F. Poitrasson, R. Freydier
Oct 20, 2005
Citations
48
Influential Citations
273
Citations
Quality indicators
Journal
Chemical Geology
Abstract
Abstract High mass resolution multicollector inductively coupled plasma mass spectrometry (MC-ICP-MS) was assessed for iron isotope measurement of natural samples after matrix separation by anion exchange chromatography. No remaining interferences were observed on the plateaus used for the mass spectrometric measurements. The approach developed and the instrument used permitted analyses in the static mode. Various mass bias corrections using Ni doping were tested, and even the assumption of similar fractionation factors for Fe and Ni did not produce significantly inaccurate data. However, the daily regression method between ln 57 Fe / 54 Fe and ln 61 Ni / 60 Ni on the standard reference material IRMM-14 to characterize the instrumental mass bias appeared to give the best precision. The reproducibility observed over four months is about 0.013‰/amu, 2 SD, on both δ 57 Fe / 54 Fe and δ 56 Fe / 54 Fe values, provided that each sample is analyzed at least six times. Accuracy, as estimated on interlaboratory comparison of natural samples that included geostandards, lies within this uncertainty. Among the bulk granitic rocks analysed, those with MgO below 0.6 wt.% and SiO 2 above 71 wt.% have δ 57 Fe / 54 Fe values significantly heavier than the bulk mafic Earth. This shows that the iron isotope composition of terrestrial igneous rocks is more scattered than previously thought. There are good correlations between the Fe isotope composition and the MgO and SiO 2 contents of the granitoids. These correlations are interpreted as reflecting the exsolution of late magmatic aqueous fluids from the granitic melt that preferentially removed the lighter isotopes of iron and enriched the residual magma in the heavier isotopes.