Some Observations Concerning U-Pb Data Quality Based on

the G-Chron Proficiency Testing Scheme

Michael Wiedenbeck¹

¹GFZ Helmholtz Centre for Geosciences, Potsdam, Germany

Starting in 2019, the International Association of Geoanalysts has run a Proficiency Testing programme designed to aid geochronology laboratories test their U-Pb zircon data for unsuspected bias. Test samples are dispatched world-wide to roughly 70 laboratories at roughly 2-year intervals. To date, three such rounds have been completed; this talk will provide some insights based the results that have been reported. The three zircon materials that have been distributed range in age from Late Archean to Mesozoic, with each material having its U-Pb homogeneity tested using the Potsdam SIMS instrument. Subsequently the target ²⁰⁶Pb/²³⁸U and ²⁰⁷Pb/²⁰⁶Pb values were defined by the results pooled from between three and seven ID-TIMS laboratories. From the rank distribution plots from the 70-odd laboratories it can be seen that roughly 20 % of those reporting have significant biases in their reported ²⁰⁶Pb/²³⁸U values. Likewise, for the two older samples it is also apparent that around 20 % of the reported ²⁰⁷Pb/²⁰⁶Pb results differ substantially from the ID-TIMS-defined target values. Another important observation is that the bias from the target values correlates poorly with the reported analytical uncertainties. A further surprise from the submitted data is that quadrupole ICPMS instruments generally report more precise results as compared to sector instruments.

The ⁸⁷Rb-⁸⁷Sr system is a versatile chronological tool and geochemical tracer. Unlike Rb-Sr measurements by thermal ionization mass spectrometry, in-situ laser ablation (LA) mass spectrometry offers high sample throughput and minimizes sample destruction. This method, however, is compromised by the presence of multiple isobaric interferences on Sr isotopes. This problem can be overcome using multi-collector inductively-coupled-plasma mass spectrometers (MC-ICP-MS) equipped with the recently developed pre-cell mass filter and collision/reaction cell (CRC), such as the Thermo Scientific^TM Neoma^TM MS/MS-MC-ICP-MS. This instrument allows the simultaneous measurements of on-mass and mass-shifted Rb and Sr isotopes using SF₆ reaction-gas [1-4]. Here we show, however, that for Y- and Zr-bearing phases, ⁸⁹Y¹⁶O⁺, ⁹⁰Zr¹⁶O⁺, and ⁹¹Zr¹⁶O⁺ can form in the CRC, when O₂/H₂O enters the CRC either from impurities in the reaction gas or molecules that formed in the plasma. These oxides interfere with ⁸⁶SrF⁺, ⁸⁷SrF⁺, and ⁸⁸SrF⁺, respectively, and cause systematic errors on measured ⁸⁷Rb/⁸⁶Sr and ⁸⁷Sr/⁸⁶Sr ratios. To correct for these effects, we synthesized glasses, doped with Rb, Sr, Y and Zr. The glasses can be used to correct for YO⁺- and ZrO⁺-induced systematic errors on measured ⁸⁷Rb/⁸⁶Sr and ⁸⁷Sr/⁸⁶Sr ratios. The effects of these interferences and their correction, using our synthesized glasses, on in-situ Rb-Sr dating using LA-MC-ICP-MS/MS will be discussed at the meeting.

[1] Craig et al. (2021), Anal. Chem. 93, 30, 10519-10527;

[2] Dauphas et al. (2022), JAAS 37, 2420-2441;

[3] Telouk et al. (2024), JAAS 39, 879-887;

[4] Huang et al. (2025), Spectrochim. Acta B 224, 107117

Fossils that provide three-dimensional anatomical detail are among the most spectacular and significant archives of past life. However, the duration and speed of tissue mineralization often remain obscure, as do the sources of the fossilizing agents. We present a novel approach to reconstruct the deep-time 3D silicification of plant tissues, using Late Pennsylvanian fossil wood from fluvial red beds of the Kyffhäuser (Siebigerode Formation, Saale Basin, central Germany). Methods applied include fossil-wood histology, sediment petrography, cathodoluminescence (CL), scanning-electron microscopy, Raman spectroscopy, electron-probe and fluid-inclusion microanalysis, LA-ICP-MS in-situ U-Pb dating and Si isotopy of silicified fossil wood. The results reveal a 200-myr history of mineralization that occurred in concert with regional basin evolution. Accordingly, initial permineralization through amorphous silica is dated to 304±10 Ma and sourced from pyroclasts related to syndepositional, late-Variscan volcanism. This stage was followed by the opal-quartz transformation during subsidence until c. 290 Ma (early Permian). Around 260 Ma, hydrothermal quartz-hematite mineralization impacted the woods following middle–late Permian inversion. Blocky quartz sourced from basinal brines crystallized in the fossil trunks during maximum burial between 180–160 Ma (Late Jurassic), at 3.5–5.0 km depth and 160–240°C. The youngest mineralization found in the trunks dates to 115 Ma (Early Cretaceous) and resulted from regionally traceable, hydrothermal quartz-baryte-calcite mineralization in the course of inversion in the Central European Basin System. These outcomes not only break new ground in understanding cell-scale fossilization but also reveal silicified plants as promising tools for reconstructing basin evolution and regional mineralization.

Uranium contents in natural garnet are typically in the range of 1–50 μg/g in andraditic garnet found in alkali-magmatic rocks and skarns, but are orders of magnitude lower in metamorphic almandine–pyrope with typically 1–100 ng/g. Nonetheless, the instrumental setup at FIERCE coupling laser-ablation with a modern MC-ICPMS system enables us to routinely extract U–Pb ages even from samples of the latter group at a geologically meaningful level of precision. However, limited knowledge of the mechanisms and systematics of U incorporation into garnet hamper the processes of sample selection and data interpretation for garnet petrochronology.

We find that U is predominantly incorporated into both andraditic and almandine–pyrope garnet in its tetravalent form and that it strongly favours the dodecahedral X site of garnet. The elbrusite substitution, which includes hexavalent U does not operate at the trace-element level of U in natural garnet. Instead, the incorporation of tetravalent U into the dodecahedral X-site dominates and is charge balanced through tetrahedral ferric Fe and/or through divalent cations in the Y-site. Many andradite samples show oscillatory zoning with a covariation of U, Th and the light REE, demonstrating that incorporation of the lanthanides and actinides may be controlled by the same crystal chemical mechanisms. Some samples additionally show clear sector zoning in U (and Th and Ce) with a 20–50 % enrichment of U in the {211} sectors compared to the {110} sectors, demonstrating a control of crystal growth rate on U incorporation.