The carbon-to-oxygen ratio in stars with planets

TL;DR

This study refines the measurement of the carbon-to-oxygen ratio in stars with planets, showing it varies slightly with metallicity and does not support the existence of carbon-rich planets, using improved spectroscopic methods.

Contribution

It provides more accurate C/O ratios in stars by using high-excitation lines and non-LTE corrections, challenging previous claims of high C/O values.

Findings

C/O ratio increases slightly with metallicity from 0.58 to 0.70

No evidence supporting the existence of carbon-rich planets

C/O ratios are consistent with well-mixed interstellar medium

Abstract

In some recent works, the C/O abundance ratio in high-metallicity stars with planets is found to vary from 0.4 to about 1.0. This has led to discussions about the existence of terrestrial planets with a carbon-dominated composition that is very different from the composition of the Earth. The C/O values were obtained by determining carbon abundances from high-excitation CI lines and oxygen abundances from the forbidden [OI] line at 6300 A. This weak line is, however, strongly affected by a nickel blend at high metallicities. Aiming for more precise C/O ratios, oxygen abundances in this paper are derived from the high-excitation OI triplet at 7774 A and carbon abundances from the CI lines at 5052 and 5380 A using MARCS model atmospheres and including non-LTE corrections. The results do not confirm the high C/O ratios previously found. C/O shows a tight, slightly increasing dependence on metallicity from C/O=0.58 at [Fe/H]=0.0 to C/O=0.70 at [Fe/H] =0.4 with an rms scatter of only 0.06. Assuming that the composition of a proto-planetary disk is the same as that of the host star, the C/O values found in this paper lend no support to the existence of carbon-rich planets. The small scatter of C/O among thin-disk stars suggests that the nucleosynthesis products of Type II supernovae and low- to intermediate-mass stars are well mixed in the interstellar medium.