Chemical-abrasion SIMS dating of zircon from the Eocene Caetano caldera, Nevada

The Eocene Caetano caldera in northern Nevada formed during eruption of ~1100 km³ of crystal-rich rhyolite. Miocene extension cut the caldera into a set of fault blocks that expose minor pre-caldera volcanic rocks, two units of intracaldera Caetano Tuff up to 4 km thick, ash-flow tuff feeder dikes and ring-fracture intrusions, caldera collapse breccias, and post-collapse resurgent intrusions. Single-crystal ⁴⁰Ar/³⁹Ar sanidine dates on all parts of the caldera system overlap, yielding a 34.01 ± 0.05 Ma (n=17, Fish Canyon sanidine = 28.201 Ma) age for the eruption. ⁴⁰Ar/³⁹Ar dating also documents several preceding episodes of magmatism: 35.69 ± 0.06 Ma (sanidine, n =13) rhyolite dikes in the nearby Cortez gold district, 35.21 ± 0.18 Ma (plagioclase, n=1) andesite lava underlying Caetano Tuff, and a 38.90 ± 0.11 Ma (biotite, n=1), dacite dike in the northeastern caldera wall. Extensive U-Pb SHRIMP dating of zircon from both the Cortez dikes and all phases of the Caetano system suggests continuous magmatism from 40-34 Ma. However, all samples contain at least some—sometimes many—zircons with U-Pb ages younger than the 34.0 Ma argon age. To determine if anomalously young zircon ages are due to Pb-loss, we analyzed representative samples of the upper Caetano Tuff and the Redrock Canyon resurgent pluton with and without chemical abrasion to mitigate Pb-loss. Bulk zircon separates were annealed at 850°C for 48 hours, then chemically abraded with 10:1 HF/HNO₃ vapor in a Parr bomb at 225°C for 8 hours, based on protocols outlined by Mattinson (2005). Both treated and untreated zircons from the same sample were mounted in epoxy and polished to their midsections, then imaged on the SEM using BSE and CL. The SHRIMP-RG at Stanford University was used to determine U-Pb ages and trace element concentrations in single spots for ~25 to 30 individual zircons per sample, using a round-robin procedure and two zircon age standards (R33 and 080) to monitor external precision. Analyses revealed distinctly different age populations for the abraded and untreated zircons. The chemically abraded populations yielded unimodal zircon age distributions with mean ages that overlap with the ⁴⁰Ar/³⁹Ar age. Untreated zircon populations yielded mean ages 0.9-1.5 Ma younger than the ⁴⁰Ar/³⁹Ar. In the untreated populations, 50-60% of zircon ages are younger than 34.0 Ma at 1σ, versus 15-20% in the chemically abraded populations. Comparison of trace element data from treated and untreated populations indicates that trace element concentrations are apparently unaffected by the chemical abrasion procedure. Further experiments are underway, but we tentatively conclude that chemical abrasion is effective for removing damaged Pb-loss portions of zircons while still enabling high spatial resolution U-Pb dating and trace element analysis. It appears to be a relatively fast and low-cost way to improve the accuracy of SIMS dating of large populations of zircon from Tertiary and older plutonic and volcanic rocks where Pb-loss is frequently an issue.