Heavy iron isotope composition of iron meteorites explained by core crystallization

Abstract

Similar to Earth, many large planetesimals in the Solar System experienced planetary-scale processes such as accretion, melting, and differentiation. As their cores cooled and solidified, significant chemical fractionation occurred due to solid metal-liquid metal fractionation. Iron meteorites -- core remnants of these ancient planetesimals -- record a history of this process. Recent Fe isotope analyses of iron meteorites found δ^57/54Fe to be heavier than chondritic by approximately 0.1 to 0.2 ‰ for most meteorites, indicating that a common parent body process was responsible. However, the mechanism for this fractionation remains poorly understood. Here we experimentally show that the Fe isotopic composition of iron meteorites can be explained solely by core crystallization. In our experiments of core crystallization at 1300 °C, we find that solid metal becomes enriched in δ^57/54Fe by 0.13 ‰ relative to liquid metal. Fractional crystallization modelling of the IIIAB iron meteorite parent body shows that observed Ir, Au and Fe isotopic compositions can be simultaneously reproduced during core crystallization. The model implies the formation of complementary S-rich components of the iron meteorite parental cores that remain unsampled by meteorite records and may be the missing reservoir of isotopically-light Fe. The lack of sulfide meteorites and previous trace element modeling predicting significant unsampled volumes of iron meteorite parent cores support our findings.

Figure 1 |. Iron isotopic compositions for various types of terrestrial and extra-terrestrial samples.

The error-weighted means and two standard errors for different planetary sample groups are shown in colored bars. Typical analytical errors are shown as the grey bar for comparison. Data are from refs.^–,–. Figure modified from ref..

Figure 2 |. Results of solid metal - liquid metal equilibrium experiments.

The experimental fractionation factors are corrected to 1300 °C to evaluate the effect of S (see Supplementary Information). Error bars of individual experiments are 2 s.e. Two sets of time-series experiments with ~20 wt% S and ~25 wt% S gave similar fractionation factors within analytical error, demonstrating that isotopic equilibrium was reached in our experiments even at the lowest temperatures. Our data did not resolve measurable dependence of Δ⁵⁷Fe_{solid-liquid metal} on sulfur content. Instead, the trend can be described well by an error-weighted mean of 0.129 ± 0.067 ‰ (2 s.d.).

Figure 3 |

Core crystallization fractionation modelling. a, Illustration of our model. Inward crystallization is shown in the illustration for the IIIAB parent body (see Supplementary Information). b and c, Modeled Ir-Au and Ir-δ⁵⁷Fe diagrams for IIIAB iron meteorites assuming an initial metallic liquid with 11.5 wt% S, 7.2 wt% Ni, 0.755 ppm Au, 2.88 ppm Ir, and 0‰ δ⁵⁷Fe. The red, blue and orange curves are modeled compositions for crystallized solid metal, crystallizing liquid metal, and residual solid metal from the trapped melt after troilite formation. Error bars in c are 2 s.e. Details about the model can be found in Methods.

Figure 4 |. Demonstration of the missing S-rich reservoir unsampled by iron meteorites.

During crystallization of iron meteorite parent cores, sulfur preferentially partitions into the remaining liquid, causing the later crystallized part of the core to be S-rich in composition. According to our experiments, this part of the core is also enriched in the light isotopes of Fe, but essentially unsampled by meteorite records on Earth. Magmatic iron meteorites predominantly come from the early-crystallized, S-poor part of the core and ended up heavy in Fe isotope compositions on average.