EXAMINING THE IRON ISOTOPE GEOCHEMISTRY OF OLIVINE IN THE SKAERGAARD LAYERED MAFIC INTRUSION, EAST GREENLAND

Abstract

Despite reports of significant fractionation of iron isotopes occurring under magmatic and hydrothermal conditions, the mechanisms that cause and control fractionation in magmatic systems remain poorly understood. Previous studies suggest that iron isotopes fractionate during fractional crystallization and hydrothermal alteration, but this has yet to be determined in a simple, closed igneous system. The Eocene-aged Skaergaard layered mafic intrusion, East Greenland underwent closed-system evolution, as it cooled and crystallized inward from the floor, roof, and walls, and later underwent hydrothermal alteration, making it a natural laboratory for studying the processes that control iron isotope fractionation under magmatic and hydrothermal conditions. This study focuses on the iron isotope composition of olivine, coexisting phases, and the host rocks of the Skaergaard intrusion, in addition to determining the extent and possible causes of fractionation. Presented here are new iron isotope compositions, relative to igneous rocks ([delta]56FeIgR), for bulk-mineral separates of olivine (n=9), coexisting Fe-Ti oxides (n=9), coexisting pyroxene (n=9), and bulk-rocks (n=6), obtained via multi-collector - inductively coupled plasma - mass spectrometry, for a suite of gabbros from the Layered Series of the Skaergaard intrusion. Samples were selected to represent the range of compositions and the varying extent of alteration within the Layered Series. The [delta]56Fe values range from -0.12 0.05[per mille] to +0.01 ± 0.01[per mille] (2-SE) for olivine, -0.09 ± 0.01[per mille] to +0.51 ± 0.01[per mille] for coexisting Fe-Ti oxides, -0.10 ± 0.03 to -0.01 ± 0.01[per mille] for coexisting pyroxene, and -0.02 ± 0.01[per mille] to +0.03 ± 0.03[per mille] for bulk-rocks. These values mostly fall within, but also vary from the average of mafic- to intermediate-composition crustal igneous rocks (0.00 ± 0.08[per mille]; 2-SD). Oxygen isotope compositions ([delta]18OVSMOW), obtained to investigate potential iron isotope fractionation due to hydrothermal alteration, range from 4.48 ± 0.13[per mille] to 4.77 ± 0.13[per mille] (2-SD) for olivine and 4.12 ± 0.08[per mille] to 5.49 ± 0.14[per mille] for bulk-rock, falling within and below the range for unaltered mafic rocks worldwide (5-6[per mille]). The [delta]56Fe values of unaltered olivine exhibit slight systematic variations in parts of the intrusion, and these changes coincide with slight variations in [delta]56Fe values of coexisting unaltered Fe-Ti oxides and pyroxene, and are interpreted as a reflection of fractionation due to fractional crystallization. Additionally, [delta]56Fe values of altered olivine appear slightly higher than expected, potentially indicating slight fractionation due to hydrothermal alteration. However, the iron isotope composition of the altered olivine fall within the range of fresh olivine, which suggests that no fractionation took place during alteration. In contrast, the [delta]56Fe values of bulk-rocks analyzed here vary by only 0.05[per mille] across the Layered Series with no systematic trends, and do not reflect fractionation due to fractional crystallization or hydrothermal alteration. Inter-mineral iron isotope fractionation factors between olivine and Fe-Ti oxides ([delta]56FeOx-Ol) from the Layered Series range from -0.02[per mille] to +0.54[per mille]. Using theoretical and experimental fractionation factors as a function of temperature, these fractionations are shown to yield, in some cases, temperatures close to those of crystallization, but in other cases produce an inaccurate wide range of temperatures. However, when oxide compositional variations are taken into account in the calculations, more realistic fractionation temperatures are obtained. In general, fractionation factors increase with evolution in the Middle Zone. Iron isotope fractionation factors between olivine and pyroxene ([delta]56FePx-Ol) have a narrower range from -0.08[per mille] to +0.06[per mille], and calculations yield unrealistically high fractionation temperatures. Overall, the unrealistic temperatures obtained are due to mineral compositional variations outside those used in the theoretical and experimental studies, extrapolations done to higher temperatures than those at which experiments were conducted, and in some cases, the effect of inclusions.