Free Access
Issue
A&A
Volume 542, June 2012
Article Number A92
Number of page(s) 10
Section Planets and planetary systems
DOI https://doi.org/10.1051/0004-6361/201118051
Published online 13 June 2012

© ESO, 2012

1. Introduction

More than 700 extrasolar planet (exoplanet) candidates were discovered so far (Schneider et al. 2011, http://www.exoplanet.eu), but the knowledge of their properties is strongly affected by observational bias and selection effects. Taking the solar system as an archetype, the target lists of exoplanet search programs so far originally consist of mostly single and solar like stars (regarding the spectral type and age). But the first planet candidate detected by the RV technique was found around the primary of the close spectroscopic binary γ   Cep (Campbell et al. 1988; Hatzes et al. 2003; Neuhäuser et al. 2007), which demonstrates the existence of planets in binaries.

In the last years, imaging campaigns found stellar companions around several dozen exoplanet host stars formerly believed to be single stars (see e.g. Raghavan et al. 2006; Mugrauer & Neuhäuser 2009, and references therein). Most of these exoplanet candidates are in the S-type orbit configuration (exoplanet surrounding one stellar component of a binary), while the orbit of a planet around both stellar binary components is called P-type orbit. Such circumbinary planets are detectable by measuring eclipse timing variations as done for NN Ser (Beuermann et al. 2010), HW Vir (Lee et al. 2009), DP Leo (Qian et al. 2010), HU Aqr (Qian et al. 2011; Hinse et al. 2012), and UZ For (Dai et al. 2010; Potter et al. 2011). Kepler-16 (AB)b, Kepler-34 (AB)b, and Kepler-35 (AB)b are detected by measuring the transit lightcurve and eclipse timing variations (Doyle et al. 2011; Welsh et al. 2012), thus these are confirmed circumbinary planets. Due to a different formation and evolution scenario for planets in a P-type orbit (compared to the more common S-type orbit), this paper only considers exoplanets found in a S-type orbit.

Multiplicity studies, as done by Mugrauer et al. (2007b) or Eggenberger et al. (2007), are looking for stellar companions around exoplanet host stars by direct imaging. As summarized in Mugrauer & Neuhäuser (2009), these studies found 44 stellar companions around stars previously not known to be multiple, which results in a multiplicity rate of about 17%, while Raghavan et al. (2006) found a host star multiplicity of about 23%. The multiplicity rate of solar like stars1 was determined by Raghavan et al. (2010) to (46 ± 2)%. Duquennoy & Mayor (1991) measured the multiplicity of 164 nearby G-dwarfs (within 22 pc) to 44% (57% considering incompleteness).

2. Extrasolar planets in stellar multiple systems

The deuterium burning minimum mass (DBMM) of 13   MJup is currently the most common criterion to distinguish a brown dwarf from a planet. However, we make use of the Extrasolar Planets Encyclopaedia (EPE) in this paper and thus apply the definition of Schneider et al. (2011) who includes all confirmed substellar companions with a mass of less than 25   MJup within a 1σ uncertainty. Due to a missing publication of the planet detection, the exoplanet candidates GJ 433 b, ρ CrB b, 91 Aqr b, ν Oph b&c, τ Gem b, HD 59686 b, HD 106515A b, HD 20781 b&c, and HD 196067 b are not included in this paper. Also, the stellar binary HD 176051, where Muterspaugh et al. (2010) detected the astrometric signal of an exoplanet around one of the two stellar components, is not included in this study: Because the planet was found by ground based astrometric observation (using an optical interferometer), this detection still need to be confirmed by other techniques and the final planetary mass depends on which of the stellar components is the host star.

The multiplicity of an exoplanet host star is defined (in this paper) by either a published common proper motion or an entry in the Catalogue of Components of Double and Multiple Stars (CCDM) by Dommanget & Nys (2000). In case, the stellar multiplicity is only mentioned in the CCDM, all stellar components were checked on common proper motion using other catalogs (see Appendix A). By searching the literature and matching the host stars of exoplanet candidates detected with transit or RV observations listed in the EPE (date: 2012/02/08) with the CCDM, 57 stellar multiple systems (47 double and 10 triple systems) with at least one exoplanet out of 477 systems in total are identified. The resulting multiplicity rate of about 12% is less than previously published values (see Table 1). An explanation for that can be the increasing number of transiting exoplanets in the last years, which are included in this paper but excluded by previous studies. The host star multiplicity of transiting exoplanets is most likely still underestimated, because multiplicity studies around such host stars, like done by Daemgen et al. (2009), have just recently started.

Table 1

Multiplicity of solar like and exoplanet host stars.

