• High-precision mass-dependent Ni isotope data are presented for Apollo lunar basalts and dunite 72415. • Diffusion modelling accounts for the extreme Ni isotope fractionation in lunar dunite 72415 (δ⁶⁰/⁵⁸Ni = +1.80‰). • Lunar mare basalts exhibit significant δ⁶⁰/⁵⁸Ni variability not primarily controlled by magmatic differentiation. • PMELTS modelling indicates Ni isotope fractionation during lunar magma ocean crystallization. • The estimated bulk silicate Moon δ⁶⁰/⁵⁸Ni (0.18 ± 0.20‰) overlaps with published bulk silicate Earth values (∼0.11‰). Although the Moon is thought to have formed through a giant impact between proto-Earth and a Mars-sized body, the processes responsible for the chemical and mass-dependent isotopic differences between Earth and Moon remain debated. We report high-precision mass-dependent Ni isotope data for 19 Apollo samples, including one dunite (72415), fifteen low-Ti basalts, and three high-Ti basalts, analyzed by double-spike technique using a multi-collector plasma-sourced mass spectrometer. The dunite 72415 shows an extremely high δ 60/58 Ni value of +1.80 ± 0.01‰, which we attribute to kinetic isotope fractionation from Ni diffusion during re-equilibration between olivine and a later melt. Diffusion modeling of Ni–Fe–Mg systematics reproduces the observed heavy Ni enrichment. In contrast, low-Ti basalts display a mean δ 60/58 Ni of 0.23 ± 0.20‰ (2SD), unaffected by cosmic-ray exposure, while high-Ti basalts are slightly isotopically lighter (0.06 ± 0.22‰, 2SD). Petrological modeling using pMELTS with recently constrained silicate mineral-melt fractionation factors suggests limited Ni isotope fractionation (<0.05‰) during lunar magma ocean crystallization and partial melting, yielding an estimated bulk silicate Moon (BSM) δ 60/58 Ni = 0.18 ± 0.20‰ (2SD). This overlaps with the bulk silicate Earth (BSE: 0.11 ± 0.07‰), indicating that Ni depletion in the lunar mantle, by a factor of ∼4 relative to Earth, can be caused by core formation (that does not fractionate Ni isotopes). However, our modelling shows evaporative loss of Ni can elevate δ 60/58 Ni value of < 0.23‰, which remains consistent with those of BSM within uncertainty. Hence, the mechanism of Ni evaporation cannot be ruled out.
Zhu et al. (Fri,) studied this question.