, Y.H., Xie, L.W., Zhang, D.P., Zhang, G.L., Yang, S.H., Wu, F.Y. (2021) Non-KREEP origin for Chang’e-5 basalts in the Procellarum KREEP Terrane. Nature 600, 59–63. https://doi.org/10.1038/s41586-021-04119-5
). Isotopic data: Elardo et al. (2014)Elardo, S.M., Shearer, C.K., Fagan, A.L., Borg, L.E., Gaffney, A.M., Burger, P.V., Neal, C.R., Fernandes, V.A., McCubbin, F.M. (2014) The origin of young mare basalts inferred from lunar meteorites Northwest Africa 4734, 032, and LaPaz Icefield 02205. Meteoritics and Planetary Science 49, 261–291. https://doi.org/10.1111/maps.12239
and references therein, Sprung et al. (2013)Sprung, P., Kleine, T., Scherer, E.E. (2013) Isotopic evidence for chondritic Lu/Hf and Sm/Nd of the Moon. Earth and Planetary Science Letters 380, 77–87. https://doi.org/10.1016/j.epsl.2013.08.018
, Beard et al. (1998)Beard, B.L., Taylor, L.A., Scherer, E.E., Johnson, C.M., Snyder, G.A. (1998) The source region and melting mineralogy of high-titanium and low-titanium lunar basalts deduced from Lu-Hf isotope data. Geochimica et Cosmochimica Acta 62, 525–544. https://doi.org/10.1016/S0016-7037(97)00373-6
.The coupled Nd–Hf isotopic compositions can serve as conclusive evidence for distinguishing mantle sources because the lunar low-Ti basalt source typically has a [Lu/Hf]n ratio (n refers to CI chondrite-normalised value) ∼4 times greater than its [Sm/Nd]n ratio, whereas the high-Ti source has subequal [Lu/Hf]n and [Sm/Nd]n ratios (∼1) (Beard et al., 1998
Beard, B.L., Taylor, L.A., Scherer, E.E., Johnson, C.M., Snyder, G.A. (1998) The source region and melting mineralogy of high-titanium and low-titanium lunar basalts deduced from Lu-Hf isotope data. Geochimica et Cosmochimica Acta 62, 525–544. https://doi.org/10.1016/S0016-7037(97)00373-6
; Sprung et al., 2013Sprung, P., Kleine, T., Scherer, E.E. (2013) Isotopic evidence for chondritic Lu/Hf and Sm/Nd of the Moon. Earth and Planetary Science Letters 380, 77–87. https://doi.org/10.1016/j.epsl.2013.08.018
). In particular, ilmenite (Ilm) has a markedly lower KD[Lu/Hf]/KD[Sm/Nd] (∼0.07) than other accumulated minerals, with KD[Lu/Hf]/KD[Sm/Nd] values of ∼0.7 to ∼4 from the LMO (e.g., olivine (Ol), orthopyroxene (Opx), clinopyroxene (Cpx), pigeonite (Pig), plagioclase (Pl); Sprung et al., 2013Sprung, P., Kleine, T., Scherer, E.E. (2013) Isotopic evidence for chondritic Lu/Hf and Sm/Nd of the Moon. Earth and Planetary Science Letters 380, 77–87. https://doi.org/10.1016/j.epsl.2013.08.018
; Yang et al., 2022Yang, W., Chen, Y., Wang, H., Tian, H.C., Hui, H., Xiao, Z., Wu, S.T., Zhang, D., Zhou, Q., Ma, H.X., Zhang, C., Hu, S., Li, Q.L., Lin, Y., Li, X.H., Wu, F.Y. (2022) Geochemistry of impact glasses in the Chang’e-5 regolith: Constraints on impact melting and the petrogenesis of local basalt. Geochimica et Cosmochimica Acta 335, 183–196. https://doi.org/10.1016/j.gca.2022.08.030
), accounting for the distinct trends of low-Ti and high-Ti basalts (Fig. 2c).The neutron capture effect corrected εNd(t) and εHf(t) values for the investigated CE5 basalt and Apollo low-Ti basalts display excellent positive correlations (R2 = 0.95; Fig. 2c). This correlation can be potentially interpreted as a result of a common LMO low-Ti cumulate with variable amounts of trapped instantaneous residual liquid (TIRL), or a hybrid source that comprises early-stage Ol and Opx cumulates containing various fractions of TIRL with a small amount of late-stage Cpx-bearing cumulates (Beard et al., 1998
Beard, B.L., Taylor, L.A., Scherer, E.E., Johnson, C.M., Snyder, G.A. (1998) The source region and melting mineralogy of high-titanium and low-titanium lunar basalts deduced from Lu-Hf isotope data. Geochimica et Cosmochimica Acta 62, 525–544. https://doi.org/10.1016/S0016-7037(97)00373-6
; Sprung et al., 2013Sprung, P., Kleine, T., Scherer, E.E. (2013) Isotopic evidence for chondritic Lu/Hf and Sm/Nd of the Moon. Earth and Planetary Science Letters 380, 77–87. https://doi.org/10.1016/j.epsl.2013.08.018
). For comparison, the corrected isotopic data are expressed as parent-to-daughter ratios according to their crystallisation ages that would be required for their mantle sources to evolve from the isotopic composition of the LMO, assuming an initial LMO crystallisation age at ∼4.42 Ga (Nyquist et al., 1995Nyquist, L.E., Wiesmann, H., Bansal, B., Shih, C.Y., Keith, J.E., Harper, C.L. (1995) 146Sm-142Nd formation interval for the lunar mantle. Geochimica et Cosmochimica Acta 59, 2817–2837. https://doi.org/10.1016/0016-7037(95)00175-y
). Here, we modelled the (147Sm/144Nd)source, (176Lu/177Hf)source, and Eu anomaly [Eu/Eu* = EuN/(SmN×GdN)1/2, subscript N represents chondrite normalised value] of different cumulate sources that formed during the LMO crystallisation. The crystallising modes followed Snyder et al. (1992)Snyder, G.A., Taylor, L.A., Neal, C.R. (1992) A chemical model for generating the sources of mare basalts: Combined equilibrium and fractional crystallization of the lunar magmasphere. Geochimica et Cosmochimica Acta 56, 3809–3823. https://doi.org/10.1016/0016-7037(92)90172-f
, Elardo et al. (2011)Elardo, S.M., Draper D.S., Shearer Jr. C.K. (2011) Lunar Magma Ocean crystallization revisited: Bulk composition, early cumulate mineralogy, and the source regions of the highlands Mg-suite. Geochimica et Cosmochimica Acta 75, 3024–3045. https://doi.org/10.1016/j.gca.2011.02.033
, Elkins-Tanton et al. (2011)Elkins-Tanton, L.T., Burgess, S., Yin, Q.Z. (2011) The lunar magma ocean: Reconciling the solidification process with lunar petrology and geochronology. Earth and Planetary Science Letters 304, 326–336. https://doi.org/10.1016/j.epsl.2011.02.004
, and Charlier et al. (2018)Charlier, B., Grove, T.L., Namur, O., Holtz, F. (2018) Crystallization of the lunar magma ocean and the primordial mantle-crust differentiation of the Moon. Geochimica et Cosmochimica Acta 234, 50–69. https://doi.org/10.1016/j.gca.2018.05.006
