Item

Abstract

In situ isotope ratio measurements provide the opportunity to detect variations at the microscale and to correlate them with textures. However, in situ approaches such as secondary ion mass spectrometry (SIMS) require matrix-matched reference materials due to analytical artefacts. On top of structural state and chemical composition, orientation of the crystal during measurements can create an additional bias for minerals such as magnetite. Here, we present a new magnetite SIMS reference material PF21mt that has a δ18OVSMOW value of -1.97 ± 0.18‰ (n = 6, 2 standard deviation – 2s – laser fluorination data) and is homogeneous in its chemical composition regarding FeO, MgO, Cr2O3, Al2O3, TiO2, MnO, NiO and ZnO. SIMS measurements of multiple chips of PF21mt in two different laboratories revealed that each chip is homogeneous in δ18O with a repeatability varying from ± 0.08 to 0.62‰ (2s). However, the repeatability of multiple δ18O measurements done under similar measurement conditions on randomly oriented chips of the same magnetite grain varied between ± 1.80 and ± 2.18‰ (2s). These results suggest a crystal orientation bias of up to ± 2‰ (2s) when using the conventional 20 keV protocol. This was further confirmed by variations in the repeatability and relative offsets in δ18O values between different magnetite chips, when rotating the samples in the SIMS holder. The repeatability of SIMS δ18O measurements improved to ± 1‰ (2s) when using a lower energy protocol at 13 keV, suggesting a smaller extent of orientation bias. We conclude that crystal orientation bias limits the precision of magnetite δ18O measurements obtained using SIMS. However, the relatively large oxygen isotopic fractionation between magnetite and silicates still justifies the use of in situ measurements for many geological applications.