40Ar/39Ar age of plagioclase from Acapulco meteorite and the problem of systematic errors in cosmochronology

Acapulcoites consistently yield some of the oldest 40Ar/39Ar ages reported for any meteorites, all indistinguishable from the age of 4510±22 Ma previously determined for whole-rock samples of Acapulco. These ages are significantly younger, at face value, than the 4557±2 Ma Pb/Pb age reported for phosphates, which dates cooling through ∼550°C. Metallographic evidence for rapid cooling between 650 and 350°C, retention of 129Xe in apatite, and the absence of shock features all suggest that the 40Ar/39Ar ages might be expected to be more consistent with the Pb/Pb age. New 40Ar/39Ar data from plagioclase, determined to be the sole K-rich phase in Acapulco, yield a high-precision isochron age of 4507±18 Ma (2σ) based on currently accepted decay constants and standards but neglecting uncertainties in these quantities. These data allow unambiguous association of the 40Ar/39Ar age with a closure temperature near 300°C and preclude the possibility of artifacts due to 39Ar recoil, low-temperature alteration products, or inherited pre-solar grains. The 40Ar/39Ar plagioclase age recalculated with decay constants believed to be more accurate than those conventionally used is 4554 Ma, weakening the basis of a previously inferred dramatic decrease in cooling rate shortly after closure of the Pb/Pb system in phosphate. Systematic errors in 40Ar/39Ar dating, associated chiefly with 40K decay constants and standards, must be redressed before the method can be applied to high-resolution inference of early Solar System evolution. Similar limitations affect ages based on other long-lived radionuclides such as 147Sm.