to fully constrain potential effects of different mineral assemblages on their sources (Fig. S-7). Considering that the compositions of the initial LMO have not been fully elucidated, three compositions were investigated, namely, a chondritic 147Sm/144Nd = 0.1967 and 176Lu/177Hf = 0.0336, as described by Snyder et al. (1992)Snyder, G.A., Taylor, L.A., Neal, C.R. (1992) A chemical model for generating the sources of mare basalts: Combined equilibrium and fractional crystallization of the lunar magmasphere. Geochimica et Cosmochimica Acta 56, 3809–3823. https://doi.org/10.1016/0016-7037(92)90172-f
; a moderately depleted composition of 147Sm/144Nd = 0.2008 and 176Lu/177Hf = 0.0375 by Johnston et al. (2022)Johnston, S., Brandon, A., McLeod, C., Rankenburg, K., Becker, H., Copeland, P. (2022) Nd isotope variation between the Earth–Moon system and enstatite chondrites. Nature 611, 501–506. https://doi.org/10.1038/s41586-022-05265-0
; and a highly depleted composition of 147Sm/144Nd = 0.2082 and 176Lu/177Hf = 0.0375 by Caro and Bourdon (2010)Caro, G., Bourdon, B. (2010) Non-chondritic Sm/Nd ratio in the terrestrial planets: consequences for the geochemical evolution of the mantle–crust system. Geochimica et Cosmochimica Acta 74, 3333–3349. https://doi.org/10.1016/j.gca.2010.02.025
. Detailed calculations are presented in Table S-14.Firstly, the possibility of a highly depleted LMO proposed by Caro and Bourdon (2010) can be ruled out. In this case, the low-Ti basalt trend cannot be produced by a simple source composing a common cumulate and variable TIRL (Figs. 3e, S-10). Instead, a complicated scenario in which KREEP mixes with a range of different LMO cumulates in just the right proportions is required, which seems to be too coincidental, as also has been mentioned by Sprung et al. (2013)
Sprung, P., Kleine, T., Scherer, E.E. (2013) Isotopic evidence for chondritic Lu/Hf and Sm/Nd of the Moon. Earth and Planetary Science Letters 380, 77–87. https://doi.org/10.1016/j.epsl.2013.08.018
. Addition of a KREEP component would also cause much higher incompatible element abundances than those observed in the low-Ti basalts.Figure 3 (176Lu/177Hf)source and (147Sm/144Nd)source of lunar sample sources with Eu/Eu*, assuming LMO differentiation at 4.42 Ga. (a, c, e) Crystallisation mode of Snyder et al. (1992)
Snyder, G.A., Taylor, L.A., Neal, C.R. (1992) A chemical model for generating the sources of mare basalts: Combined equilibrium and fractional crystallization of the lunar magmasphere. Geochimica et Cosmochimica Acta 56, 3809–3823. https://doi.org/10.1016/0016-7037(92)90172-f
with different initial LMO compositions. (b, d, f) Eu/Eu* values associated with their (147Sm/144Nd)source. Sequences 1 to 5 are presented in Figure S-7. The labels“Seq 1 + 2 & Seq 4” and “Seq 1 + 2 & Seq 5”indicate hybrid sources of early-forming Ol–Opx (Seq 1 + 2) and late-stage Cpx-bearing (Seq 4) or Cpx–Ilm-bearing (Seq 5) cumulates comprising various TIRL in the Ol–Opx cumulates. The black dotted curves represent mixtures of KREEP and Seq 3 cumulate. Data source: Sprung et al. (2013)Sprung, P., Kleine, T., Scherer, E.E. (2013) Isotopic evidence for chondritic Lu/Hf and Sm/Nd of the Moon. Earth and Planetary Science Letters 380, 77–87. https://doi.org/10.1016/j.epsl.2013.08.018
and Unruh et al. (1984)Unruh, D.M., Stille, P., Patchett, P.J., Tatsumoto, M. (1984) Lu-Hf and Sm-Nd evolution in lunar mare basalts. Journal of Geophysical Research 89, B459–B477. https://doi.org/10.1029/jb089is02p0b459
.For the initial LMO with chondritic or moderately depleted compositions, the calculated results are similar: the mantle sources of the low-Ti basalts can be explained by a moderate-stage LMO cumulate (e.g., sequence 3 from Snyder et al. (1992)
Snyder, G.A., Taylor, L.A., Neal, C.R. (1992) A chemical model for generating the sources of mare basalts: Combined equilibrium and fractional crystallization of the lunar magmasphere. Geochimica et Cosmochimica Acta 56, 3809–3823. https://doi.org/10.1016/0016-7037(92)90172-f
, Charlier et al. (2018)Charlier, B., Grove, T.L., Namur, O., Holtz, F. (2018) Crystallization of the lunar magma ocean and the primordial mantle-crust differentiation of the Moon. Geochimica et Cosmochimica Acta 234, 50–69. https://doi.org/10.1016/j.gca.2018.05.006
, and Elkins-Tanton et al. (2011)Elkins-Tanton, L.T., Burgess, S., Yin, Q.Z. (2011) The lunar magma ocean: Reconciling the solidification process with lunar petrology and geochronology. Earth and Planetary Science Letters 304, 326–336. https://doi.org/10.1016/j.epsl.2011.02.004
) with different proportions of TIRL. Taking Snyder’s model as an example, the sequence 3 cumulate consists of 53 % Pl, 25 % Ol, and 22 % Pig at 86 percent crystallised solid (PCS). Meanwhile, upward separation of buoyant Pl leads to limited Pl entrainment in the residual cumulate. An intermediate proportion of TIRL (∼2 %) trapped in sequence 3 could reproduce the [176Lu/177Hf] and [147Sm/144Nd] of the CE5 basalt (black solid curves in Fig. 3), which falls between those for A15 (>2 % TIRL) and A12 (<1.5% TIRL) low-Ti basalts, consistent with the prediction by Hallis et al. (2014)Hallis, L., Anand, M., Strekopytov, S. (2014) Trace-element modelling of mare basalt parental melts: implications for a heterogeneous lunar mantle. Geochimica et Cosmochimica Acta 134, 289–316. https://doi.org/10.1016/j.gca.2014.01.012
.Alternatively, in all models, hybrid sources comprised of early-stage Ol–Opx cumulates with TIRL and a trace amount (<0.5 %) of late-stage Cpx-bearing cumulates could also account for the low-Ti basalt trend (purple and adjacent red dashed curves in Figs. 3, S-8–10). To further evaluate this possibility, Eu/Eu* might be the best candidate due to the unique and ubiquitous negative Eu anomalies in mare basalts. The early-stage Ol and Opx cumulates with variable TIRL values display limited variable Eu anomalies (Figs. 3, S-8–10). Addition of a low proportion (<0.1 %) of late-stage Cpx-bearing cumulate may account for the Nd–Hf isotopic compositions, but has little influence on the Eu anomaly, resulting in hybrid sources deviating from low-Ti basalts (red dashed curves marked by “seq 1 + 2 & 0.01 % seq 4” in Figs. 3, S-8–10). Instead, the sequence 3 cumulate with all Pl floatation and different proportions of TIRL can account for both Nd–Hf isotopic compositions and Eu/Eu* for the low-Ti mare basalts (black solid curves in Figs. 3, S-8, S-9).
