Abstract:Introduction: Meteorites provide a “window” into planetary interiors; allowing us to better understand evolution of the early solar system and planetary differentiation. Iron meteorites in particular present a unique opportunity to study core formation processes. The variety in chemical compositions between iron meteorite groups points to complex accretion and differentiation histories of their parent bodies. Understanding coremantle segregation processes under different accretionary conditions can provide insight into the range of chemical compositions recorded in meteorites. Two mechanisms are often cited to describe metal segregation and core formation: (1) the existence of a magma ocean and (2) metallic melt percolation through a solid silicate matrix [1]. Planets may experience different segregation processes depending on heat source and thermal evolution, but differentiation via percolation is likely the dominant process for smaller planetary bodies when heated to temperatures high enough to melt the metallic component [2]. Metallic melt migration through a porous matrix is primarily controlled by melt volume and the chemical composition [3–5]; the latter most notably affects the wettability of the liquid. Iron meteorites, dominantly composed of Fe and Ni, also contain small amounts of light elements including sulfur, silicon, carbon, and phosphorous [6]. Light elements have proven to reduce melt/olivine interfacial energy, though show a range in this effect. For example the addition of sulfur or oxygen reduces the dihedral angle while silicon has no effect [1,7–11]. However, studies to date have largely determined percolation behavior for binary or ternary systems via dihedral angle measurements. Our objective is to better understand the physical process of percolation in a complex system of multiple light elements. Our experiments address the following questions: (1) What are the compositions of the immiscible liquids in the Fe-Ni-Si-S-C system? (2) What is the percolation behavior of the immiscible liquids (i.e., can the two phases physically separate during transport through cracks/grain boundaries)? (3) How does temperature affect percolation behavior, especially when partial melting of metal occurs? (4) Can we relate composition evolution during percolation to meteorite compositions? Methods: To simulate liquid metal percolation in a small planetary interior, we performed piston-cylinder experiments at temperatures between 1100–1600 °C at 1 GPa. Two compositions were employed to observe the effects of light elements on percolation behavior: Fe + 5 wt.% Ni + 6 wt.% Si + 2 wt.% S and Fe + 5 wt.% Ni + 2 wt.% Si + 6 wt.% S. Powders were loaded into graphite capsules to allow C saturation in the experiments (£ 6.4 wt.%). Within the graphite capsule, we used two capsule configurations: (1) San Carlos (SC) olivine inner “capsule”, where powdered SC olivine fully surrounded the metal sample or (2) a SC olivinemetal-SC olivine “sandwich”. The capsule along with MgO spacers was placed within a half inch talc-pyrex assembly with graphite heaters. Samples were pressurized to 1 GPa and heated to 800 °C and sintered for two hours. The experiments were then heated to the target temperatures and held for 3–12 hours before quenching. The experimental products were polished and their chemical compositions analyzed with the electron microprobe (EMP). Apparent dihedral angles were measured on 2D images captured during EMP analysis or with the scanning electron microscope (SEM). Results: Both starting compositions form immiscible Fe-Ni-S and Fe-Ni-C liquids (Fig. 1) that coexist with a solid Fe-Ni-Si-C alloy at lower experimental temperatures. The Fe-Ni-S melt wets the surrounding olivine grains and carries the non-wetting, C-rich Fe-Ni alloy into the olivine matrix. The C-rich melt is always found suspended within the Fe-Ni-S melt and shows no evidence of segregation from the wetting phase (Fig. 2).

PERCOLATION BEHAVIOR IN THE Fe-Ni-Si-S-C SYSTEM AND INFLUENCE ON METEORITE COMPOSITIONS