Precise isotope determination of sub-microgram Mg by the critical mixture double spike technique and its application to fluid inclusions in halite†

With only three isotopes, Mg isotopic measurements have been routinely performed using the sample-standard-bracketing (SSB) method. When using SSB to correct mass bias, δ²⁶Mg becomes negatively biased with decreasing sample loading mass, which hinders the application of Mg isotopes to fluid inclusions in halite due to their low Mg concentration and extremely high Na/Mg ratio. In this study, we used 0.5 mol per L HNO₃ and AG50W-X8 resin to separate Mg from Na. Even when the initial Na/Mg ratio of samples is as high as 5000, Na can be efficiently removed with almost 100% Mg recovery. This process is accomplished within 5 hours. The residual matrix can be removed using procedures modified from Li et al. (2016). To ensure optimal performance of the separation protocol, reducing the sample loading size is the most straightforward approach. Critical mixture double spike (CMDS) technique enables accurate and precise Mg isotope measurements with reduced sample size, as indicated by near-zero Δ²⁶Mg_{measured-recommended} values for GSP-2, BHVO-2 and high-Na/Mg synthetic solutions at 0.9–5.8 μg of Mg loading. Minimum Mg mass loaded onto the column can reach 0.9 μg, which is 11–22 times lower than the amount (typically 10–20 μg) required for accurate SSB measurements. The precision and accuracy of δ²⁶Mg are better than 0.05% for the established procedures at sub-μg Mg loading. Finally, this method is applied to halite from core ISL1A in the Qarhan Salt Lake, which shows highly variable δ²⁶Mg ranging from −0.663% to −1.222%. Negative fluctuation of these halite δ²⁶Mg values may reflect either water recharge during wetter climatic condition or the precipitation of Mg-bearing minerals (e.g., carnallite) during the evaporation process, validating the potential of Mg isotopes in tracing the evolution of paleolake water.