Towards understanding planets around evolved stars
Posted:
Ji-Lin Zhou
Department of Astronomy, Nanjing university, Nanjing 210093 china; zhoujl@nju.edu.cn
To reveal planets outside of our solar system, which are called exoplanets, is a persisting objective of humanity. More than 17 years have pasted since the first exoplanet was discovered around a billion-year-old neutron star PSR 1257C12 in 1992 (Wolszczan & Frail 1992), followed by the detection of the first exoplanet around a main sequence star 51 Pegasi in 1995 (Mayor & Queloz 1995). Now, the number of confirmed exoplanets in our solar neighborhood has increased to 3331. A new planet candidate with a minimum mass of 2.7 Jupiter masses is reported in this issue of Research in Astronomy and Astrophysics (Liu et al. 2009). It is a companion to the intermediate-mass giant HD 173416, which is an evolved star with a mass of around 2.0 
. The observations were conducted at Xinglong station of the National Astronomical Observatory of China and Okayama Astrophysical Observatory of Japan. This is the second sub stellar companion reported from a joint planet-search program between these two groups; the first one is possibly a brown dwarf with a minimum mass of 19.4 Jupiter masses around 11 Comae, a giant star with a mass of 2.7 
(Liu et al. 2008).
Both of the sub-stellar companions were found by a method called the Doppler radial velocity technique. Due their small albedo and the small angular distances to their host stars, planets are very difficult to detect with direct imaging. For example, in the visible wavelengths, the reflected light from a Jupiter analog has brightness that is 10-9 that of its host star. Thus, some indirect methods are used to detect exoplanet systems. Among them, radial velocity measurements of star motion due to planets have proved to be the most effective one to date, and have contributed most of the discovered planets. However, due to the intrinsic activities of stellar photospheres, the radial velocity technique has a limiting precision of around 3 meter per second, although the limit can be substantially lowered for relatively quiet host stars (Marcy & Butler 1998).
Most of the projects that use the Doppler technique search for planets around main sequence stars of spectral types F or later, since for earlier type stars, the velocity precision is reduced by line broadening due to the rapid stellar rotation. However, this difficulty can be overcome by looking at stars that have left the main sequence. Evolved stars with intermediate mass, like giant and sub-giant stars, have lower surface temperatures and rotational velocities than their main-sequence progenitors, which makes them ideal targets in Doppler detections. Recently, observations have revealed more than 20 planets and a few brown dwarfs around such evolved stars.
The detection of planets around evolved stars complements our knowledge of planet formation. According to the modern theory, a planet forms in the circumstellar disk during the formation of the star. Based on how gas giant planets were formed, there are mainly two different scenarios: gravitational instability (GI) and core accretion (CA). The GI model assumes the gas giants formed in a relatively thick and cold disk so that gravitational contraction dominates over the gas pressure and centrifugal force, making planetary-mass clumps easily form within several thousand years (Boss 1997). As the requirement of gravitational instability is very strict, the GI scenario is thought to be effective only in some extreme cases. On the other hand, the CA model is also regarded as the major scenario for making planets (Ida & Lin 2008). In this scenario, planetary embryos merge in protostellar disks through dust condensation, cohesive planetesimal coagulation and giant impacts. As embryos exceed the critical mass (around 10 Earth masses), efficient gas accretion begins and gas giants form prior to the dispersal of the protostellar disk, with a timescale of 3-5 million years.
One of the crucial issues that can tell whether a planet is formed through the GI or CA models relies on the metallicity in the stellar disk, or equivalently the stellar gas. High metallicity is helpful for generating planets in the CA model since the density of planetesimals as well as their collision rates are higher, but it does not affect the formation of planets in the GI model, which depends only on the disk mass and temperature. Statistics show that planet-harboring main-sequence stars are, on average, metal-rich compared with stars that do not harbor planets (Fischer & Valenti 2005), indicating that most of the observed planets were formed through the CA model, a conclusion consistent with numerical simulations (Matsuo et al. 2007). However, whether the same conclusion applies to evolved stars with intermediate masses is still unclear. Statistics are unreliable due to the limited number of samples. For example, it is still controversial whether or not planet-harboring evolved stars have high metallicity (Pasquini et al. 2007; Hekker & Meléndez 2007).
Recently, a correlation was reported between stellar mass and planet occurrence, with a detection rate of 9% within 2.5 AU among the high-mass sample, compared to 4.5% for Sun-like stars and less than 2% for M dwarfs (Johnson et al. 2007). Because a larger mass star has a relative higher disk mass and thus has much more solid material, which favors both the GI and CA models, it is intuitive to believe that stars with higher masses can more easily harbor planets. However, several factors may complicate this picture. Firstly, the disk age is relatively short for larger mass disks (Haisch et al. 2001), making them more suitable for forming planets under the GI model. Secondly, planet migration will be speeded up due to the high disk mass, which is a disadvantage for both models, but more of a disaster for the CA model, since type I migration will clear all giant embryos before they can form gas giants. A paucity of planets within 1 AU of stars with masses greater than 1.5 solar masses was found (Johnson et al. 2008), indicating that migration could indeed be important in sculpting the architecture of planetary systems. Thirdly, for planets around evolved stars, were they formed at the stage of the protostar, and if so, how could they survive under the evolution of the host star transforming from a main sequence to a giant? The underlying mechanism is still not quite clear.
In the next few years, more and more space missions will join the team of planet-hunters, like Kepler (NASA's first mission capable of detection Earth-sized and smaller exoplanets in the habitable zone, scheduled for launch in Feb. 2009), ESO's GAIA (scheduled for launch in Dec. 2011), JPL's SIM PlanetQuest (scheduled after 2015) and NASA's TPF (Terrestrial Planet Finder, under study) .
Together with the ongoing missions like CNES' CoRoT (COnvection, ROtation & planetary Transits, Dec. 2006) and lots of ground-based projects, we are facing an explosion of exoplanet observations, and our knowledge of exoplanets, their formation, physics and dynamical evolution, will be more complete.
Johnson, J. A., Butler, R. P., Marcy, G. W., et al. 2007, ApJ, 670, 833
Johnson, J. A., Marcy, G. W., Fischer, D. A., et al. 2008, ApJ, 675,
Marcy, G. W., & Butler, R. P. 1998, Annu.Rev.Astron.Astrophys. 36, 57
Matsuo, T., Shibai, H., Ootsubo, T., Tamura, M. 2007, ApJ, 662, 1282
Pasquini, L., Döllinger, M. P., Weiss, A., et al. 2007, 473, 979
附件下载：
It accepts original submissions from all over the world and is internationally published and distributed by IOP