As mentioned above, several models of the LMO cumulate source for these CE5 clasts have been proposed. These models have also been assessed, e.g., based on the Snyder’s model, an early-stage Ol cumulate assimilated with 30 % late-stage Cpx-Ilm-rich cumulate (Seq 5) (Jiang et al., 2023
Jiang, Y., Kang, J., Liao, S., Elardo, S.M., Zong, K., Wang, S., Nie, C., Li, P., Yin, Z., Huang, F., Hsu, W. (2023) Fe and Mg Isotope Compositions Indicate a Hybrid Mantle Source for Young Chang’E 5 Mare Basalts. The Astrophysical Journal Letters 945, L26. https://doi.org/10.3847/2041-8213/acbd31
), or an early-stage Ol-Opx cumulate (Seq 1 + 2) assimilated with 20 % late-stage Cpx-Ilm-rich cumulate (Seq 5) (Su et al., 2022Su, B., Yuan, J., Chen, Y., Yang, W., Mitchell, R.N., Hui, H., Wang, H., Tian, H., Li, X.H., Wu, F.Y. (2022) Fusible mantle cumulates trigger young mare volcanism on the Cooling Moon. Science Advances 8, eabn2103. https://doi.org/10.1126/sciadv.abn2103
), or a late-stage Cpx-Ilm dominated cumulate (Seq 4) (Wang et al., 2024Wang, C., Xu, Y.G., Zhang, L., Chen, Z., Xia, X., Lin, M., Guo, F. (2024) A shallow (< 100 km) ilmenite-bearing pyroxenitic source for young lunar volcanism. Earth and Planetary Science Letters 639, 118770. https://doi.org/10.1016/j.epsl.2024.118770
), or a moderate-stage Cpx-rich (40–60 %) cumulate assimilated with 0.3–1.5 % KREEP components (Zong et al., 2022Zong, K., Wang, Z., Li, J., He, Q., Li, Y., Becker, H., Zhang, W., Hu, Z., He, T., Cao, K., She, Z., Wu, X., Xiao, L., Liu, Y. (2022) Bulk compositions of the Chang’E-5 lunar soil: Insights into chemical homogeneity, exotic addition, and origin of landing site basalts. Geochimica et Cosmochimica Acta 335, 284–296. https://doi.org/10.1016/j.gca.2022.06.037
). The simple late-stage cumulate (e.g., seq 4) could be easily ruled out due to that both Cpx and Ilm have lower [Lu/Hf]n/[Sm/Nd]n relative to early-crystallised minerals (Fig. 3a). According to our calculations, even incorporation of a relatively small proportion (>∼2 %) of late-stage cumulate (Cpx-Ilm-rich) will intensely reduce the [Lu/Hf]n of the hybrid source at a given [Sm/Nd]n (Figs. 3a, S-8–9). Given the high enrichments of Lu–Hf and Sm–Nd and subchondritic (Lu/Hf)source (∼0.0187) and (Sm/Nd)source (∼0.1630) in KREEP (Warren, 1989Warren, P.H. (1989) KREEP: major-element diversity, trace-element uniformity (almost). Workshop on Moon in transition: Apollo 14, KREEP, and evolved lunar rocks. 149–153.
), KREEP component incorporation into the cumulate would lead to rapid decreases in both (Lu/Hf)n and (Sm/Nd)n ratios (Fig. 3). Using Snyder’s mode, less than 0.1 % KREEP was needed to reproduce the CE5 basalt data, consistent with its depleted Sr isotopic feature. Therefore, the whole rock Nd–Hf–Sr isotopic compositions and Eu anomalies for CE5C0000YJYX052 clearly point to a pristine LMO cumulate source similar to the Apollo low-Ti basalts.top
Implications for the Petrogenesis of the CE5 Basalt Clasts
Using the sequence 3 cumulate from Snyder’s model, the parental magma of the CE5 basalts (calculated from the most primitive pyroxene with the highest Mg# and partition coefficients of low-Ca Opx at 1300 °C from Yao et al. (2012)
Yao, L., Sun, C., Liang, Y. (2012) A parameterized model for REE distribution between low-Ca pyroxene and basaltic melts with applications to REE partitioning in low-Ca pyroxene along a mantle adiabat and during pyroxenite-derived melt and peridotite interaction. Contributions to Mineralogy and Petrology 164, 261–280. https://doi.org/10.1007/s00410-012-0737-5
) was generated by 2–4 % partial melting, followed by 39–83 % fractional crystallisation (Fig. 4a, Table S-15). The whole rock REE concentration was 2 to 5 times greater than that in the parental magma (Fig. 4a). If REE enrichments were generated only by fractional crystallisation, then a degree of fractional crystallisation of >90 % would be required (particularly LREEs) (Fig. 4a). However, such a high degree would generate andesitic or even rhyolitic residual melt, inconsistent with the major elements.Figure 4 (a) Modelling of partial melting and fractional crystallisation. Partial melts were calculated based on 86 PCS cumulate with 2 % TIRL. The yellow area represents 2 % to 4 % partial melts. The purple and orange areas represent 39 to ∼83 % and 71 to ∼93 % of the degree of fractional crystallisation to match the LREE and HREE for the calculated parental magma and whole rock, respectively. (b) REE patterns for minerals and glasses in CE5C0000YJYX052. The basaltic glass REE is calculated based on the highest Fe pigeonite (HF Pig) (McKay et al., 1991
McKay, G., Wagstaff, J., Le, L. (1991) REE distribution coefficients for pigeonite: constraints on the origin of the mare basalt europium anomaly, III. Mare volcanism and basalt petrogenesis workshop 91-03, 27–28.