The complete list of the 57 multiple systems harboring exoplanet candidates can be found in Tables A.4A.6. Furthermore, the proper motions of all these stars gathered from online catalogs are shown in Tables A.1A.3. The latest published summary, done by Mugrauer & Neuhäuser (2009), listed 44 planetary systems in a stellar multiple system. However, two of these systems are excluded in this study, namely HD 156846 AB (after Reffert & Quirrenbach 2011, published astrometric mass limits of mpl = (10.5...660.9)   MJup, the EPE planetary status changed to unconfirmed), and 91 Aqr (the planet detection itself is still not published in a refereed paper). In addition to that 42 systems, 15 new systems are listed and marked by the symbol ? in Tables A.4A.6.

Table 2

Critical semi-major axis acrit for planets in close stellar binaries, calculated according to Holman & Wiegert (1999).

3. Comparison of extrasolar planets in stellar multiple systems and around single stars

Marcy et al. (2005a) fitted the histogram of all known RV exoplanet minimum masses by a simple power-law and found an exponent of  − 1.05 for the mass distribution. That exponent is in good agreement with a sample of synthetic exoplanets detectable by current RV observations modeled by Mordasini et al. (2009). In our work, planets currently found by RV or transit observations are analyzed (see Fig. 1). Using also a simple power-law (see Fig. 2), an exponent of  − 1.03 was found for the mass distribution of all exoplanet candidates, which is similar to the results of previous works. For exoplanet candidates in stellar multiple systems and around single stars, the exponent is  − 0.97 and  − 1.04, respectively. The mean of the planetary masses is about 2.5   MJup for planets around single stars and 3.1   MJup for the case of stellar multiplicity. In addition to the power-law, we also fit a log-normal distribution to the planetary masses. The probability distribution function (PDF) and the expectation value  of a log-normal distribution for a measure x can be calculated by (1)where μ and σ are the mean value and the standard deviation of the distribution. To determine the χ2 value (shown in Fig. 2) the Python package “SciPy” (Jones et al. 2001) was used.

thumbnail Fig. 1

The (minimum) mass of extrasolar planets detected by RV (circles) or transit (squares) observations over their orbital period. Exoplanets around single stars are shown as open markers, while exoplanets in stellar multiple systems are coded by filled symbols. Jupiter is shown as a filled triangle and the filled diamond marks Neptune.

Open with DEXTER

thumbnail Fig. 2

Mass distribution of exoplanets detected by RV or transit observations around single stars (left column), in stellar multiple systems (middle column), and for all kind of host stars (right column) fitted by a power law (dashed line, upper values) and a log-normal distribution (solid line, lower values).

Open with DEXTER

As one can see in Fig. 2, the log-normal fit results in a better than the power-law fit. The expectation values for the mass of exoplanets around single stars and in stellar multiple systems differ, hence the power-law as well as the log-normal fit lead to the conclusion that the mass distribution of exoplanets in stellar multiple systems are pushed towards higher planetary masses, compared to the mass distribution of exoplanets around single stars.

However, the statistic of the exoplanet host star multiplicity is still affected by observational bias and selection effects of the originally planet search programs. Most multiplicity studies so far were carried out after the planet detection. Hence, most of the host stars are solar like (regarding the age and spectral type), but they are also originally selected as single stars. To avoid the adaption of such selection effects, systematic searches for planets in stellar multiple systems, like described in Desidera et al. (2007) or Roell et al. (2010), are needed.

thumbnail Fig. 3

Planetary (minimum) mass over the projected separation of the exoplanet host star and its nearest stellar companion. The markers represent the number of planets per system (dots ... one planet, squares ... two planets, triangles ... three or more planets). The size of the symbols represent the mass of the exoplanet host star.

Open with DEXTER

4. Influence of a close stellar companion on planet properties

In the previous section, the difference in the mass of exoplanets around single stars and in stellar multiple systems was discussed. In order to unveil the cause of this difference, a closer look on the influence of a stellar companion around the exoplanet host star is advisable. In Fig. 3 we plot the planetary (minimum) mass over the projected separation of the exoplanet host star and its nearest stellar companion. Because all systems analyzed in this paper are hierarchical, the exoplanet host star and its nearest stellar companion can be treated as a binary system. The order of the stellar multiplicity is not relevant, but the planetary minimum mass decreases with an increasing projected stellar separation (dashed line in Fig. 3). Furthermore, multi-planet systems are only present in stellar systems with a projected stellar separation larger than about 100 AU and up to now, no planet was found in a stellar binary with a projected separation of less than 10 AU. The two planets below the dashed line in Fig. 3 are the planetary system around GJ 667 C, a component of a hierarchical triple star system at a distance of 7 pc. The true semi-major axis is likely larger than the measured projected separation and the true planetary mass could also be larger than the measured minimum mass. These observational bias effects could explain, why GJ 667 C is the only system left of that dashed line in Fig. 3.

