[S-IT14] 地球深部科学

[SIT14-P03] Five-step fractional crystallisation of a late Lunar Magma Ocean implication for primitive KREEP

*Dilan Harsha Fernando Diyalanthonige¹、Takashi Yoshino¹、Robert Myhill²、Richard Brooker²、Amanda Lindoo² (1.Institute for Planetary Materials, Okayama University、2.University of Bristol)

キーワード：Moon, Fractional crystallisation, KREEP, Lunar magma ocean

After its accretion, the Moon underwent the Lunar Magma Ocean (LMO) [1], [2]. Differentiation started, dense minerals such as olivine and pyroxene would have sunk, and light minerals such as plagioclase would have floated, forming the crust [1], [2]. The KREEP (K, REE, P, Th, U enriched) basalts were one of the discoveries that proved internal differentiation [2], [3]. Since then, research led to the idea that the LMO had been highly convective at the early stages, and mineral deposition would have slowly cumulated by equilibrium crystallisation at the lower mantle [2], [3], [4], [5], [6]. The upper mantle would have subsequently followed with continuous fractional crystallisation. This caused incompatible elements, such as K, REE, P, U, Th and Ti, to be left in the melt, increasing their concentration until forming an enriched layer beneath the crust [2], [3], [4], [5], [6]. This study concentrates on recreating the primitive KREEP using the Taylor Whole Moon (TWM) starting composition at 50 per cent solid (PCS) of the LMO [5]. Previous experiments have defined a nearly homogenous lower mantle mainly composed of olivine + pyroxenes and a possible small amount of garnets [6], [7]. The experiments have been primarily run using a piston-cylinder to recreate a 5-step fractional crystallisation from 50 to 99 PCS with isobaric pressure from 1.73 to 0.5 GPa using the melt of the previous experiment as a new starting composition [4], [5], [6]. Then analysed using SEM and EPMA. The result shows that the crystallisation follows the trend of 50 PCS olivine + orthopyroxene + spinel followed by 70 PCS with orthopyroxene + clinopyroxene + spinel + olivine + plagioclase then 85 PCS orthopyroxene + clinopyroxene + plagioclase + spinel +quartz + ilmenite, 95 PCS clinopyroxene + plagioclase + quartz + spinel + ilmenite + orthopyroxene and finishing with 99 PCS clinopyroxene + plagioclase + quartz + spinel + ilmenite. The fractional crystallisation follows similar trends as seen in previous experiments [4], [5], [6]. Plagioclase formed at 70 PCS and ilmenite at 85 PCS due to the high Ti content. However, the formation of quartz at past 85 PCS was unexpected, indicating that the previous models [5] would be present beneath the anorthosite crust. Similarly, in earlier experiments, FeO increased up to 20 wt%, and Ti increased to over 4 wt% in the melt at 99PCS. At this stage, it would have been denser than the beneath layers starting the gravitational overturn. These experiments have shown a pyroxenitic upper mantle. As in each step, plagioclase and quartz are expected to be removed from each cumulate since less dense than the surrounding melt. The primitive KREEP REE trend shows that the primitive KREEP was highly concentrated above any sample retrieved and the last models. Then primitive KREEP would have been an extremely enriched layer that partially reset the lunar upper mantle, leading to volcanism that formed present KREEP and Ti basalts.

[1] R. M. Canup and K. Righter, "Origin of the Earth and Moon," University of Arizona Press, 2000.
[2] B. Jolliff et al., "Origin and Evolution of the Moon's Procellarum KREEP Terrane," Bull. AAS, vol. 53, no. 4, Mar. 2021, doi: 10.3847/25c2cfeb.8668f714.
[3] S. R. Taylor, "The Moon re-examined," Geochim. Cosmochim. Acta, vol. 141, pp. 670–676, Sep. 2014, doi: 10.1016/j.gca.2014.06.031.
[4] J.-J. Jing, "Garnet stability in the deep lunar mantle: Constraints on the physics and chemistry of the interior of the Moon," Earth Planet. Sci. Lett., p. 13, 2022.
[5] T. E. Johnson et al., "The phases of the Moon: Modelling crystallisation of the lunar magma ocean through equilibrium thermodynamics," Earth Planet. Sci. Lett., vol. 556, p. 116721, Feb. 2021, doi: 10.1016/j.epsl.2020.116721.
[6] S. M. Elardo et al., "Lunar Magma Ocean crystallisation revisited: Bulk composition, early cumulate mineralogy, and the source regions of the highlands Mg-suite," Geochim. Cosmochim. Acta, vol. 75, no. 11, pp. 3024–3045, Jun. 2011, doi: 10.1016/j.gca.2011.02.033.