Argon diffusion in Apollo 16 impact glass spherules: Implications for 40Ar/39Ar dating of lunar impact events

Abstract

The 40Ar/39Ar technique applied to impact glass has been used to date both terrestrial and lunar impact events. The ability to utilize the 40Ar/39Ar technique rests on the assumption that impact glasses are closed to the loss of daughter product, 40Ar∗, after formation. Diffusion experiments were performed on three Apollo 16 lunar impact glasses and yielded activation energies for 39Ar of ∼17 to 20kcalmol−1 and log10(D0/a2) values of −5.2 to −6.0 s−1. The resulting diffusion coefficients are interpreted as minimum values and the Apollo 16 glass is probably some of the least retentive of lunar glasses, as the degree of non-bridging oxygen is at one end of the range in lunar glasses. At temperatures below the glass transition temperature (i.e., ∼660°C), the data can be explained by volume diffusion from a single diffusion domain. Modeling shows that Apollo 16 composition glass could lose significant quantities of radiogenic argon (40Ar∗) (∼90–100% over 20–40Myr assuming a diffusion domain size (a) of 75μm) due to diurnal temperature variations on the lunar surface, although 40Ar∗ loss is highly sensitive to exposure duration and effective diffusion domain size. Modeling shows that loss from transient thermal events (e.g., heating to ∼200°C for 102yr duration) can also cause partial resetting of apparent 40Ar/39Ar ages. In small (a=75μm) glasses a maximum of 50–60% of 40Ar∗ is lost over 4Ga when buried to depths corresponding to temperatures of −15°C. Results indicate that caution should be exercised in interpreting lunar impact glass 40Ar/39Ar ages, as the assumption of closed system behavior may have been violated, particularly in glasses with low fractions of non-bridging oxygen.