) or the felsic glass (Fg) (Shearer et al., 2001Shearer, C.K., Papike, J.J., Spilde, M.N. (2001) Trace-element partitioning between immiscible lunar melts: An example from naturally occurring lunar melt inclusions. American Mineralogist 86, 238–246. https://doi.org/10.2138/am-2001-2-305
). The chondritic and KREEP data are from Anders and Grevesse (1989)Anders, E., Grevesse, N. (1989) Abundances of the elements: Meteoritic and solar. Geochimica et Cosmochimica Acta 53, 197–214. https://doi.org/10.1016/0016-7037(89)90286-X
and Warren (1989)Warren, P.H. (1989) KREEP: major-element diversity, trace-element uniformity (almost). Workshop on Moon in transition: Apollo 14, KREEP, and evolved lunar rocks. 149–153.
, respectively.The mesostasis represents last-stage residual liquids during magmatic evolution, which are highly enriched in incompatible elements. These residual liquids undergo immiscibility splitting into Fe-rich (basaltic glass) and Si-K-rich (felsic glass) conjugate liquids (Shearer et al., 2001
Shearer, C.K., Papike, J.J., Spilde, M.N. (2001) Trace-element partitioning between immiscible lunar melts: An example from naturally occurring lunar melt inclusions. American Mineralogist 86, 238–246. https://doi.org/10.2138/am-2001-2-305
). The analysed felsic glass displayed similar REE concentrations to the whole rock concentrations, whereas the Fe-rich basaltic glass was more enriched (Fig. 4b). Although basaltic glasses were too small to be directly analysed, their pre-existence was evidenced by fayalite, Fe-rich pigeonite, and phosphates directly crystallised from them (Fig. S-2). These minerals display obviously higher REE concentrations than those of in-magma-crystallised olivine and moderate pigeonite (Fig. 4b). Thus, we proposed that the heterogeneous and high proportion of such immiscible liquids trapped in local regions generated the mare basalts with enriched REE patterns and depleted Sr–Nd–Hf isotopic compositions, as observed in CE5 samples studied here.The pristine low-Ti cumulate source for Chang’e 5 basalts leaves the formation mechanism of young volcanics enigmatic. Various models on mantle hybridisations have been proposed to explain the melting mechanism. However as stated above, such mixings cannot accommodate the Sr–Nd–Hf isotopic systematics. We note that the Fe-Mg isotopic compositions of the CE5 basalts observed by Jiang et al. (2023)
Jiang, Y., Kang, J., Liao, S., Elardo, S.M., Zong, K., Wang, S., Nie, C., Li, P., Yin, Z., Huang, F., Hsu, W. (2023) Fe and Mg Isotope Compositions Indicate a Hybrid Mantle Source for Young Chang’E 5 Mare Basalts. The Astrophysical Journal Letters 945, L26. https://doi.org/10.3847/2041-8213/acbd31
are within the range of Apollo low-Ti basalt data, which does not necessitate the introduction of a hybrid source. Although the petrological and mineralogical signature of CE5 basalts can be fulfilled by hybridisation of an early-stage cumulate by a late-stage Cpx- and Ilm- rich cumulate (Su et al., 2022Su, B., Yuan, J., Chen, Y., Yang, W., Mitchell, R.N., Hui, H., Wang, H., Tian, H., Li, X.H., Wu, F.Y. (2022) Fusible mantle cumulates trigger young mare volcanism on the Cooling Moon. Science Advances 8, eabn2103. https://doi.org/10.1126/sciadv.abn2103
; Wang et al., 2024Wang, C., Xu, Y.G., Zhang, L., Chen, Z., Xia, X., Lin, M., Guo, F. (2024) A shallow (< 100 km) ilmenite-bearing pyroxenitic source for young lunar volcanism. Earth and Planetary Science Letters 639, 118770. https://doi.org/10.1016/j.epsl.2024.118770
), the highly differentiated nature and uncertainties of melting conditions (e.g., temperature, pressure, geothermal gradient) will likely impose a large uncertainty on the estimation of the source composition. Therefore, any further model for the CE5 low-Ti basalt formation needs to take into account the radiogenic isotopic results.top
Acknowledgements
We thank the staff of the Chang’e Lunar Exploration Project and the China National Space Administration for providing access to the CE5 sample. We also thank Dr. Richard W. Carlson and two anonymous reviewers, whose comments improved the manuscript, and Francis McCubbin for his work as editor. This study was funded by NSFC (41930216, 42241102, 41973004, 42373010), the Strategic Priority Research Program (B) of the CAS (XDB41000000), USTC Research Funds of the Double First-Class Initiative (YD3410002001), pre-research Projects on Civil Aerospace Technologies (D020204), and Fundamental Research Funds for the Central Universities of China (WK3410000019, KY2080000101). L.Q. also acknowledges the support from the Tencent Foundation through the XPLORER PRIZE.
Editor: Francis McCubbin
top
References
Anders, E., Grevesse, N. (1989) Abundances of the elements: Meteoritic and solar. Geochimica et Cosmochimica Acta 53, 197–214. https://doi.org/10.1016/0016-7037(89)90286-X
Show in context The basaltic glass REE is calculated based on the highest Fe pigeonite (HF Pig) (McKay et al., 1991) or the felsic glass (Fg) (Shearer et al., 2001). The chondritic and KREEP data are from Anders and Grevesse (1989) and Warren (1989), respectively.
View in article
Beard, B.L., Taylor, L.A., Scherer, E.E., Johnson, C.M., Snyder, G.A. (1998) The source region and melting mineralogy of high-titanium and low-titanium lunar basalts deduced from Lu-Hf isotope data. Geochimica et Cosmochimica Acta 62, 525–544. https://doi.org/10.1016/S0016-7037(97)00373-6
Show in context In particular, Hf–Nd isotopic compositions can provide insight into the mantle source for mare basalts, because Hf–Nd isotopic covariations are sensitive to different mineral assemblages and source hybridisations (Beard et al., 1998; Sprung et al., 2013).
View in article
Isotopic data: Elardo et al. (2014) and references therein, Sprung et al. (2013), Beard et al. (1998).
View in article
The coupled Nd–Hf isotopic compositions can serve as conclusive evidence for distinguishing mantle sources because the lunar low-Ti basalt source typically has a [Lu/Hf]n ratio (n refers to CI chondrite-normalised value) ∼4 times greater than its [Sm/Nd]n ratio, whereas the high-Ti source has subequal [Lu/Hf]n and [Sm/Nd]n ratios (∼1) (Beard et al., 1998; Sprung et al., 2013).
View in article
This correlation can be potentially interpreted as a result of a common LMO low-Ti cumulate with variable amounts of trapped instantaneous residual liquid (TIRL), or a hybrid source that comprises early-stage Ol and Opx cumulates containing various fractions of TIRL with a small amount of late-stage Cpx-bearing cumulates (Beard et al., 1998; Sprung et al., 2013).
