Evaluating U-Pb accuracy and precision by comparing zircon ages from 12 standards using TIMS and LA-ICP-MS methods

Abstract

U-Pb geochronological studies based on LA-ICP-MS isotopic measurements often require a profound understanding of the accuracy, precision, and preferred ages derived from this methodology. To enhance such awareness, an approximately 60,000 record U-Pb database is compiled from LA-ICP-MS zircon analyses of 12 standards (reference materials), with the standards having ages ranging from 3500 Ma to 0 Ma. Because the 12 standards are dated with the highly accurate and precise TIMS methods, a means exists for assessing the accuracy and precision of corresponding LA-ICP-MS ages. For each standard, the LA-ICP-MS ages are separated by concordance classes, and then age distributions are calculated for each concordance class using four age models: 206Pb/238U, 207Pb/206Pb, IsoplotR concordia, and non-iterative probability ages. Preferred U-Pb ages are determined by the model from which LA-ICP-MS ages consistently align with TIMS ages across the entire age continuum from 3500 Ma to present. The results favor the non-iterative probability model, which indicate that LA-ICP-MS ages have a mean accuracy of ±1.4 myr when comparing median LA-ICP-MS ages with mean TIMS ages; however, LA-ICP-MS ages have relatively poor 2σ internal precision, which is age-dependent and generally ranges from ±3 myr to ±25 myr, with mean 2σ internal precision of 15 myr. Additionally, the results suggest that U-Pb discordance has at least two primary sources: (a) inherent LA-ICP-MS imprecision due to repeated measurement errors of 206Pb/238U and 207Pb/206Pb ratios, is likely the primary cause of proximal-Wetherill discordance, whereas (b) Pb-loss is likely the primary contributor to distal-Wetherill discordance.