Argon‐40/argon‐39 age spectra of Apollo 17 highlands breccia samples by laser step heating and the age of the Serenitatis basin

We have obtained high‐resolution (21–63 steps) 40 Ar/ 39 Ar age spectra using a continuous laser system on 19 submilligram samples of melt rocks and clasts from Apollo 17 samples collected from the pre‐Imbrian highlands in the easternmost part of the Serenitatis basin. The samples include poikilitic melt rocks inferred to have been formed in the Serenitatis basin‐forming impact, aphanitic melt rock whose compositions vary and whose provenance is uncertain, and granulite, gabbro, and melt clasts. Three of the poikilitic melts have similar age spectrum plateau ages (72395,96, 3893 ± 16 Ma {2σ} 72535,7, 3887 ± 16 Ma; 76315,150, 3900 ± 16 Ma) with a weighted mean age of 3893 ± 9 Ma, which we interpret as the best age for the Serenitatis basin‐forming impact. Published 40 Ar/ 39 Ar age spectrum ages of Apollo 17 poikilitic melts are consistent with our new age but are much less precise. Two poikilitic melts did not give plateaus and the maxima in their age spectra indicate ages of ≥3869 Ma (72558,7) and ≥3743 Ma (77135,178). Plateau ages of two poikilitic melts and two gabbro clasts from 73155 range from 3854 ± 16 Ma to 3937 ± 16 Ma and have probably been affected by the ubiquitous (older?) clasts and by post‐formation heating (impact) events. Plateau ages from two of the aphanitic melt “blobs” and two granulites in sample 72255 fall in the narrow range of 3850 ± 16 Ma to 3869 ± 16 Ma with a weighted mean of 3862 ± 8 Ma. Two of the aphanitic melt blobs from 72255 have ages of 3883 ± 16 Ma and ≥3894 Ma, whereas a poikilitic melt clast (of different composition from the “Serenitatis” melts) has an age of 3835 ± 16 Ma, which is the upper limit for the accretion of 72255. These data suggest that either the aphanitic melts vary in age, as is also suggested by their varying chemical compositions, or they formed in the 72255 accretionary event about 3.84–3.85 Ga and older relict material is responsible for the dispersion of ages. In any case the aphanitic melts do not appear to be Serenitatis products. Our age for the Serenitatis impact shows, on the basis of the isotopic age evidence alone, that Serenitatis is >20–25 Ma and probably >55–60 Ma older than Imbrium (≤3870 Ma and probably ≤3836 Ma [ Dalrymple and Ryder , 1993]). Noritic granulite sample 78527 has a plateau age of 4146 ± 17 Ma, representing a minimum age for cooling of this sample in the early lunar crust. So far there is no convincing evidence in the lunar melt rock record for basin‐forming impacts significantly older than 3.9 Ga.