Reassessing the uranium decay constants for geochronology using ID-TIMS U–Pb data

As the internal precision of radiometric dates approaches the 0.1% level, systematic biases between different methods have become apparent. Many workers have suggested that calibrating other decay constants against the U–Pb system is a viable solution to this problem. We test this assertion empirically and quantitatively by analyzing U–Pb systematics of zircon and xenotime on the single- to sub-grain scale by high-precision ID-TIMS geochronology on 11 rock samples ranging from 0.1 to 3.3 Ga. Large statistically equivalent datasets give 207Pb/ 206Pb dates that are systematically older than 206Pb/ 238U dates by ∼0.15% in Precambrian samples to as much as ∼3.3% in Mesozoic samples, suggesting inaccuracies in the mean values of one or both of the U decay constants. These data are used to calculate a ratio of the U decay constants that is lower than the accepted ratio by 0.09% and is a factor of 5 more precise. Four of the samples are used to augment existing data from which the U–Pb and 40Ar/ 39Ar systems can be compared. The new data support most previous observations that U–Pb and 207Pb/ 206Pb dates are older than 40Ar/ 39Ar by ⩽1%, though scatter in the amount of offset in samples as a function of age suggests that the bias is not entirely systematic, and may incorporate interlaboratory biases and/or geologic complexities. Studies that calibrate other decay schemes against U–Pb should include an assessment of inaccuracies in the U decay constants in addition to other systematic biases and non-systematic geologic uncertainty.