Holman & Wiegert (1999) determine a formula to calculate the critical semi-major axis acrit for a stable planetary orbit coplanar to the stellar orbit with the semi-major axis abin, which varies from acrit ≃ (0.02...0.45)   abin, depending on the mass ratio μbin and the eccentricity ebin of the stellar binary. Table 2 listed the five systems, where the apparent separation is less than 50 times the planetary semi-major axis (see Fig. 4) including the corresponding critical semi-major axis. Except for the exoplanet HD 196885 Ab, which grazes an “unstable region” during the apastron passage, all these systems are clearly stable. However, considering the age of the F8V star HD 196885 A of 2.0    ±    0.5 Gyr (Correia et al. 2008), the planetary system can also be regarded as long-term stable.

thumbnail Fig. 4

Planetary semi-major axis over the projected stellar separation (dots ... stellar binary, triangles ... triple star). The five systems in the upper left () are shown in more detail in Table 2. The size of the symbols represent the (minimum) mass of the exoplanet.

Open with DEXTER

5. Summary

Analyzing the host star multiplicty of exoplanets detected by RV or transit observations, 57 exoplanet host stars with stellar companions are identified and presented in the Appendix A, including 15 new systems (compared to the latest published summary in 2009). The resulting multiplicity rate for exoplanet host stars of at least 12% is about four times smaller than the multiplicity of solar like stars. No planet is found so far in stellar binaries with a projected separation of less than 10 AU and multi-planet systems were only found in stellar systems with a projected separation larger than 100 AU. The planetary (minimum) mass decreases with an increasing projected stelllar separation.


1

Defined as all main-sequence stars with a spectral type from F6 to K3 within 25 parsec, see Raghavan et al. (2010).

Acknowledgments

T.R. and R.N. thank the DFG (Deutsche Forschungsgemeinschaft) for financial support under the project numbers NE 515/23-1, NE 515/30-1, and NE 515/33-1 (SPP 1385: “First ten million years of the solar systems”). A.S. acknowledge support from the National Science Foundation under grant NSF AST-0708074. This research has made use of the SIMBAD database and the VizieR catalog access tool, both operated at CDS, Strasbourg, France.

References

Appendix A: Updated tables of extrasolar planets in stellar multiple systems

Table A.1

Exoplanet host stars having a common proper motion with another star. The ID of the host stars is the same as it is used in the EPE and the planet status is coded by R... published in a Refereed paper, S... Submitted to a professional journal, and C... announced by astronomers in professional Conferences. A planet status followed by the letter R in brackets means, the status on the EPE is not up-to-date and the planet detection is already published in a refereed paper. The fourth and fifth column show the proper motion listed in the catalog mentioned in the last column. For easy identification of the stellar companions in the online catalogs, the third column contains either the separation from the primary as calculated by the used online catalog or the latest separation measurement in case of published relative astrometric measurements.

Table A.2

Exoplanet host stars listed in the CCDM, but unlikely a common proper motion pair (same columns as in A.1).

Table A.3

Exoplanet host stars with companion candidates, but further epoch observations are needed (same columns as in A.1).

Table A.4

Extrasolar planets detected with transit or RV observations in closer binaries with a projected stellar separation of , sorted by an increasing stellar separation. For the four closest systems a value for the binary semi-major axis (abin) is known from multi-epoch observations (listed in brackets in the column). If RV and transit measurements are available the true mass of the exoplanet candidate is given in the table. ?  ... New system compared to the latest published overview by Mugrauer & Neuhäuser (2009). ?  ... B component is a brown dwarf, see Mugrauer et al. (2006b). ?  ... B component is a white dwarf, see Mugrauer & Neuhäuser (2005). §  ... B component is a white dwarf, see Chauvin et al. (2007). d  ... Reffert & Quirrenbach (2011) determined an astrometric mass range for the planet candidate, which is given within the brackets in the planetary mass column.

Table A.5

Extrasolar planets detected with transit or RV observations in wider binaries with a projected stellar separation of ), sorted by an increasing stellar separation. If RV and transit measurements are available the true mass of the exoplanet candidate is given in the table. ?  ... New system compared to the latest published overview by Mugrauer & Neuhäuser (2009). ?  ... Closer component listed in the CCDM (formerly called B) was disproved by proper motion measurements, but a new wide stellar companion was found and confirmed by common proper motion (Mugrauer et al. 2005). §  ... B component is a white dwarf, see Porto de Mello & da Silva (1997).

