An investigation of factors affecting high-precision Sr isotope analyses (87Sr/86Sr and δ88/86Sr) by MC-ICP-MS

The stable strontium (Sr) isotope system is being increasingly utilized to move forward our understanding of environmental, geological, and cosmological processes. Two analytical techniques are commonly used to mea-sure stable Sr isotopes: 1) double-spike thermal ionization mass spectrometry (DS-TIMS) and 2) Zr-doped sample -standard bracketing multi-collector inductively coupled plasma mass spectrometry (Zr-doped SSB via MC-ICP-MS). Relative to DS-TIMS, Zr-doped SSB via MC-ICP-MS allows simultaneous determination of both 87Sr/86Sr and 88Sr/86Sr ratios, increasing measurement efficiency and sample throughput. However, several factors limit the analytical precision and accuracy of this technique, including matrix effects and potential intra-and inter -session offsets of measured isotope ratios. In this study, we evaluate potential causes for the observed offsets. We performed a series of experiments to assess the effect of different solution matrices and sources of contamination on Sr isotope measurements and to explore potential underlying mechanisms. Our tests show that incomplete Sr recovery during chromatographic separation, mismatches of Sr and Zr concentrations and acid molarity between samples and bracketing standards, and cation contamination all affect the precision and ac-curacy of Sr isotope measurements. We present evidence that analyses using our recommended updated prep-aration procedure for Zr-doped SSB via MC-ICP-MS can achieve a long-term reproducibility (2 sigma SD: 87Sr/86Sr = +/- 0.000015 and delta 88/86Sr = +/- 0.03 parts per thousand) comparable to that of DS-TIMS.