View in article
Carlson, R.W., Borg, L.E., Gaffney, A.M., Boyet, M. (2014) Rb-Sr, Sm-Nd and Lu-Hf isotope systematics of the lunar Mg-suite: the age of the lunar crust and its relation to the time of Moon formation. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 372, 20130246. https://doi.org/10.1098/rsta.2013.0246
Show in context However, the Nd-Hf isotopic compositions of lunar materials can be altered by cosmic ray irradiation (Nyquist et al., 1995; Sprung et al., 2010; Carlson et al., 2014).
View in article
Caro, G., Bourdon, B. (2010) Non-chondritic Sm/Nd ratio in the terrestrial planets: consequences for the geochemical evolution of the mantle–crust system. Geochimica et Cosmochimica Acta 74, 3333–3349. https://doi.org/10.1016/j.gca.2010.02.025
Show in context Considering that the compositions of the initial LMO have not been fully elucidated, three compositions were investigated, namely, a chondritic 147Sm/144Nd = 0.1967 and 176Lu/177Hf = 0.0336, as described by Snyder et al. (1992); a moderately depleted composition of 147Sm/144Nd = 0.2008 and 176Lu/177Hf = 0.0375 by Johnston et al. (2022); and a highly depleted composition of 147Sm/144Nd = 0.2082 and 176Lu/177Hf = 0.0375 by Caro and Bourdon (2010).
View in article
Che, X., Nemchin, A., Liu, D., Long, T., Wang, C., Norman, M.D., Joy, K.H., Tartese, R., Head, J., Jolliff, B., Snape, J.F., Neal, C.R., Whitehouse, M.J., Crow, C., Benedix, G., Jourdan, F., Yang, Z., Yang, C., Liu, J., Xie, S., Bao, Z., Fan, R., Li, D., Li, Z., Webb, S.G. (2021) Age and composition of young basalts on the Moon, measured from samples returned by Chang’e-5. Science 374, 887–890. https://doi.org/10.1126/science.abl7957
Show in context The data yield an age of 1962 ± 90 Ma (MSWD = 1.6, n = 9; Fig. 1), in accordance with 1963 ± 57 Ma (Che et al., 2021) and 2030 ± 4 Ma (Li et al., 2021) for other CE5 clasts.
View in article
Charlier, B., Grove, T.L., Namur, O., Holtz, F. (2018) Crystallization of the lunar magma ocean and the primordial mantle-crust differentiation of the Moon. Geochimica et Cosmochimica Acta 234, 50–69. https://doi.org/10.1016/j.gca.2018.05.006
Show in context
The crystallising modes followed Snyder et al. (1992), Elardo et al. (2011), Elkins-Tanton et al. (2011), and Charlier et al. (2018) to fully constrain potential effects of different mineral assemblages on their sources (Fig. S-7).
View in article
For the initial LMO with chondritic or moderately depleted compositions, the calculated results are similar: the mantle sources of the low-Ti basalts can be explained by a moderate-stage LMO cumulate (e.g., sequence 3 from Snyder et al. (1992), Charlier et al. (2018), and Elkins-Tanton et al. (2011)) with different proportions of TIRL.
View in article
Elardo, S.M., Shearer, C.K., Fagan, A.L., Borg, L.E., Gaffney, A.M., Burger, P.V., Neal, C.R., Fernandes, V.A., McCubbin, F.M. (2014) The origin of young mare basalts inferred from lunar meteorites Northwest Africa 4734, 032, and LaPaz Icefield 02205. Meteoritics and Planetary Science 49, 261–291. https://doi.org/10.1111/maps.12239
Show in context (a) 87Rb/86Sr values of mantle sources were calculated assuming a single-stage differentiated evolution from 4558 Ma with a lunar initial 87Sr/86Sr of 0.69903 following Elardo et al. (2014).
View in article
(b-c) The εNd and εHf values of mantle sources were calculated assuming source formation at the time of LMO crystallisation (4.42 Ga) with a chondritic composition (Nyquist et al., 1995; Elardo et al., 2014).
View in article
Isotopic data: Elardo et al. (2014) and references therein, Sprung et al. (2013), Beard et al. (1998).
View in article
Elardo, S.M., Draper D.S., Shearer Jr. C.K. (2011) Lunar Magma Ocean crystallization revisited: Bulk composition, early cumulate mineralogy, and the source regions of the highlands Mg-suite. Geochimica et Cosmochimica Acta 75, 3024–3045. https://doi.org/10.1016/j.gca.2011.02.033
Show in context The crystallising modes followed Snyder et al. (1992), Elardo et al. (2011), Elkins-Tanton et al. (2011), and Charlier et al. (2018) to fully constrain potential effects of different mineral assemblages on their sources (Fig. S-7).
View in article
Elkins-Tanton, L.T., Burgess, S., Yin, Q.Z. (2011) The lunar magma ocean: Reconciling the solidification process with lunar petrology and geochronology. Earth and Planetary Science Letters 304, 326–336. https://doi.org/10.1016/j.epsl.2011.02.004
Show in context The crystallising modes followed Snyder et al. (1992), Elardo et al. (2011), Elkins-Tanton et al. (2011), and Charlier et al. (2018) to fully constrain potential effects of different mineral assemblages on their sources (Fig. S-7).
View in article
For the initial LMO with chondritic or moderately depleted compositions, the calculated results are similar: the mantle sources of the low-Ti basalts can be explained by a moderate-stage LMO cumulate (e.g., sequence 3 from Snyder et al. (1992), Charlier et al. (2018), and Elkins-Tanton et al. (2011)) with different proportions of TIRL.
View in article
Hallis, L., Anand, M., Strekopytov, S. (2014) Trace-element modelling of mare basalt parental melts: implications for a heterogeneous lunar mantle. Geochimica et Cosmochimica Acta 134, 289–316. https://doi.org/10.1016/j.gca.2014.01.012
Show in context An intermediate proportion of TIRL (∼2 %) trapped in sequence 3 could reproduce the [176Lu/177Hf] and [147Sm/144Nd] of the CE5 basalt (black solid curves in Fig. 3), which falls between those for A15 (>2 % TIRL) and A12 (<1.5% TIRL) low-Ti basalts, consistent with the prediction by Hallis et al. (2014).
View in article
Haskin, L.A., Gillis, J.J., Korotev, R.L., Jolliff, B.L. (2000) The materials of the lunar Procellarum KREEP Terrane: A synthesis of data from geomorphological mapping, remote sensing, and sample analyses. Journal of Geophysical Research: Planets 105, 20403–20415. https://doi.org/10.1029/1999je001128
Show in context Furthermore, radiogenic heat from the KREEP has been proposed to induce extensive lunar mantle melting (Haskin et al., 2000).
