Draft version September 20, 2018 Preprint typeset using LATEX style emulateapj v. 11/10/09 RETIRED A STARS AND THEIR COMPANIONS VII. EIGHTEEN NEW JOVIAN PLANETS1 John Asher Johnson2,9, Christian Clanton2,9, Andrew W. Howard4, Brendan P. Bowler3, Gregory W. Henry5, Geoffrey W. Marcy4, Justin R. Crepp2, Michael Endl6, William D. Cochran6, Phillip J. MacQueen6, Jason T. Wright7,8, Howard Isaacson3 Draft version September 20, 2018 ABSTRACT We report the detection of eighteen Jovian planets discovered as part of our Doppler survey of subgiant stars at Keck Observatory, with follow-up Doppler and photometric observations made at McDonald and Fairborn Observatories, respectively. The host stars have masses 0.927 ≤ M⋆/M⊙ ≤ 1.95, radii 2.5 ≤ R⋆/R⊙ ≤ 8.7, and metallicities −0.46 ≤ [Fe/H] ≤ +0.30. The planets have minimum masses 0.9 MJup ≤ MP sin i . 13 MJup and semimajor axes a ≥ 0.76 AU. These detections represent a 50% increase in the number of planets known to orbit stars more massive than 1.5 M⊙ and provide valuable additional information about the properties of planets around stars more massive than the Sun. Subject headings: techniques: radial velocities—planetary systems: formation—stars: individ- ual (HD1502, HD5891, HD18742, HD28678, HD30856, HD33142, HD 82886, HD96063, HD98219, HD99706, HD102329, HD106270, HD108863, HD 116029, HD131496, HD142245, HD152581, HD158038) 1. INTRODUCTION This result has both informed models of giant Jupiter-mass planets are not uniformly distributed planet formation (Ida & Lin 2004; Laughlin et al. around all stars in the galaxy. Rather, the rate of 2004; Thommes et al. 2008; Kennedy & Kenyon 2008; planet occurrence is intimately tied to the physical prop- Mordasini et al. 2009) and pointed the way toward addi- erties of the the stars they orbit (Johnson et al. 2010a; tional exoplanet discoveries (Laughlin 2000; Marois et al. Howard et al. 2011b; Schlaufman & Laughlin 2011). Ra- 2008). dial velocity surveys have demonstrated that the likeli- The increased abundance of giant planets around hood that a star harbors a giant planet with a mini- massive, metal-rich stars may be a reflection of their mum mass M sin i & 0.5 M increases with both stel- more massive, dust-enriched circumstellar disks, which P Jup form protoplanetary cores more efficiently (Ida & Lin lar metallicity and mass10 (Gonzalez 1997a; Santos et al. 2004; Fischer & Valenti 2005; Thommes & Murray 2006; 2004; Fischer & Valenti 2005; Johnson et al. 2010a; Wyatt et al. 2007). In the search for additional Schlaufman & Laughlin 2010; Brugamyer et al. 2011). planets in the Solar neighborhood, metallicity-biased Doppler surveys have greatly increased the number of [email protected] 1 close-in, transiting exoplanets around nearby, bright Based on observations obtained at the W.M. Keck Observa- stars, thereby enabling detailed studies of exoplanet tory, McDonald Observatory and the Hobby-Ebberly Telescope. Keck is operated jointly by the University of California and the atmospheres (Fischer et al. 2005; da Silva et al. 2006; California Institute of Technology. Keck time has been granted Charbonneau et al. 2007). Similarly, future high- by Caltech, the University of Hawaii, NASA and the University contrast imaging surveys will likely benefit from enrich- of California. 2 ing their target lists with intermediate-mass A- and F- Department of Astrophysics, California Institute of Technol- ogy, MC 249-17, Pasadena, CA 91125 type stars (Marois et al. 2008; Crepp & Johnson 2011). 3 arXiv:1108.4205v1 [astro-ph.EP] 21 Aug 2011 Institute for Astronomy, University of Hawai’i, 2680 Wood- Occurrence rate is not the only aspect of exoplanets lawn Drive, Honolulu, HI 96822 4 that correlates with stellar mass. Just when exoplanet Department of Astronomy, University of California, Mail researchers were growing accustomed to short-period and Code 3411, Berkeley, CA 94720 5 Center of Excellence in Information Systems, Tennessee highly eccentric planets around Sun-like stars, surveys State University, 3500 John A. Merritt Blvd., Box 9501, of evolved stars revealed that the orbital properties of Nashville, TN 37209 planets are very different at higher stellar masses. Stars 6 McDonald Observatory, University of Texas at Austin, TX, 78712-0259, USA more massive than 1.5 M⊙ may have a higher overall 7 Department of Astronomy & Astrophysics, The Pennsylva- occurrence of Jupiters than do Sun-like stars, but they nia State University, University Park, PA 16802 exhibit a marked paucity of planets with semimajor axes 8 Center for Exoplanets and Habitable Worlds, The Pennsyl- a . 