Evolution of the early Moon: insights from experimental petrology and returned lunar samples

This thesis presents my main findings exploring different aspects of the Moon's history and evolution including (1) the structure and composition of the early Moon, (2) the volatile (F) evolution of the lunar magma ocean (LMO), (3) the olivine-melt partition coefficients of first-row transition elements application into lunar basalts genesis, (4) genesis and source region of Chang’E-5 basalts. Chapter 2 examines the stability of garnet in the deep lunar mantle. The chapter presents results from high-pressure, high-temperature experiments at deep lunar mantle conditions to determine the garnet stability field. The results indicate that garnet is stable at pressures above 3 GPa and temperatures below 1700 ℃, yielding a smaller stability field than previously suggested on the basis of thermodynamic calculations. This study also used experimental data to model equilibrium crystallization in a ‘two-stage’ model of LMO crystallization starting from a fully molten Moon. The findings suggest that garnet in the deep lunar mantle would significantly decrease the Al2O3 content of the residual LMO and impact HREE/MREE fractionation. Chapter 3 investigates the F evolution during LMO crystallization. The study conducted high-pressure, high-temperature experiments to quantify F mineral-melt partitioning for lunar minerals (plagioclase, orthopyroxene, and ilmenite). The results constrain the F abundance in the magma ocean to 21-41 ppm at the time crust-forming plagioclase started crystallizing. Starting at this new anchor point, the F contents in initial LMO and the urKREEP reservoir have been modeled. Based on the calculated 4.1-8.2 ppm F abundance in the initial LMO, this chapter discusses the possible ways in which the Moon lost F and provides evidence suggesting that the depletion of F could have occurred during or prior to the early LMO stages, whereas the onset of crust formation appears to have provided an efficient lid on the magma ocean, limiting degassing. Chapter 4 focuses on determining partition coefficients between olivine and silicate melt of various elements in the simple system FeO-CaO-MgO-Al2O3-SiO2 (FCMAS) and re-assessing three aspects of lunar basalt generation using the newly derived partition coefficients. The first-row transition element (FRTE) partition coefficients between olivine and melt were experimentally obtained under oxygen fugacities ranging from very reducing to very oxidizing. The new results were applied to re-assess the origin of lunar basalts in three aspects including (1) the source region lithology by Cr systematics, (2) redox state by Co-Ni systematics and (3) melting temperature by an updated Mn geothermometer. In addition to the applications shown in this chapter, these new partition coefficients data can be further applied in the studies of origin of other planetary basalts. Chapter 5 reports findings from the Chang'E-5 mission that provide new insights into the thermal and compositional state of the lunar interior around 2 billion years ago. The study provides new constraints on the source region of the Chang’E-5 basalts by measuring the concentrations of Sc, Co, Ni, and Ge in olivine from basaltic clasts. The results suggest (1) that 1% batch melting of shallow lunar mantle cumulates (olivine pyroxenite) can reproduce magma with chemistry characteristics consistent with our and previous sample analyses; (2) that the presence of a metal phase persisted in the lunar magma ocean after lunar core formation, leading to Ni-poor LMO cumulates at great depths and a Ni-rich Chang'E-5 shallow source. Low Ge contents of the lunar core (by core formation at mildly reducing conditions) or late addition of Ge-rich meteoritic materials is required to explain the observed high Ge contents of the Chang’E-5 basalts. These results illustrate clearly the strong and direct connection between the earliest evolution of the Moon and the formation of the youngest lunar rocks 2.5 billion years later.