View in article
Jiang, Y., Kang, J., Liao, S., Elardo, S.M., Zong, K., Wang, S., Nie, C., Li, P., Yin, Z., Huang, F., Hsu, W. (2023) Fe and Mg Isotope Compositions Indicate a Hybrid Mantle Source for Young Chang’E 5 Mare Basalts. The Astrophysical Journal Letters 945, L26. https://doi.org/10.3847/2041-8213/acbd31
Show in context However, recent analyses of CE5 mare basalt clasts have provided opposite conclusions about whether the mantle source was pristine or hybridised (mixing of LMO cumulates or melts at different cooling stages), which is critical for determining the mechanism of this young melting event (Tian et al., 2021; Su et al., 2022; Zong et al., 2022; Jiang et al., 2023).
View in article
These models have also been assessed, e.g., based on the Snyder’s model, an early-stage Ol cumulate assimilated with 30 % late-stage Cpx-Ilm-rich cumulate (Seq 5) (Jiang et al., 2023), or an early-stage Ol-Opx cumulate (Seq 1 + 2) assimilated with 20 % late-stage Cpx-Ilm-rich cumulate (Seq 5) (Su et al., 2022), or a late-stage Cpx-Ilm dominated cumulate (Seq 4) (Wang et al., 2024), or a moderate-stage Cpx-rich (40–60 %) cumulate assimilated with 0.3–1.5 % KREEP components (Zong et al., 2022).
View in article
We note that the Fe-Mg isotopic compositions of the CE5 basalts observed by Jiang et al. (2023) are within the range of Apollo low-Ti basalt data, which does not necessitate the introduction of a hybrid source.
View in article
Johnston, S., Brandon, A., McLeod, C., Rankenburg, K., Becker, H., Copeland, P. (2022) Nd isotope variation between the Earth–Moon system and enstatite chondrites. Nature 611, 501–506. https://doi.org/10.1038/s41586-022-05265-0
Show in context Considering that the compositions of the initial LMO have not been fully elucidated, three compositions were investigated, namely, a chondritic 147Sm/144Nd = 0.1967 and 176Lu/177Hf = 0.0336, as described by Snyder et al. (1992); a moderately depleted composition of 147Sm/144Nd = 0.2008 and 176Lu/177Hf = 0.0375 by Johnston et al. (2022); and a highly depleted composition of 147Sm/144Nd = 0.2082 and 176Lu/177Hf = 0.0375 by Caro and Bourdon (2010).
View in article
Li, Q.-L., Zhou, Q., Liu, Y., Xiao, Z., Lin, Y., Li, J.H., Ma, H.X., Tang, G.Q., Guo, S., Yuan, J.Y., Li, J., Wu, F.Y., Ouyang, Z., Li, C., Li, X.H. (2021) Two-billion-year-old volcanism on the Moon from Chang’e-5 basalts. Nature 600, 54–58. https://doi.org/10.1038/s41586-021-04100-2
Show in context The data yield an age of 1962 ± 90 Ma (MSWD = 1.6, n = 9; Fig. 1), in accordance with 1963 ± 57 Ma (Che et al., 2021) and 2030 ± 4 Ma (Li et al., 2021) for other CE5 clasts.
View in article
McKay, G., Wagstaff, J., Le, L. (1991) REE distribution coefficients for pigeonite: constraints on the origin of the mare basalt europium anomaly, III. Mare volcanism and basalt petrogenesis workshop 91-03, 27–28.
Show in context The basaltic glass REE is calculated based on the highest Fe pigeonite (HF Pig) (McKay et al., 1991) or the felsic glass (Fg) (Shearer et al., 2001). The chondritic and KREEP data are from Anders and Grevesse (1989) and Warren (1989), respectively.
View in article
Nyquist, L.E., Wiesmann, H., Bansal, B., Shih, C.Y., Keith, J.E., Harper, C.L. (1995) 146Sm-142Nd formation interval for the lunar mantle. Geochimica et Cosmochimica Acta 59, 2817–2837. https://doi.org/10.1016/0016-7037(95)00175-y
Show in context However, the Nd-Hf isotopic compositions of lunar materials can be altered by cosmic ray irradiation (Nyquist et al., 1995; Sprung et al., 2010; Carlson et al., 2014).
View in article
(b-c) The εNd and εHf values of mantle sources were calculated assuming source formation at the time of LMO crystallisation (4.42 Ga) with a chondritic composition (Nyquist et al., 1995; Elardo et al., 2014).
View in article
For comparison, the corrected isotopic data are expressed as parent-to-daughter ratios according to their crystallisation ages that would be required for their mantle sources to evolve from the isotopic composition of the LMO, assuming an initial LMO crystallisation age at ∼4.42 Ga (Nyquist et al., 1995).
View in article
Shearer, C.K., Papike, J.J., Spilde, M.N. (2001) Trace-element partitioning between immiscible lunar melts: An example from naturally occurring lunar melt inclusions. American Mineralogist 86, 238–246. https://doi.org/10.2138/am-2001-2-305
Show in context The basaltic glass REE is calculated based on the highest Fe pigeonite (HF Pig) (McKay et al., 1991) or the felsic glass (Fg) (Shearer et al., 2001). The chondritic and KREEP data are from Anders and Grevesse (1989) and Warren (1989), respectively.
View in article
These residual liquids undergo immiscibility splitting into Fe-rich (basaltic glass) and Si-K-rich (felsic glass) conjugate liquids (Shearer et al., 2001).
View in article
Shearer, C.K., Hess, P.C., Wieczorek, M.A., Pritchard, M.E., Parmentier, E.M., Borg, L.E., Longhi, J., Elkins-Tanton, L.T., Neal, C.R., Antonenko, I., Canup, R.M., Halliday, A.N., Grove, T.L., Hager, B.H., Lee, D.-C., Wiechert, U. (2006) Thermal and magmatic evolution of the Moon. Reviews in Mineralogy and Geochemistry 60, 365–518. https://doi.org/10.2138/rmg.2006.60.4
Show in context Based on the Apollo and Luna samples, most previous studies have linked regional incompatible element enrichments to pre-existing KREEP components (Shearer et al., 2006 and references therein), which represent residues after extensive fractional crystallisations (>99 %) of the Lunar Magma Ocean (LMO) (Warren, 1989).
View in article
Shearer, C.K., Neal, C.R., Glotch, T.D., Prissel, T.C., Bell, A.S., Fernandes, V.A., Gaddis, L.R., Jolliff, B.L., Laneuville, M., Magna, T., Simon, J. (2023) Magmatic evolution II: A new view of post-differentiation magmatism. Reviews in Mineralogy and Geochemistry 89, 147–206. https://doi.org/10.2138/rmg.2023.89.04
Show in context The Chang’e 5 (CE5) mission returned mare basalts with eruption ages of ∼2.0 Ga, indicating that the volcanic event in the PKT mare region lasted an additional 0.8 to 1.0 Gyr after the youngest lunar volcanic magmatism event (2.8 to 3.0 Ga; Shearer et al., 2023) dated before the CE5 mission.