Table A.6

Extrasolar planets detected with transit or RV observations in stellar systems with more than two components, sorted by the increasing projected separation of the host star and the nearest stellar component (). For the closest systems a value for the binary semi-major axis (abin) is known from multi-epoch observations (listed in brackets in the column). If RV and transit measurements are available the true mass of the exoplanet candidate is given in the table. ?  ... New system compared to the latest published overview by Mugrauer & Neuhäuser (2009).

All Tables

Table 1

Multiplicity of solar like and exoplanet host stars.

Table 2

Critical semi-major axis acrit for planets in close stellar binaries, calculated according to Holman & Wiegert (1999).

Table A.1

Exoplanet host stars having a common proper motion with another star. The ID of the host stars is the same as it is used in the EPE and the planet status is coded by R... published in a Refereed paper, S... Submitted to a professional journal, and C... announced by astronomers in professional Conferences. A planet status followed by the letter R in brackets means, the status on the EPE is not up-to-date and the planet detection is already published in a refereed paper. The fourth and fifth column show the proper motion listed in the catalog mentioned in the last column. For easy identification of the stellar companions in the online catalogs, the third column contains either the separation from the primary as calculated by the used online catalog or the latest separation measurement in case of published relative astrometric measurements.

Table A.2

Exoplanet host stars listed in the CCDM, but unlikely a common proper motion pair (same columns as in A.1).

Table A.3

Exoplanet host stars with companion candidates, but further epoch observations are needed (same columns as in A.1).

Table A.4

Extrasolar planets detected with transit or RV observations in closer binaries with a projected stellar separation of , sorted by an increasing stellar separation. For the four closest systems a value for the binary semi-major axis (abin) is known from multi-epoch observations (listed in brackets in the column). If RV and transit measurements are available the true mass of the exoplanet candidate is given in the table. ?  ... New system compared to the latest published overview by Mugrauer & Neuhäuser (2009). ?  ... B component is a brown dwarf, see Mugrauer et al. (2006b). ?  ... B component is a white dwarf, see Mugrauer & Neuhäuser (2005). §  ... B component is a white dwarf, see Chauvin et al. (2007). d  ... Reffert & Quirrenbach (2011) determined an astrometric mass range for the planet candidate, which is given within the brackets in the planetary mass column.

Table A.5

Extrasolar planets detected with transit or RV observations in wider binaries with a projected stellar separation of ), sorted by an increasing stellar separation. If RV and transit measurements are available the true mass of the exoplanet candidate is given in the table. ?  ... New system compared to the latest published overview by Mugrauer & Neuhäuser (2009). ?  ... Closer component listed in the CCDM (formerly called B) was disproved by proper motion measurements, but a new wide stellar companion was found and confirmed by common proper motion (Mugrauer et al. 2005). §  ... B component is a white dwarf, see Porto de Mello & da Silva (1997).

Table A.6

Extrasolar planets detected with transit or RV observations in stellar systems with more than two components, sorted by the increasing projected separation of the host star and the nearest stellar component (). For the closest systems a value for the binary semi-major axis (abin) is known from multi-epoch observations (listed in brackets in the column). If RV and transit measurements are available the true mass of the exoplanet candidate is given in the table. ?  ... New system compared to the latest published overview by Mugrauer & Neuhäuser (2009).

All Figures

thumbnail Fig. 1

The (minimum) mass of extrasolar planets detected by RV (circles) or transit (squares) observations over their orbital period. Exoplanets around single stars are shown as open markers, while exoplanets in stellar multiple systems are coded by filled symbols. Jupiter is shown as a filled triangle and the filled diamond marks Neptune.

Open with DEXTER
In the text
thumbnail Fig. 2

Mass distribution of exoplanets detected by RV or transit observations around single stars (left column), in stellar multiple systems (middle column), and for all kind of host stars (right column) fitted by a power law (dashed line, upper values) and a log-normal distribution (solid line, lower values).

Open with DEXTER
In the text
thumbnail Fig. 3

Planetary (minimum) mass over the projected separation of the exoplanet host star and its nearest stellar companion. The markers represent the number of planets per system (dots ... one planet, squares ... two planets, triangles ... three or more planets). The size of the symbols represent the mass of the exoplanet host star.

Open with DEXTER
In the text
thumbnail Fig. 4

Planetary semi-major axis over the projected stellar separation (dots ... stellar binary, triangles ... triple star). The five systems in the upper left () are shown in more detail in Table 2. The size of the symbols represent the (minimum) mass of the exoplanet.

Open with DEXTER
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.