Magnesium isotopic composition of igneous rock standards measured by MC-ICP-MS

To provide inter-lab comparison for high-precision Mg isotope analysis, Mg isotope compositions (expressed as δ26Mg relative to DSM-3) for commercially accessible peridotite, basalt, andesite, and granite geo-standards have been measured by multi-collector inductively coupled mass-spectrometry (Nu-Plasma) using sample-standard bracketing method. There is a large tolerance of matrix cations during the measurement of Mg isotopes, as intensity ratios of 23Na/24Mg and 27Al/24Mg of about 20% only change the δ26Mg by less than 0.1‰, and low 55Mn/24Mg (<0.1) and 58Ni/24Mg (<0.01) do not cause significant mass bias either. Concentration match between samples and standards within 90% is adequate to obtain accurate isotope analysis, which also mitigates the isobaric interference of 12C14N+ on 26Mg. Organic matrix from chemical purification can cause significant analytical errors when the mass of Mg processed is small. The long-term reproducibility of δ26MgDSM-3 for samples with relatively higher MgO content is about 0.11‰ (2SD), and granites with lower MgO content is about 0.2‰ (2SD). Although the standards in this study have wide ranges of major element compositions with SiO2 from 40 to 70 wt.% and MgO from 0.75 to 49.6 wt.%, they exhibit a variation of Mg isotopic compositions with δ26Mg from −0.07 to −0.40‰. δ26Mg do not correlate with SiO2 or MgO contents, suggesting homogenous Mg isotope compositions in igneous rocks at the level of current precision, relative to low temperature samples including sediments and riverine and sea waters. Our data do not support a non-chondritic Mg isotope composition of the Earth.