Abstract The performance of multi-collector secondary ion mass spectrometry (MC-SIMS) for Mg isotope ratio analysis was evaluated using 17 olivine and 5 pyroxene reference materials (RMs). The Mg isotope composition of these RMs was accurately and precisely determined by multi-collector inductively coupled plasma mass spectrometry (MC-ICP-MS), and these measured isotope ratios were used to evaluate SIMS instrumental mass bias as a function of the forsterite (Fo) content of olivine. The magnitude of the Mg isotope matrix effects were ~3‰ in δ25Mg, and are a complex function of olivine Fo content, that ranged from Fo59.3 to Fo100. In addition to these Mg isotope matrix effects, Si+ ion yields and Mg+/Si+ ion ratios varied as a complex function of the Fo content of the olivine RMs. For example, Si+ ion yields varied by ~33%. Based on the observations, we propose instrumental bias correction procedures for SIMS Mg isotope analysis of olivine using a combination of Mg+/Si+ ratios and Fo content of olivine. Using this correction method, the accuracy of δ25Mg analyses is 0.3‰, except for analysis of olivine with Fo86–88 where instrumental biases and Mg+/Si+ ratios change dramatically with Fo content, making it more difficult to assess the accuracy of Mg isotope ratio measurements by SIMS over this narrow range of Fo content. Five pyroxene RMs (3 orthopyroxenes and 2 clinopyroxenes) show smaller ranges of instrumental bias (~1.4‰ in δ25Mg) as compared to the olivine RMs. The instrumental bias for the 3 orthopyroxene RMs do not define a linear relationship with respect to enstatite (En) content, that ranged from En85.5–96.3. The clinopyroxene RMs have similar En and wollastonite (Wo) contents but have δ25Mg values that differ by 0.5‰ relative to their δ25Mg values determined by MC-ICP-MS. These results indicate that additional factors (e.g., minor element abundances) likely contribute to SIMS instrumental mass fractionation. In order to better correct for these SIMS matrix effects, additional pyroxene RMs with various chemical compositions and known Mg isotope ratios are needed.

中文翻译：

---

通过 SIMS 分析橄榄石和辉石的镁同位素：评估基质效应

摘要 使用 17 种橄榄石和 5 种辉石参考材料 (RM) 评估了多接收器二次离子质谱 (MC-SIMS) 对镁同位素比分析的性能。这些 RM 的 Mg 同位素组成通过多接收器电感耦合等离子体质谱法 (MC-ICP-MS) 准确而精确地确定，这些测量的同位素比率用于评估 SIMS 仪器质量偏差作为镁橄榄石 (Fo ) 橄榄石的含量。δ25Mg 中 Mg 同位素基体效应的大小约为 3‰，并且是橄榄石 Fo 含量的复杂函数，范围从 Fo59.3 到 Fo100。除了这些 Mg 同位素基质效应之外，Si+ 离子产率和 Mg+/Si+ 离子比作为橄榄石 RM 的 Fo 含量的复杂函数而变化。例如，Si+ 离子产率变化约 33%。根据观察，我们建议使用 Mg+/Si+ 比率和橄榄石的 Fo 含量的组合对橄榄石进行 SIMS Mg 同位素分析的仪器偏差校正程序。使用这种校正方法，δ25Mg 分析的准确度为 0.3‰，除了用 Fo86-88 分析橄榄石外，仪器偏差和 Mg+/Si+ 比随 Fo 含量发生显着变化，从而更难评估 Mg 同位素比测量的准确性通过 SIMS 在这个狭窄的 Fo 含量范围内。与橄榄石 RM 相比，五个辉石 RM（3 个斜辉石和 2 个单斜辉石）显示出较小的仪器偏差范围（δ25Mg 中约为 1.4‰）。3 种斜方辉石 RM 的仪器偏差未定义与顽火石 (En) 含量的线性关系，其范围为 En85.5–96.3。单斜辉石 RM 具有相似的 En 和硅灰石 (Wo) 含量，但其 δ25Mg 值与其通过 MC-ICP-MS 确定的 δ25Mg 值相差 0.5‰。这些结果表明其他因素（例如，微量元素丰度）可能有助于 SIMS 仪器质量分馏。为了更好地校正这些 SIMS 基质效应，需要具有各种化学成分和已知 Mg 同位素比的额外辉石 RM。