[S-CG57] ハイブリッド年代学ー年代値の意味とは？ー

[SCG57-06] Limitations on the accuracy of in situ U-Pb dating

*Axel Gerdes¹ (1.Goethe Universty Frankfurt, Germany)

キーワード：Zircon, U-Pb, LA-ICP-MS, downhole fractionation, garnet, carbonate

The direct dating of geological events, e.g. by U-Pb analysis of zircon, has grown exponentially in the last 30 years (e.g. ISI Web of Science). New analytical techniques and new types of mass spectrometers made this possible. The impetus came with the development of SHRIMP (1978) and laser probes coupled to ICP-MS (1990). Driven by cost and speed, LA-ICP-MS in particular has evolved and is now often the method of choice. As LA-ICP-MS labs worldwide grow, U-Pb ID-TIMS labs are decreasing significantly. Optimized and automated, the LA-ICP-MS allows 10x or 150x higher sample throughput compared to SHRIMP or ID-TIMS. The method's weakness lies in evaluation and interpretation of the data; complex data sets and large numbers of analyses quickly become a bottleneck. The diversity of hardware, software, ablation parameters and reference materials used and the level of experience are also weak points. The accuracy limit of the method appears to be 2-3% (²⁰⁶Pb/²³⁸U), as shown by inter-laboratory comparisons on gem quality zircon. This is in contrast to the frequently quoted age uncertainties of even below 1% in many studies. Natural zircons, especially detrital ones, are often not gem-quality. They show large variations in Pb/U downhole fractionation, which is positively correlated with the alpha dose of the grains. A higher alpha dose means a higher ablation rate. However, ablation rates, and therefore downhole fractionation, can vary even between gem quality zircon used as reference material (RM). This is due to differences in optical properties and is enhanced at high and very low energy densities, i.e. close to the ablation threshold. It is therefore problematic to standardise to an RM zircon assuming identical Pb/U fractionation. The expected inaccuracies can be up to 10%, or even more. The individual correction of the Pb/U fractionation of each individual analysis or of the same grains leads to much more consistent and accurate results. The analysis of metamict areas that have lost or gained Pb is of course also problematic. Unlike zircon, many other minerals, e.g. carbonates, sulphates and garnets, show less variability in Pb/U fractionation during ablation of different samples of each group. In addition to laser-induced fractionation, plasma-induced Pb/U fractionation occurs due to variations in mineral composition and ablation rates/volumes. It is still poorly understood and has not yet been systematically studied in detail. Results of Pb/U fractionation in various minerals as a function of spot size, ablation rate and energy density are presented and discussed. They are based on investigations using a 193 nm Excimer laser (RESOlution-LR) coupled to an Element XR ICPMS.
There are also systematic sources of error, such as incorrect determination of detector dead time and SEM plateau voltage, and calibration of the dual-mode SEM (ACF or PA factor). Especially the latter, when relying on the software algorithms of the ICP-MS manufacturer, can lead to inaccuracies of 4-10%. Provided that the Pb and/or U isotopes of RM and samples have been measured in the different modes.
After about 30 years of LA ICPMS U-Pb dating, studies of inter-element fractionation during ablation and ionisation in the ICP are still of great importance. In order to make the corrections more reliable and to better quantify the associated uncertainties. It is hoped that this contribution will stimulate further research studies.