1 AU (Johnson et al. 2007; Sato et al. 2008a). This vania State University, University Park, PA 16802 9 NASA Exoplanet Science Institute (NExScI), CIT Mail Code is not an observational bias since close-in, giant planets 100-22, 770 South Wilson Avenue, Pasadena, CA 91125 produce readily detectable Doppler signals. There is also 10 Some studies indicate a lack of a planet-metallicity re- growing evidence that planets around more massive stars lationship among planet-hosting K-giants (Pasquini et al. 2007; tend to have larger minimum masses (Lovis & Mayor Sato et al. 2008b). However, a planet-metallicity correlation is evi- dent among subgiants, which probe an overlapping range of stellar 2007; Bowler et al. 2010), and occupy less eccentric or- masses and convective envelope depths (Fischer & Valenti 2005; bits compared to planets around Sun-like stars (Johnson Johnson et al. 2010a; Ghezzi et al. 2010). 2008). 2 M-type dwarfs also exhibit a deficit of “hot Jupiters,” albeit with a lower overall occurrence of giant planets 1.5 at all periods (Endl et al. 2003; Johnson et al. 2010a). However, a recent analysis of the transiting planets 1.0 detected by the spaced-based Kepler mission shows that the occurrence of close-in, low-mass planets (P < ) 0.5 50 days, MP . 0.1 MJup) increases steadily with decreas- ing stellar mass (Howard et al. 2011b). Also counter Sun to the statistics of Jovian planets, low-mass planets log(L/L 0.0 are found quite frequently around low-metallicity stars (Sousa et al. 2008; Valenti et al. 2009). These results strongly suggest that stellar mass is a key variable in the −0.5 formation and subsequent orbital evolution of planets, and that the formation of gas giants is likely a threshold process that leaves behind a multitude of “failed cores” −1.0 6500 6000 5500 5000 4500 with masses of order 10 M⊕. Teff [K] To study the properties of planets around stars more massive than the Sun, we are conducting a Doppler sur- Fig. 1.— Distribution of the effective temperatures and lumi- vey of intermediate-mass subgiant stars, also known as nosities of the Keck sample of subgiants (filled circles) compared the “retired” A-type stars (Johnson et al. 2006). Main- to the full CPS Keck target sample (gray diamonds). sequence stars with masses greater than ≈ 1.3 M⊙ (spec- tral types . F8) are challenging targets for Doppler sur- target sample of the California Planet Survey (CPS). veys because they are hot and rapidly rotating (Teff > −1 6300, Vr sin i& 30 km s ; Galland et al. 2005). How- 2.2. Spectra and Doppler-Shift Measurements ever, post-main-sequence stars located on the giant and We obtained spectroscopic observations of our sample subgiant branches are cooler and have much slower rota- of subgiants at Keck Observatory using the HIRES spec- tion rates than their main-sequence cohort. Their spec- trometer with a resolution of R ≈ 55, 000 with the B5 tra therefore exhibit a higher density of narrow absorp- ′′ decker (0. 86 width) and red cross-disperser (Vogt et al. tion lines that are ideal for precise Doppler-shift mea- 1994). We use the HIRES exposure meter to ensure that surements. all observations receive uniform flux levels independent Our survey has resulted in the detection of 16 plan- of atmospheric transparency variations, and to provide ets around 14 intermediate-mass (M & 1.5 M⊙) stars, ⋆ the photon-weighted exposure midpoint which is used for including two multiplanet systems, the first Doppler- the barycentric correction. Under nominal atmospheric detected hot Jupiter around an intermediate-mass star, conditions, a V = 8 target requires an exposure time of and 4 additional Jovian planets around less massive 90 seconds and results in a signal-to-noise ratio of 190 subgiants (Johnson et al. 2006, 2007, 2008; Bowler et al. ˚ 2010; Peek et al. 2009; Johnson et al. 2010c,b, 2011). In at 5800 A for our sample comprising mostly early K-type this contribution we announce the detection of 18 new gi- stars. ant exoplanets orbiting subgiants spanning a wide range Normal program observations are made through a of stellar physical properties. temperature-controlled Pyrex cell containing gaseous io- dine, which is placed just in front of the entrance slit 2. OBSERVATIONS AND ANALYSIS of the spectrometer. The dense set of narrow molecu- lar lines imprinted on each stellar spectrum from 5000 A˚ 2.1. Target Stars to 6200 A˚ provides a robust, simultaneous wavelength The details of the target selection of our Doppler sur- calibration for each observation, as well as informa- vey of evolved stars at Keck Observatory have been de- tion about the shape of the spectrometer’s instrumen- scribed in detail by Johnson et al. (e.g. 2006); Peek et al. tal response (Marcy & Butler 1992). Radial veloci- (e.g. 2009); Johnson et al. (e.g. 2010c). In summary, ties (RV) are measured with respect to an iodine-free we have selected subgiants from the Hipparcos catalog “template” observation that has had the HIRES in- (van Leeuwen 2007) based on B − V colors and absolute strumental profile removed through deconvolution. Dif- magnitudes MV so as to avoid K-type giants that are ob- ferential Doppler shifts are measured from each spec- served as part of other Doppler surveys (e.g.