arXiv:astro-ph/0504214v2 2 May 2005Extrasolar Carbon PlanetsMarc J. Kuchner1Princeton University Department of Astrophysical SciencesPeyton Hall, Princeton, NJ 0854 4S. SeagerCarneg i e Institution of Washington, 5241 Broad Branch Rd. NW, Washington DC [email protected] suggest that some extrasolar planets ! 60 M!will form substantiallyfrom silicon carbide and other carbon compounds. Pulsar planets and low-masswhite dwarf planets are especially good candidate members of this new class ofplanets, but these objects could also conceivably form around stars like the Sun.This planet-formation pathway requires only a factor of two local enhancementof the pro t oplanetary disk’s C/O ratio above solar, a condition that pileups ofcarbonaceous grains may create in ordinary protoplanetary disks. Hot, Neptune-mass carbon planets should show a significant paucity of water vapor in theirspectra compared to hot pla nets with solar abundances. Cooler, less massivecarbon planets may show hydrocarbon-rich spectra and tar-covered surfaces. Thehigh sublimation t emperatures of diamond, SiC, and other carbon compoundscould protect these planets from carbon depletion at high temperatures.Subject headings: astrobiology — planets and satellites, individual (Mercury,Jupiter) — planetary systems: formation — pulsars, individual (PSR 1257+12)— white dwa r fs1. INTRODUCTIONThe recent discoveries of Neptune-mass extrasolar planets by the radial velocity method(Santos et al. 2004; McArthur et al. 2004; Butler et al. 2004) and the rapid development1Hubble Fellow– 2 –of new technologies to study the compositions of low-mass extrasolar pla nets (see, e.g., thereview by Kuchner & Spergel 2003) have compelled several authors to consider planets withchemistries unlike those found in the solar system (Stevenson 2 004) such as water planets(Kuchner 2003; Leger et al. 20 04). Here we describe a new possibility: extrasolar planets inwhich carbon is the most abundant component by number—carbon planets.Recently, Lodders (2004) argued t ha t t he Galileo-measured abundances of CH4andH2O in Jupiter’s atmosphere imply that O is depleted by a factor of 4 and C is enrichedby 1.7 relative to solar abundances, giving C/O = 1.8. She suggested that the planetaryembryo that grew into Jupiter may have formed where the nebula was locally carbon-rich,and tha t Jupiter’s embryo was a carbon planet. This suggestion inspired our research;if Jupiter could have formed from a carbo n-rich embryo, we would expect that carbon-rich embryos should be relatively common and occasionally observable as carbon planets.Although Lodders (2004) did not provide the only possible interpretation of the Galileo data,it seems reasonable that the solar nebula and other planetary systems could have formed largecarbon-rich bodies. This Letter discusses formation scenarios for carbon planets (Section 2),their likely compositions (Section 3), and t heir possible appeara nces (Section 4).2. FORMATION OF CARBON PLANET S2.1. CondensationIf solar composition gas at 1 0−4bars is cooled slowly from high t emperatures, severalmajor building blocks of the solar system condense out one by one. Fir st metal oxides andiron-peak elements condense at ∼1500 K, then silicates condense at 1200–1400 K, water at∼180 K, and eventually, ammonia and methane at lower temperatures (e.g., Lodders 2003).This equilibrium condensation sequence apparently describes the gross compositions of theinner solar system planets: Fe and Ni cores surrounded by silicate mantles, topped by morecomplicated veneers containing water and more volatile compounds. Lewis (1974) connectedthese two trends, suggesting that the locations of the planets determined their compositions;the solar system terrestrial planets formed in hot regions of the solar nebula from high-temperature condensates while the planets with larg er semi-major axes formed in coolerregions of the solar nebula fro m lower-temperature condensates. Some lower temperaturecondensates spread around the planetary system via small bodies late in the process of planetformation, coating planetary surfaces with volatiles. Although the details of this picture haveevolved—chemical processing is thought to be quenched in the outer solar system, planets arenow known to migrate, etc.—the equilibrium condensation sequence survives as a standardreference point for understanding the compositions of the solar system planets.– 3 –In g as with C/O ratio > 0 .98, the condensation sequence changes dramatically (Larimer1975). In carbon- r ich gas, the highest temperature condensates (T ≈ 1200–1600 K) a r ecarbon-rich compounds: graphite, carbides, nitrides, and sulfides. Since the sun has C/Oratio 0.5 (Asplund et al. 2005 ) , carbon-rich condensation sequences are not ordinarily asso-ciated with planet formation, though they have been investigated at length in the contextof the formation of silicon carbide (SiC) grains in meteorites (e.g., Lodders & Fegley 1995)and also in the context of outflows from evolved stars (e.g., Lodders & Fegley 1997, 1999).Low-mass planets formed via these carbon-rich condensation sequences would be carbonplanets, initially composed largely of the high-temperature condensates formed in carbon-rich gas, like graphite and silicon carbide.2.2. Formation ScenariosSome protoplanetary disks may spawn many carbon planets simply because they areesp ecially rich in carbon overall, and planet formation proceeds by a carbon-rich condensationsequence. The planets around the pulsar PSR 1257+12 (Wolszczan & Frail 1992) might havebeen formed in a carbon-rich nebula created by t he disruption of either a carbon-rich star orof a white dwarf (Tutukov 1991; Podsiadlowski et al. 1991; Phinney & Hansen 1993); perhapsthese pulsar planets are carbon planets. These mechanisms may also operate around whitedwarfs (Livio et al. 1992). In general, C/O ratios in stars and H II regions increase withmetallicity and towards the ga lactic center (e.g., Esteban et al. 2005), as reflected in galacticchemical evolution models (e.g., G avil´an et al. 2005); planets detected by microlensing maystand a better-than-average chance of being carbon planets because they are closer to thegalactic center than the Sun. Stars that host extrasolar planets are on average enhanced inmetals, including carbon, and often show 10–15 % enhancements in C/O ratio compared tothe sun (Gonzalez et al. 2001).The formation of carb on