Experimental Evidence Supporting an Overturned Iron‐Titanium‐Rich Melt Layer in the Deep Lunar Interior

Abstract

Dense Fe‐Ti‐rich cumulates, formed as the last dregs of the lunar magma ocean, are thought to have driven a large‐scale overturn of the lunar mantle over 4 Ga ago. Analysis of lunar seismic data has implied that some of the overturned bodies may have reached the lunar core‐mantle boundary and remained there until the present day as a partially molten layer. However, whether such a molten layer could be stable during >4 Ga of post‐magma‐ocean lunar history and explain lunar seismic observations remains poorly constrained. Here, we report the first sound velocity measurements on a Fe‐Ti‐rich lunar melt up to conditions of the lowermost lunar mantle. Our results suggest that a partial melt layer with at least 20% overturned Fe‐Ti‐rich melt can be trapped atop the lunar core‐mantle boundary until the present day, strongly influencing the thermochemical evolution of the lunar interior.

The Volcanic and Radial Expansion/Contraction History of the Moon Simulated by Numerical Models of Magmatism in the Convective Mantle

“…(see Equations and and Figure 10b as well as the assumed values of Δ ρ ∞ / ρ 0 = 0.005, Δ ρ zero / ρ 0 = 0.22, and λ = 16.42 GPa in Equation .) However, some studies suggest that IBC‐rich magma can be denser than the coexisting matrix in the deep mantle of the Moon (e.g., Sakamaki et al., 2010; van Kan Parker et al., 2012; Xu et al., 2022). To see how this possible density inversion affects our numerical results, we calculated the two additional cases (vl‐change and beta0202).…”

“…A comparison with earlier classical 1-D models of lunar thermal history in the literature shows the crucial role that heat transport by migrating magma and mantle convection plays in the reference case. In earlier 1-D models where only internal heating and thermal conduction are considered, the temperature in the deep mantle monotonously increases to the solidus temperature due to internal heating, while the lithosphere monotonously thickens with time owing to cooling from the surface boundary throughout the calculated history (e.g., Solomon & Chaiken, 1976;Solomon & Toksöz, 1973;Toksőz & Solomon, 1973;Wood, 1972). These models show that the deep mantle becomes extensively molten and the lithosphere becomes as thick as around 600 km at present.…”

“…We assumed that the solid phase is always denser than the melt phase. Some earlier studies, however, suggest that IBC‐rich magma can be denser than the coexisting matrix in the deep mantle of the Moon (e.g., Sakamaki et al., 2010; van Kan Parker et al., 2012; Xu et al., 2022). We also examined how this melt‐solid density inversion can influence our numerical results (Section 3.4).…”

“…This density inversion does not influence the overall upwelling flow, because melting of the mantle always causes a decrease in the bulk density (see Movie S10); the density inversion just lets magma migrate downward within upwelling partially molten plumes in the deep mantle. Xu et al, 2022). In (b), the black dashed line is the solidus curve for the lunar mantle materials (Garcia et al, 2011(Garcia et al, , 2012Katz et al, 2003).…”

“…10.1029/2023JE007845 20 of 29 mantle (e.g., Laneuville et al, 2018;Solomon & Toksöz, 1973;Spohn et al, 2001;U et al, 2022;Wieczorek & Phillips, 2000;Wood, 1972). These models assume that the magma ascends from partially molten regions as deep as 200-800 km to the crust.…”