View in article
Snyder, G.A., Taylor, L.A., Neal, C.R. (1992) A chemical model for generating the sources of mare basalts: Combined equilibrium and fractional crystallization of the lunar magmasphere. Geochimica et Cosmochimica Acta 56, 3809–3823. https://doi.org/10.1016/0016-7037(92)90172-f
Show in context The crystallising modes followed Snyder et al. (1992), Elardo et al. (2011), Elkins-Tanton et al. (2011), and Charlier et al. (2018) to fully constrain potential effects of different mineral assemblages on their sources (Fig. S-7).
View in article
Considering that the compositions of the initial LMO have not been fully elucidated, three compositions were investigated, namely, a chondritic 147Sm/144Nd = 0.1967 and 176Lu/177Hf = 0.0336, as described by Snyder et al. (1992); a moderately depleted composition of 147Sm/144Nd = 0.2008 and 176Lu/177Hf = 0.0375 by Johnston et al. (2022); and a highly depleted composition of 147Sm/144Nd = 0.2082 and 176Lu/177Hf = 0.0375 by Caro and Bourdon (2010).
View in article
(a, c, e) Crystallisation mode of Snyder et al. (1992) with different initial LMO compositions.
View in article
For the initial LMO with chondritic or moderately depleted compositions, the calculated results are similar: the mantle sources of the low-Ti basalts can be explained by a moderate-stage LMO cumulate (e.g., sequence 3 from Snyder et al. (1992), Charlier et al. (2018), and Elkins-Tanton et al. (2011)) with different proportions of TIRL.
View in article
Sprung, P., Kleine, T., Scherer, E.E. (2013) Isotopic evidence for chondritic Lu/Hf and Sm/Nd of the Moon. Earth and Planetary Science Letters 380, 77–87. https://doi.org/10.1016/j.epsl.2013.08.018
Show in context In particular, Hf–Nd isotopic compositions can provide insight into the mantle source for mare basalts, because Hf–Nd isotopic covariations are sensitive to different mineral assemblages and source hybridisations (Beard et al., 1998; Sprung et al., 2013).
View in article
Isotopic data: Elardo et al. (2014) and references therein, Sprung et al. (2013), Beard et al. (1998).
View in article
The coupled Nd–Hf isotopic compositions can serve as conclusive evidence for distinguishing mantle sources because the lunar low-Ti basalt source typically has a [Lu/Hf]n ratio (n refers to CI chondrite-normalised value) ∼4 times greater than its [Sm/Nd]n ratio, whereas the high-Ti source has subequal [Lu/Hf]n and [Sm/Nd]n ratios (∼1) (Beard et al., 1998; Sprung et al., 2013).
View in article
In particular, ilmenite (Ilm) has a markedly lower KD[Lu/Hf]/KD[Sm/Nd] (∼0.07) than other accumulated minerals, with KD[Lu/Hf]/KD[Sm/Nd] values of ∼0.7 to ∼4 from the LMO (e.g., olivine (Ol), orthopyroxene (Opx), clinopyroxene (Cpx), pigeonite (Pig), plagioclase (Pl); Sprung et al., 2013; Yang et al., 2022), accounting for the distinct trends of low-Ti and high-Ti basalts (Fig. 2c).
View in article
This correlation can be potentially interpreted as a result of a common LMO low-Ti cumulate with variable amounts of trapped instantaneous residual liquid (TIRL), or a hybrid source that comprises early-stage Ol and Opx cumulates containing various fractions of TIRL with a small amount of late-stage Cpx-bearing cumulates (Beard et al., 1998; Sprung et al., 2013).
View in article
Instead, a complicated scenario in which KREEP mixes with a range of different LMO cumulates in just the right proportions is required, which seems to be too coincidental, as also has been mentioned by Sprung et al. (2013).
View in article
The black dotted curves represent mixtures of KREEP and Seq 3 cumulate. Data source: Sprung et al. (2013) and Unruh et al. (1984).
View in article
Su, B., Yuan, J., Chen, Y., Yang, W., Mitchell, R.N., Hui, H., Wang, H., Tian, H., Li, X.H., Wu, F.Y. (2022) Fusible mantle cumulates trigger young mare volcanism on the Cooling Moon. Science Advances 8, eabn2103. https://doi.org/10.1126/sciadv.abn2103
Show in context However, recent analyses of CE5 mare basalt clasts have provided opposite conclusions about whether the mantle source was pristine or hybridised (mixing of LMO cumulates or melts at different cooling stages), which is critical for determining the mechanism of this young melting event (Tian et al., 2021; Su et al., 2022; Zong et al., 2022; Jiang et al., 2023).
View in article
These models have also been assessed, e.g., based on the Snyder’s model, an early-stage Ol cumulate assimilated with 30 % late-stage Cpx-Ilm-rich cumulate (Seq 5) (Jiang et al., 2023), or an early-stage Ol-Opx cumulate (Seq 1 + 2) assimilated with 20 % late-stage Cpx-Ilm-rich cumulate (Seq 5) (Su et al., 2022), or a late-stage Cpx-Ilm dominated cumulate (Seq 4) (Wang et al., 2024), or a moderate-stage Cpx-rich (40–60 %) cumulate assimilated with 0.3–1.5 % KREEP components (Zong et al., 2022).
View in article
Tian, H.-C., Wang, H., Chen, Y., Yang, W., Zhou, Q., Zhang, C., Lin, H.L., Huang, C., Wu, S.T., Jia, L.H., Xu, L., Zhang, D., Li, X.G., Chang, R., Yang, Y.H., Xie, L.W., Zhang, D.P., Zhang, G.L., Yang, S.H., Wu, F.Y. (2021) Non-KREEP origin for Chang’e-5 basalts in the Procellarum KREEP Terrane. Nature 600, 59–63. https://doi.org/10.1038/s41586-021-04119-5
Show in context
However, recent analyses of CE5 mare basalt clasts have provided opposite conclusions about whether the mantle source was pristine or hybridised (mixing of LMO cumulates or melts at different cooling stages), which is critical for determining the mechanism of this young melting event (Tian et al., 2021; Su et al., 2022; Zong et al., 2022; Jiang et al., 2023).
View in article
In situ Sr isotopic composition of plagioclase and Nd isotopic composition of merrillite suggests a depleted mantle source for the CE5 basalts (Tian et al., 2021).
View in article
Second, its whole rock Sr and Nd isotopic compositions are comparable to in situ analysed data within errors (Fig. 2; Tian et al., 2021).
View in article
The grey bars represent in situ Sr and Nd isotopic compositions (Tian et al., 2021).
