Academic Commons

Theses Doctoral

High Pressure Melting of Iron with Nonmetals Sulfur, Carbon, Oxygen, and Hydrogen: Implications for Planetary Cores

Buono, Antonio Salvatore

The earth's core consists of a solid metallic center surrounded by a liquid metallic outer layer. Understanding the compositions of the inner and outer cores allows us to better understand the dynamics of the earth's core, as well as the dynamics of the cores of other terrestrial planets and moons. The density and size of the earth's core indicate that it is approximately 90% metallic, predominantly iron, with about 10% light elements. Iron meteorites, believed to be the remnants of planetary cores, provide further constraints on the composition of the earth's core, indicating a composition of 86% iron, 4% nickel, and 10% light elements. Any potential candidate for the major light element core component must meet two criteria: first, it must have high cosmic abundances and second, it must be compatible with Fe. Given these two constraints there are five plausible elements that could be the major light element in the core: H, O, C, S, and Si. Of these five possible candidates this thesis focuses on S and C as well exploring the effect of minor amounts of O and H on the eutectic temperature in a Fe-FeS core. We look at two specific aspects of the Fe-FeS system: first, the shape of the liquidus as a function of pressure, second, a possible cause for the reported variations in the eutectic temperature, which draws on the effect of H and O. Finally we look at the effect of S and C on partitioning behavior of Ni, Pt, Re,Co, Os and W between cohenite and metallic liquid. We are interested in constraining the shape of the Fe-FeS liquidus because as a planet with a S-enriched core cools, the thermal and compositional evolution of its core is constrained by this liquidus. In Chapter 1 I present an equation that allows for calculation of the temperature along the liquidus as a function of pressure and composition for Fe-rich compositions and pressures from 1 bar to 10 GPa. One particularly interesting feature of the Fe -rich side of the Fe-FeS eutectic is the sigmoidal shape of the liquidus. This morphology indicates non-ideal liquid solution behavior and suggests the presence of a metastable solvus beneath the liquidus. An important consequence of such curved liquidi is that isobaric, uniform cooling requires substantial variations in the solidification rate of the core. Additionally, in bodies large enough for P variation within the core to be significant, solidification behavior is further complicated by the P dependence of the liquidus shape. Brett and Bell (1969) show that at 3 GPa, the liquidus curvature relaxes, implying that the liquid solution becomes more ideal. By 10 GPa, the liquidus approaches nearly ideal behavior (Chen et al., 2008b). However, at 14 GPa, the liquidus again assumes a sigmoidal curvature (Chen et al., 2008a; Chen et al., 2008b), suggesting a fundamental change in the thermodynamic behavior of the liquid. Chapter 1 of this thesis accounts for the observed complexity in the liquidus up to 10 GPa thus enabling more accurate modeling of the evolution of the cores of small planets (Buono and Walker, 2011). Accurately knowing the eutectic temperature for the Fe-FeS system is important because it places a minimum bound on the temperature of a S-enriched core that has a solid and liquid component which are in equilibrium. Unfortunately literature values for the 1 bar to 10 GPa eutectic temperature in the Fe-FeS system are highly variable making the estimation of core temperature, an important geodynamic parameter, very difficult. In Chapter 2 we look at a possible cause of this observed variation by experimentally investigating the effects of H on the eutectic temperature in the Fe-FeS system at 6 and 8 GPa. We find that H causes a decrease in the eutectic temperature (but that O does not) and that this decrease can explain some of the observed scatter in the available data. The effect of H on the eutectic temperature increases with increasing pressure (i.e. the eutectic temperature is more depressed at higher pressures), matching the trend reported for the Fe-FeS system (Fei et al., 1997). Our work suggests a significantly higher eutectic temperature than is commonly used in the Fe-S system and explains the lower observed eutectic temperatures by employing the ternary Fe-S-H system. Additionally, we report an equation which allows for accurate prediction of the composition of the eutectic in the Fe-FeS system. The constraints presented here (eutectic temperature in the Fe-FeS system are 990 °C up to at least 8 GPa in conjunction with the equation presented in Chapter 1, allows for complete prediction of the Fe-rich liquidus in the Fe-FeS system to 8 GPa. It is important to understand the partitioning behavior of trace elements between the solid and liquid components of a system because it fundamentally informs our understanding of that systems chemical evolution. In light of this, we investigate partitioning behavior in the context of the Fe-S-Ni-C system in Chapter 3. Choice of this system was motivated by work outside the scope of this thesis investigating the liquidus relationships in the Fe-S-C system (Dasgupta et al., 2009). In these experiments, cohenite (Fe<sub>3</sub>C) is the stable solid phase, instead of Fe-metal and we find that the partition coefficients between cohenite and Fe-C-S liquids are significantly lower than those between Fe-metal and Fe-S liquids. There are two potential situations to which this work can be applied. With respect to core formation, although it is unlikely that any planets entire inner core is carbide, it is possible that in a C-rich planet, as the Fe core crystallizes, C in the liquid phase could be enriched to the point where cohenite is a stable crystalizing phase. Under these circumstances, we would predict smaller depletions of the elements studied in the outer core than would be the case for Fe-metal crystallization. This work can also be applied to the earth's upper mantle which is thought to become Fe-Ni metal-saturated as shallow as 250 km. Under these circumstances, the sub-system Fe-Ni-C (diamond) -S (sulfide) becomes relevant and Fe-Ni carbide rather than metallic Fe-Ni alloy could become the crystalline phase of interest. Our study implies that if cohenite and Fe-C-S melt are present in the mantle, the mantle budget of Ni, Co, and Pt may be dominated by Fe-C-S liquid. Additionally, in the case of a S-free system, W, Re, and Os will also be slightly enriched in Fe-Ni-C liquid over cohenite. In total this body of work better constrains several key aspects of the compositional and thermal evolution of cores in small planetary bodies and has potential implications for the earth's mantle.


  • thumnail for Buono_columbia_0054D_10433.pdf Buono_columbia_0054D_10433.pdf application/pdf 3.64 MB Download File

More About This Work

Academic Units
Earth and Environmental Sciences
Thesis Advisors
Walker, David
Ph.D., Columbia University
Published Here
November 11, 2011
Academic Commons provides global access to research and scholarship produced at Columbia University, Barnard College, Teachers College, Union Theological Seminary and Jewish Theological Seminary. Academic Commons is managed by the Columbia University Libraries.