Copper stable isotopes as tracers of metal–sulphide segregation and fractional crystallisation processes on iron meteorite parent bodies

Abstract

We report high precision Cu isotope data coupled with Cu concentration measurements for metal, troilite and silicate fractions separated from magmatic and non-magmatic iron meteorites, analysed for Fe isotopes (δ 57Fe; permil deviation in 57Fe/ 54Fe relative to the pure iron standard IRMM-014) in an earlier study ( Williams et al., 2006). The Cu isotope compositions (δ 65Cu; permil deviation in 65Cu/ 63Cu relative to the pure copper standard NIST 976) of both metals (δ 65Cu M) and sulphides (δ 65Cu FeS) span much wider ranges (−9.30 to 0.99‰ and −8.90 to 0.63‰, respectively) than reported previously. Metal–troilite fractionation factors (Δ 65Cu M–FeS = δ 65Cu M − δ 65Cu FeS) are variable, ranging from −0.07 to 5.28‰, and cannot be explained by equilibrium stable isotope fractionation coupled with either mixing or reservoir effects, i.e. differences in the relative proportions of metal and sulphide in the meteorites. Strong negative correlations exist between troilite Cu and Fe (δ 57Fe FeS) isotope compositions and between metal–troilite Cu and Fe (Δ 57Fe M–FeS) isotope fractionation factors, for both magmatic and non-magmatic irons, which suggests that similar processes control isotopic variations in both systems. Clear linear arrays between δ 65Cu FeS and δ 57Fe FeS and calculated Cu metal–sulphide partition coefficients (D Cu = [Cu] metal/[Cu] FeS) are also present. A strong negative correlation exists between Δ 57Fe M–FeS and D Cu; a more diffuse positive array is defined by Δ 65Cu M–FeS and D Cu. The value of D Cu can be used to approximate the degree of Cu concentration equilibrium as experimental studies constrain the range of D Cu between Fe metal and FeS at equilibrium to be in the range of 0.05–0.2; D Cu values for the magmatic and non-magmatic irons studied here range from 0.34 to 1.11 and from 0.04 to 0.87, respectively. The irons with low D Cu values (closer to Cu concentration equilibrium) display the largest Δ 57Fe M–FeS and the lowest Δ 65Cu M–FeS values, whereas the converse is observed in the irons with large values D Cu that deviate most from Cu concentration equilibrium. The magnitudes of Cu and Fe isotope fractionation between metal and FeS in the most equilibrated samples are similar: 0.25 and 0.32‰/amu, respectively. As proposed in an earlier study ( Williams et al., 2006) the range in Δ 57Fe M–FeS values can be explained by incomplete Fe isotope equilibrium between metal and sulphide during cooling, where the most rapidly-cooled samples are furthest from isotopic equilibrium and display the smallest Δ 57Fe M–FeS and largest D Cu values. The range in Δ 65Cu M–FeS, however, reflects the combined effects of partial isotopic equilibrium overprinting an initial kinetic signature produced by the diffusion of Cu from metal into exsolving sulphides and the faster diffusion of the lighter isotope. In this scenario, newly-exsolved sulphides initially have low Cu contents (i.e. high D Cu) and extremely light δ 65Cu FeS values; with progressive equilibrium and fractional crystallisation the Cu contents of the sulphides increase as their isotopic composition becomes less extreme and closer to the metal value. The correlation between Δ 65Cu M–FeS and Δ 57Fe M–FeS is therefore a product of the superimposed effects of kinetic fractionation of Cu and incomplete equilibrium between metal and sulphide for both isotope systems during cooling. The correlations between Δ 65Cu M–FeS and Δ 57Fe M–FeS are defined by both magmatic and non-magmatic irons record fractional crystallisation and cooling of metallic melts on their respective parent bodies as sulphur and chalcophile elements become excluded from crystallised solid iron and concentrated in the residual melt. Fractional crystallisation processes at shallow levels have been implicated in the two main classes of models for the origin of the non-magmatic iron meteorites; at (i) shallow levels in impact melt models and (ii) at much deeper levels in models where the non-magmatic irons represent metallic melts that crystallised within the interior of a disrupted and re-aggregated parent body. The presence of non-magmatic irons with a range of Fe and Cu isotope compositions, some of which record near-complete isotopic equilibrium implies crystallisation at a range of cooling rates and depths, which is most consistent with cooling within the interior of a meteorite parent body. Our data therefore lend support to models where the non-magmatic irons are metallic melts that crystallised in the interior of re-aggregated, partially differentiated parent bodies.