Iron Abundance and the Formation of Terrestrial Exoplanets

Title: A compositional link between rocky exoplanets and their host stars

Authors: Vardan Adibekyan, Caroline Dorn, Sérgio G. Sousa, Nuno C. Santos, Bertram Bitsch, Garik Israelian, Christoph Mordasini, Susana C. C. Barros, Elisa Delgado Mena, Olivier D. S. Demangeon, João P. Faria, Pedro Figuiera, Artur A. Hakobyan, Mahmoudreza Oshagh, Bárbara M. T. B. Soares, Masanobu Kunitomo, Yoichi Takeda, Emiliano Jofré, Romina Petrucci, and Eder Martioli

First Author’s Institution: Instituto de Astrofísica e Ciências do Espaço & Departamento de Física e Astronomia, Universidade do Porto, Porto, Portugal

Status: Published in Science [closed access]


A diagram showing the process of star and planet formation. There are 6 steps in the image. The first shows the gas cloud collapsing, and the five following show the protostar forming, the protoplanetary disk forming, material clumping together to form planetesimals, protoplanets forming from those planetesimals, the rest of the gas disappating, and the resultant system (in this case, our Solar System).

Though we don’t yet have a complete picture of planetary formation – especially for types of planets we don’t find in the Solar System – we do have a good idea of the overall process. It starts with a cloud of gas and dust somewhere in space. Turbulence will then cause clumps of mass to form within the cloud. When there’s enough mass, the clump will collapse in on itself due to gravity, heating up and forming a protostar. Over time, a disk forms around the protostar from the remaining gas and dust – the protoplanetary disk. Within that disk, more collisions occur, and dust grains clump together, eventually becoming protoplanets. Mass will continue to accrete onto the protostars and protoplanets, essentially meaning that the gas and dust “fall” onto the mass clumps, making them bigger. Eventually, the protostar grows into a main sequence star, and the protoplanets are either destroyed or become full-blown planets.

Figure 1: The steps of star and planet formation are shown here, as follows: a) the gas cloud collapses, b) the protostar begins to form, c) the protoplanetary disk forms around the protostar, d) material starts to clump together, forming protoplanets, e) the protoplanets have grown larger and the remaining gas and dust has dissipated, and f) the resulting system (in this case, the Solar System)

The chemical composition of the protoplanetary disk and the resulting planets can tell us important information about the formation mechanisms of the planets, as well as the potential characteristics of the planets. With current observational technology, we can’t always determine the composition of the planet itself. But, since this whole system originates from the same cloud of gas and dust, learning information about the chemical composition of the stars can also tell us information about the planets around them. We can do that relatively easily using stellar spectroscopy, since different elements present in a star or planet will emit and absorb different wavelengths of light in a spectrum.


Today’s authors investigated 32 known exoplanets with masses M < 10 MEarth orbiting 27 different sun-like stars to find a correlation between terrestrial planet formation and the composition of planetary host stars. All 32 of the planets have known masses from radial velocity measurements and known radii from transit measurements. The authors plotted them on a mass-radius diagram and removed 10 planets that are likely mini-Neptunes from the sample (see Figure 2). They then determined the abundances of Mg, Si, and Fe – all major rock-forming elements – in the host stars using their spectra. Using pre-existing stellar composition models and these abundances, the authors estimate the iron abundance of the stars and protoplanetary disks relative to Mg and Si. Using the masses and radii of the planets and existing planetary interior models, they then estimate the iron abundance – again relative to Mg and Si – of the planets, independent of the star. 

A plot showing the radii of the sample planets versus their masses.
Figure 2: The mass-radius diagram used to determine which exoplanets the authors kept. The mass axis is in terms of Earth masses, and the radius axis is in terms of Earth radii. The dark blue line indicates the mass-radius parameters at which they expect Earth-like exoplanets. The dotted blue lines indicate the radius gap the authors used to distinguish between the terrestrial exoplanets and the mini-Neptune exoplanets. The grey line indicates the minimum planetary radius predicted by the collision model of planet formation (an explanation of this model can be found in this bite). Figure 1 in the paper.


Figure 3 shows the results of plotting each planet’s resulting iron abundance versus its host star’s iron abundance. Indeed, we see a positive correlation between the iron abundance of these terrestrial planets and their host stars. Interestingly, the results suggest that five of the planets in the sample are likely “super-Mercuries”, planets which have similar compositions to Earth, but much higher masses relative to their radii, similar to Mercury in our solar system. All five super-Mercuries orbit stars with high iron-to-silicate ratios and high iron abundances, which suggests that the planets’ compositions may be related to their stellar and protoplanetary disk compositions. Under these conditions, more collisions would likely occur during the planet formation process, lending credence to the theory that Mercury and Mercury-like exoplanets were formed through collision processes. Though more data is needed on super-Mercury populations, this study could potentially be adding another piece to the puzzle of planet formation!

Two plots of the iron-to-silicate ratio for the planets versus the iron-to-silicate ratio of their host stars.
Figure 3: For each planet, the iron abundance of its host star relative is plotted versus its iron abundance. The color of the points indicates the mass of the planets in Earth masses, with the lighter colors being the less massive and the darker colors being more massive. The solid black line indicates the correlation between the planet iron abundance and the stellar iron abundance with the super-Mercuries included in the sample, while the dashed line shows the correlation if they are excluded. The Solar System planets are included in the plot for reference, but are not included in the calculations. Panel A shows what the distribution would be if all the iron was in the core of the exoplanets, while Panel B shows what the distribution would be if the iron was distributed between the core and the mantle of the exoplanets. Figure 3 in the paper.

Astrobite edited by Katya Gozman

Featured image credit: ESO/L. Calçada

About Ali Crisp

I'm a fourth year grad student at Louisiana State University. I study hot Jupiter exoplanets in the Galactic Bulge. I am originally from Tennessee and attended undergrad at Christian Brothers University, where I studied physics and history. In my "free time," I enjoy cooking, hiking, and photography.

1 Comment

  1. I had never thought of similarities in composition between planets & their host stars before. Thank you.


Leave a Reply