Calibrating volatile loss from the Moon using the U-Pb systemOPEN ACCESS
J.N. Connelly, A.A. Nemchin, R.E. Merle, J.F. Snape, M.J. Whitehouse, M. Bizzarro
Geochimica et Cosmochimica Acta
Available online 2 March 2022
“Previous isotope studies of lunar samples have demonstrated that volatile loss was an important part of the early history of the Moon. The radiogenic U-Pb system, where Pb has a significantly lower T50% condensation temperature than U, has the capacity to both recognize and calibrate the extent of volatile loss but this approach has been hindered by terrestrial Pb contamination of samples. We employ a novel method that integrates analyses of individual samples by Ion Microprobe and Thermal Ionization mass spectrometry to correct for ubiquitous common Pb contamination, a method that results in significantly higher estimates for µ-values (238U/204Pb) than previously reported. Using this method, six of seven samples of low-Ti basaltic meteorites return µ-values between 1900 and 9600, values that are consistent with a re-evaluation of published results that return µ-values of 510-2900 for both low- and high-Ti basalts. While some degree of fractionation during partial melting may increase µ-values, we infer that the source region(s) for the basalts must also have had elevated µ-values by the time the lunar magma ocean solidified. Models to account for the available initial Pb isotopic compositions of lunar basalts in light of timing constraints from thermal modelling imply that their source regions had a µ-value of at least 280, consistent with the elevated µ-values of lunar basalts and that inferred for their sources. Such high µ-values are attributed to the preferential loss of more than 99% of the Pb over U relative to a precursor with a Mars-like composition in the aftermath of the giant impact. Such an extensive loss of Pb is consistent with previously reported losses of other elements with comparable volatility, namely Zn, Rb, Ga and CrO2. Finally, our modelling of highly-radiogenic lunar Pb isotopes assuming crystallization of the lunar magma ocean over 10’s of millions of years shows that the elevated µ-values allows for, but does not require, a young Moon formation age.”