View in article
Unruh, D.M., Stille, P., Patchett, P.J., Tatsumoto, M. (1984) Lu-Hf and Sm-Nd evolution in lunar mare basalts. Journal of Geophysical Research 89, B459–B477. https://doi.org/10.1029/jb089is02p0b459
Show in context The black dotted curves represent mixtures of KREEP and Seq 3 cumulate. Data source: Sprung et al. (2013) and Unruh et al. (1984).
View in article
Wang, C., Xu, Y.G., Zhang, L., Chen, Z., Xia, X., Lin, M., Guo, F. (2024) A shallow (< 100 km) ilmenite-bearing pyroxenitic source for young lunar volcanism. Earth and Planetary Science Letters 639, 118770. https://doi.org/10.1016/j.epsl.2024.118770
Show in context
These models have also been assessed, e.g., based on the Snyder’s model, an early-stage Ol cumulate assimilated with 30 % late-stage Cpx-Ilm-rich cumulate (Seq 5) (Jiang et al., 2023), or an early-stage Ol-Opx cumulate (Seq 1 + 2) assimilated with 20 % late-stage Cpx-Ilm-rich cumulate (Seq 5) (Su et al., 2022), or a late-stage Cpx-Ilm dominated cumulate (Seq 4) (Wang et al., 2024), or a moderate-stage Cpx-rich (40–60 %) cumulate assimilated with 0.3–1.5 % KREEP components (Zong et al., 2022).
View in article
Although the petrological and mineralogical signature of CE5 basalts can be fulfilled by hybridisation of an early-stage cumulate by a late-stage Cpx- and Ilm- rich cumulate (Su et al., 2022; Wang et al., 2024), the highly differentiated nature and uncertainties of melting conditions (e.g., temperature, pressure, geothermal gradient) will likely impose a large uncertainty on the estimation of the source composition.
View in article
Warren, P.H. (1989) KREEP: major-element diversity, trace-element uniformity (almost). Workshop on Moon in transition: Apollo 14, KREEP, and evolved lunar rocks. 149–153.
Show in context Based on the Apollo and Luna samples, most previous studies have linked regional incompatible element enrichments to pre-existing KREEP components (Shearer et al., 2006 and references therein), which represent residues after extensive fractional crystallisations (>99 %) of the Lunar Magma Ocean (LMO) (Warren, 1989).
View in article
Given the high enrichments of Lu–Hf and Sm–Nd and subchondritic (Lu/Hf)source (∼0.0187) and (Sm/Nd)source (∼0.1630) in KREEP (Warren, 1989), KREEP component incorporation into the cumulate would lead to rapid decreases in both (Lu/Hf)n and (Sm/Nd)n ratios (Fig. 3).
View in article
The basaltic glass REE is calculated based on the highest Fe pigeonite (HF Pig) (McKay et al., 1991) or the felsic glass (Fg) (Shearer et al., 2001). The chondritic and KREEP data are from Anders and Grevesse (1989) and Warren (1989), respectively.
View in article
Yao, L., Sun, C., Liang, Y. (2012) A parameterized model for REE distribution between low-Ca pyroxene and basaltic melts with applications to REE partitioning in low-Ca pyroxene along a mantle adiabat and during pyroxenite-derived melt and peridotite interaction. Contributions to Mineralogy and Petrology 164, 261–280. https://doi.org/10.1007/s00410-012-0737-5
Show in context Using the sequence 3 cumulate from Snyder’s model, the parental magma of the CE5 basalts (calculated from the most primitive pyroxene with the highest Mg# and partition coefficients of low-Ca Opx at 1300 °C from Yao et al. (2012)) was generated by 2–4 % partial melting, followed by 39–83 % fractional crystallisation (Fig. 4a, Table S-15).
View in article
Yang, W., Chen, Y., Wang, H., Tian, H.C., Hui, H., Xiao, Z., Wu, S.T., Zhang, D., Zhou, Q., Ma, H.X., Zhang, C., Hu, S., Li, Q.L., Lin, Y., Li, X.H., Wu, F.Y. (2022) Geochemistry of impact glasses in the Chang’e-5 regolith: Constraints on impact melting and the petrogenesis of local basalt. Geochimica et Cosmochimica Acta 335, 183–196. https://doi.org/10.1016/j.gca.2022.08.030
Show in context In particular, ilmenite (Ilm) has a markedly lower KD[Lu/Hf]/KD[Sm/Nd] (∼0.07) than other accumulated minerals, with KD[Lu/Hf]/KD[Sm/Nd] values of ∼0.7 to ∼4 from the LMO (e.g., olivine (Ol), orthopyroxene (Opx), clinopyroxene (Cpx), pigeonite (Pig), plagioclase (Pl); Sprung et al., 2013; Yang et al., 2022), accounting for the distinct trends of low-Ti and high-Ti basalts (Fig. 2c).
View in article
Zong, K., Wang, Z., Li, J., He, Q., Li, Y., Becker, H., Zhang, W., Hu, Z., He, T., Cao, K., She, Z., Wu, X., Xiao, L., Liu, Y. (2022) Bulk compositions of the Chang’E-5 lunar soil: Insights into chemical homogeneity, exotic addition, and origin of landing site basalts. Geochimica et Cosmochimica Acta 335, 284–296. https://doi.org/10.1016/j.gca.2022.06.037
Show in context However, recent analyses of CE5 mare basalt clasts have provided opposite conclusions about whether the mantle source was pristine or hybridised (mixing of LMO cumulates or melts at different cooling stages), which is critical for determining the mechanism of this young melting event (Tian et al., 2021; Su et al., 2022; Zong et al., 2022; Jiang et al., 2023).
View in article
These models have also been assessed, e.g., based on the Snyder’s model, an early-stage Ol cumulate assimilated with 30 % late-stage Cpx-Ilm-rich cumulate (Seq 5) (Jiang et al., 2023), or an early-stage Ol-Opx cumulate (Seq 1 + 2) assimilated with 20 % late-stage Cpx-Ilm-rich cumulate (Seq 5) (Su et al., 2022), or a late-stage Cpx-Ilm dominated cumulate (Seq 4) (Wang et al., 2024), or a moderate-stage Cpx-rich (40–60 %) cumulate assimilated with 0.3–1.5 % KREEP components (Zong et al., 2022).
View in article
top
Supplementary Information
The Supplementary Information includes:
- Materials and Methods
- Variations of TiO2 for Whole-Rock and Mineral
- Figures S-1 to S-11
- Tables S-1 to S-17
- Supplementary Information References
Download the Supplementary Information (PDF)
Download Tables S-4 to S-10 (.xlsx)
Download Table S-11 (.xlsx)
Download Table S-12 (.xlsx)
Download Table S-13 (.xlsx)
Download Table S-14 (.xlsx)
