Detailed analysis of the chemical composition of stars harboring Earth-like planets

Sammanfattning: Context. The planet-metallicity correlation for gaseous giants is widely accepted through spectroscopic studies. However, whether a similar correlation exists for terrestrial planets is a debated subject. High-precision spectroscopic abundance analysis on Sun-like stars suggested that the Sun is depleted in refractory elements with respect to its solar twins without exoplanets, likely due to the formation of terrestrial planets in the Solar system. Data. We use high-resolution (R=67000), high signal-to-noise ratio (S/N 200-300) optical spectra of 5 stars hosting terrestrial planets, 12 comparison stars and the Sun. These have been obtained with the High Resolution Echelle Spectrometer HIRES, on the Keck I telescope in Mauna Kea observatory. Aims. The goal of this work is to obtain the chemical elemental abundances of Sun-like stars with and without exoplanets, with a high precision of 0.01-0.02 dex. This precision will allow us to detect the planet formation signature in planet hosts, and to estimate the bulk chemical composition of their exoplanets. Methods. We determine the stellar parameters and elemental abundances by performing a strict line-by-line differential analysis. Equivalent widths (EW) are measured individually for all spectral lines with IRAF. Elemental abundances are computed with MOOG, using unidimensional, local thermodynamic equilibrium (1D LTE) model atmospheres. In addition, a chemical fractionation model that considers Earth’s devolatilisation with respect to the Sun is applied to estimate the bulk composition of exoplanets, giving our differential elemental abundances as input. Results. Stellar differential parameters and differential elemental abundances are calculated. The latter are obtained with a precision of 0.01-0.02 dex for 19 elements. The differential stellar abundances as a function of condensation temperature show a linear trend, whose slope (Tc slope) shows a dependence on stellar age due to Galactic chemical evolution (GCE) effects. This contribution to the Tc slope is corrected. Corrected differential abundances prove that the Sun is depleted in refractory elements with respect to the comparison stars. However, only one planet host shows the planet formation signature. On the other hand, most of the estimated planetary bulk compositions in our sample seem to have a similar core mass fraction compared to Earth, although their mantles are more enriched in SiO2. Conclusion. Our corrected differential abundances confirm that the Sun is depleted in refractories compared to Sun-like stars without exoplanets. The planet formation signature is only present in one out of five planet hosts in our sample. Therefore our results do not favor the planet formation signature hypothesis, although they are limited by small statistics. It is necessary to spectroscopically analyse more planet hosts with high precision, and to constrain planet densities and compositions in order to understand why the Sun presents these peculiar abundances. In addition, the bulk chemical composition of an exoplanet contributes to estimate its volatile reservoir and formation history.