Abstract
We document the utility for in situ Rb-Sr dating of a one-of-a-kind tribrid mass spectrometer, 'Proteus', coupled to a UV laser ablation system. Proteus combines quadrupole mass-filter, collision cell and sector magnet with a multicollection inductively-coupled plasma mass spectrometer (CC-MC-ICPMS/MS). Compared to commercial, single collector, tribrid inductively-coupled plasma mass spectrometers (CC-ICPMS/MS) Proteus has enhanced ion transmission and offers simultaneous collection of all Sr isotopes using an array of Faraday cups. These features yield improved precision in measured 87Sr/86Sr ratios, for a given mass of Sr analysed, approximately a factor of 25 in comparison to the Thermo Scientific™ iCAP TQ™ operated under similar conditions. Using SF6 as a reaction gas on Proteus, measurements of Rb-doped NIST SRM (standard reference material) 987 solutions, with Rb/Sr ratios from 0.01-100, yield 87Sr/86Sr that are indistinguishable from un-doped NIST SRM 987, demonstrating quantitative 'chemical resolution' of Rb from Sr. We highlight the importance of mass-filtering before the collision cell for laser ablation 87Sr/86Sr analysis, using an in-house feldspar standard and a range of glass reference materials. By transmitting only those ions with mass-to-charge ratios 82-92 u/e into the collision cell, we achieve accurate 87Sr/86Sr measurements without any corrections for atomic or polyatomic isobaric interferences. Without the pre-cell mass-filtering, measured in situ 87Sr/86Sr ratios are inaccurate. Combining in situ measurements of Rb/Sr and radiogenic Sr isotope ratios we obtain mineral isochrons. We utilise a sample from the well-dated Dartmoor granite (285 ± 1 Ma) as a calibrant for our in situ ages and, using the same conditions, produce accurate Rb-Sr isochron ages for samples of the Fish Canyon tuff (28 ± 2 Ma) and Shap granite pluton (397 ± 1 Ma). Analysing the same Dartmoor granite sample using identical laser conditions and number of spot analyses using the Thermo Scientific™ iCAP TQ™ yielded an isochron slope 5× less precise than Proteus. We use an uncertainty model to illustrate the advantage of using Proteus over single collector CC-ICPMS/MS for in situ Rb-Sr dating. The results of this model show that the improvement is most marked for samples that have low Rb/Sr (<10) or are young (<100 Ma). We also report the first example of an in situ, internal Rb-Sr isochron from a single potassium-feldspar grain. Using a sample from the Shap granite, we obtained accurate age and initial 87Sr/86Sr with 95% confidence intervals of ±1.5% and ±0.03% respectively. Such capabilities offer new opportunities in geochronological studies.