[S-IT16] Deep Earth Sciences

[SIT16-12] Five-step fractional crystallisation of a late Lunar Magma Ocean implication for a primitive KREEP

*Dilan Harsha Fernando Diyalanthonige^1,2, Takashi Yoshino¹, Robert Myhill², Richard Brooker², Amanda Lindoo² (1.Institute for Planetary Materials, Okayama University, Misasa, 682-0193 Tottori, Japan, 2.School of Earth Sciences, University of Bristol, Bristol, United Kingdom)

Keywords:Moon, KREEP, Fractional crystallisation, Lunar magma ocean

The impact that led to the formation of the Moon was followed by a stage named lunar magma ocean (LMO). During this stage, the Moon was fully molten, causing the differentiation of its interior. Evidence of this hypothesis has been given after studying returned samples from missions (e.g. Apollo missions), meteorites recovered on Earth and remote sensing of the lunar surface. However, one of the most fundamental discoveries is the KREEP basalts. These are basalts enriched with K, REE, P, Th, U and other incompatible elements. Since their discovery, the origin and formation of the KREEP basalts have been the subject of debate. Later, it was proposed that the lower mantle was formed by equilibrium crystallisation, and the upper mantle was formed by fractional crystallisation as a solution to the origin of KREEP and the surface rocks. The idea of fractional crystallisation is that as it crystallises different minerals, the melt would be left with a mixture of less-compatible elements. As this continues, it would lead to a last dreg highly concentrated of incompatible elements, such as the one seen in KREEP basalts. However, this last residue (>99PCS) is the primitive KREEP. This study is concentrated on recreating the primitive KREEP. The starting composition is Taylor Whole Moon (TWM), specifically the TWM at 50 PCS. Then, run a set of experiments using a piston-cylinder to make a 5-step fractional crystallisation (50 - 70 - 85- 95 - 99 PCS), isobaric pressure (1.73 - 0.5 GPa), and using the melt of the previous step as new starting composition. The resultant crystallisation descent seems to follow (50 PCS) olivine + orthopyroxene + spinel ->(70 PCS) orthopyroxene + clinopyroxene + spinel + olivine + plagioclase ->(85 PCS) orthopyroxene + clinopyroxene + plagioclase + spinel +quartz + Fe/Ti oxide ->(95 PCS) clinopyroxene + plagioclase + quartz + spinel + Fe/Ti oxide + orthopyroxene ->(99 PCS) clinopyroxene + plagioclase + quartz + Spinel Fe/Ti Oxide. However, the early formation of quartz at 85 PCS indicates their presence beneath a few km of felspathic crust. Fe/Ti Oxide (ilmenite) formation at 85 PCS follows the past trends. The presence of Fe at 20 wt% and Ti over 4 wt% minerals increase into the melt as we reach our 99 PCS, which, similarly to past literature, follows the idea that KREEP may have formed to the very last stage with a melt enriched of Fe-Ti that may have caused the consequent gravitational overturn. We point to a set of different points compared to past papers, such as the removal of olivine at 70 PCS and orthopyroxene at 99 PCS or the addition of quartz at 85 PCS. Another difference is that within the melt phases, the composition increases Si content between the steps, which differs from previous literature. The implication is that the evolution of the lunar upper mantle during the LMO fractional crystallisation continues to crystallise minerals such as olivine - orthopyroxene – spinel – Fe/Ti oxides, increasing the concentration of Si in the melt. REE elements model was used to define a concentration of the primitive KREEP prior to the gravitational overturn. First, we define the cumulate REE concentration showing that increments of plagioclase lead to the Eu anomaly. Then we use this model and compare it with REE elements defined by FXMOTR from Davenport et al. (2014). Lastly, calculate the cumulate concentration and use these with the idea that a overturn may have led to remelting and volcanism that formed the KREEP basalts.
Preliminary results show that primitive KREEP can only form as a consequence of the extreme fractional crystallisation after the 99 PCS, and that needs a highly enriched Fe - Ti cumulate to gravitational overturn to start. The REE elements model confirms that a high concentration of KREEP occurs only after 99 PCS. In addition, the composition between the steps indicates that the melts went from gabbroic to basaltic in the final step.