δ<scp>²⁶Mg</scp> Values of Thirty‐One Geological Reference Materials Analysed by the Critical Mixture Double Spike Technique

Stable Mg delta values (the relative deviation from a certain reference material, e.g., DSM‐3 here and expressed as δ 26 Mg DSM‐3 ) were routinely measured by the sample‐standard bracketing (SSB) method on a multi‐collector ICP‐MS, as only three isotopes (i.e., 24 Mg, 25 Mg and 26 Mg) naturally exist. Due to potential inaccuracy in correcting mass bias during measurements, considerable measurement bias has been reported among laboratories. Recently, a critical mixture double spike (CMDS) technique has been developed and demonstrated to accurately correct mass bias of measurement results with precision of ± 0.03‰ for δ 26 Mg. Here, we measured thirty‐one geological reference materials including igneous, metamorphic and sedimentary rocks, sediments and minerals using the CMDS technique, with the purpose of better characterising their δ 26 Mg DSM‐3 values. Aligning with the data previously reported, uncorrected bias, on average measured by ∆ 26 Mg SSB‐CMDS (i.e., δ 26 Mg SSB ‐ δ 26 Mg CMDS ) as ‐0.071 ± 0.092‰ (2 s , n = 42), has been reaffirmed for the traditional SSB method. Such uncorrected bias positively correlates with sample Mg/(Si+Al+Ca), and thus may result from the accumulative effect of residual matrix elements. The new data set herein can aid future inter‐laboratory comparison and data quality control.