2012ApJ...747L..27F

Recent high-resolution spectroscopic analysis of nearby FGK stars suggests that a high C/O ratio of greater than 0.8, or even 1.0, is relatively common. Two published catalogs find C/O > 0.8 in 25%-30% of systems, and C/O > 1.0 in ∼6%-10%. It has been suggested that in protoplanetary disks with C/O > 0.8 that the condensation pathways to refractory solids will differ from what occurred in our solar system, where C/O = 0.55. The carbon-rich disks are calculated to make carbon-dominated rocky planets, rather than oxygen-dominated ones. Here we suggest that the derived stellar C/O ratios are overestimated. One constraint on the frequency of high C/O is the relative paucity of carbon dwarf stars (10^–3-10^–5) found in large samples of low-mass stars. We suggest reasons for this overestimation, including a high C/O ratio for the solar atmosphere model used for differential abundance analysis, the treatment of a Ni blend that affects the O abundance, and limitations of one-dimensional LTE stellar atmosphere models. Furthermore, from the estimated errors on the measured stellar C/O ratios, we find that the significance of the high C/O tail is weakened, with a true measured fraction of C/O > 0.8 in 10%-15% of stars, and C/O > 1.0 in 1%-5%, although these are still likely overestimates. We suggest that infrared T-dwarf spectra could show how common high C/O is in the stellar neighborhood, as the chemistry and spectra of such objects would differ compared to those with solar-like abundances. While possible at C/O > 0.8, we expect that carbon-dominated rocky planets are rarer than others have suggested.