EDP Sciences
Free Access
Issue
A&A
Volume 583, November 2015
Article Number A85
Number of page(s) 24
Section Catalogs and data
DOI https://doi.org/10.1051/0004-6361/201526795
Published online 30 October 2015

© ESO, 2015

1. Introduction

Table 1

Sources of the β Pictoris stellar sample.

Wide binaries provide valuable information about key questions in astrophysics; for example, halo-wide pairs contribute to constraining the properties of dark matter (Weinberg et al. 1987; Yoo et al. 2004; Quinn et al. 2009), some star formation theories depend on the frequency and separation of wide young binaries (Parker et al. 2009; Ward-Duong et al. 2015; Marks et al. 2015), and relatively bright FGK-type primaries with M-dwarf companions provide a metallicity calibration yardstick for cool stars (Bonfils et al. 2005; Rojas-Ayala et al. 2012; Newton et al. 2014; Li et al. 2014). However, the maximum projected physical separation of a wide binary is still a matter of discussion: Some authors consider a cutoff in the number of wide binaries at 2 × 104 au (~0.1 pc), which is the typical size of protostellar cores (Tolbert 1964; Abt 1988; Wasserman & Weinberg 1991; Allen et al. 2000; Tokovinin & Lépine 2012), while others contemplate separations of 2 × 105 au (~1 pc) or more (Jiang & Tremaine 2009; Caballero 2009; Shaya & Olling 2011). Such wide common proper-motion pair candidates, which give their name to the title of this series of papers, can be either unbound members of the same young stellar kinematic group that by chance are co-moving (Tokovinin 2014a) or bound “binaries” of very low binding energies at the limit of disruption (Caballero 2010).

The younger a weakly bound system is, the less time it has had to be disrupted (Bahcall & Soneira 1981; Retterer & King 1982; Weinberg et al. 1987; Saarinen & Gilmore 1989; Poveda & Allen 2004). As a result, a search for multiple systems within a young stellar kinematic group (moving group or stellar association) offers a unique opportunity for finding new faint benchmark objects hardly influenced by the Galactic gravitational potential, but instead by their formation process. In other words, the shape of young wide binaries is dominated by nature instead of by nurture.

In this work, we use the profitable technique of searching for common proper-motion pairs of wide separation (e.g., Luyten 1979; Chanamé & Gould 2004) in a close and very young moving group, namely β Pictoris (Zuckerman et al. 2001b; Ortega et al. 2002; Song et al. 2003). Although there is no consensus in the literature, the β Pictoris age lies in a relatively narrow interval between 11 Ma and 26 Ma (Barrado y Navascués 1998; Torres et al. 2006; Yee & Jensen 2010; Binks & Jeffries 2014; Mamajek & Bell 2014, and references therein). Known moving group members and member candidates lie at between 6 pc and 80 pc from our Sun with a median distance of 40 pc.

Because of its youth and proximity, the β Pictoris moving group has been relevant for studying resolved debris discs with high angular resolution observations (Smith & Terrile 1984; Metchev et al. 2005; Boccaletti et al. 2009; Churcher et al. 2011; Wahhaj et al. 2013; Dent et al. 2013) and exoplanets through direct imaging (Mouillet et al. 1997; Neuhäuser et al. 2003; Kasper et al. 2007; Lagrange et al. 2009, 2010; Bonnefoy et al. 2011, 2013; Biller et al. 2013; Rameau et al. 2013; Males et al. 2014; Bowler et al. 2015; Macintosh et al. 2015) or for comparing observations with evolutionary models (Crifo et al. 1997; Song et al. 2002; Cruz et al. 2009; Biller et al. 2010; Mugrauer et al. 2010; Jenkins et al. 2012; Montet et al. 2015). Therefore, increasing the number of members via common proper-motion companionship, especially at low masses, can help to inform the previously mentioned fields and to constrain the age of the group. Besides that, identiying bright M-dwarf targets of ~10–30 Ma for extremely precise radial velocity surveys is becoming critical for understanding the formation and early evolution of terrestrial planets in habitable zones (Lissauer 2007; Ramírez & Kaltenegger 2014; Luger et al. 2015; Tian 2015; Tian & Ida 2015). Preliminary results of this work, including the discovery of two new stellar members in the β Pictoris moving group, were given in Alonso-Floriano et al. (2011).

2. Analysis

2.1. Stars sample

We have compiled in Table A.1 a list of 185 β Pictoris members and member candidates around which we looked for common proper-motion companions. We gathered them from 35 previous works published in the past 16 years from the first articles of Barrado y Navascués et al. (1999) and Zuckerman et al. (2001b) to the last investigations published in the SACY (Search for Associations Containing Young stars – Torres et al. 2006; Elliott et al. 2014, 2015) and BANYAN series (Bayesian Analysis for Nearby Young AssociatioNs – Malo et al. 2014a,b; Gagné et al. 2015). Table 1 lists all works that we searched through.

We cross-matched our list with the latest Geneva-Copenhagen catalogue (Holmberg et al. 2009) and identified 17 bright stars for which metallicity was available. From these data, we determined a solar metallicity of the β Pictoris moving group of [Fe/H] = –0.2±0.2. In Table A.1, we provide for each star: discovery (or recommended) name, right ascension and declination from the Two-Micron All-Sky Survey (Skrutskie et al. 2006), heliocentric distance, its uncertainty when available, and corresponding reference. We follow the nomenclature convention of Alonso-Floriano et al. (2015). In particular, we provide for the first time the ROSAT precovery names (1RXS) for several stars for which no X-ray counterpart had been identified by subsequent proper-motion surveys.

In the last column, we also list a flag indicating the quality of the star membership in β Pictoris:

  • 1.

    Uncontrovertible moving group members for which at leasttwo independent research groups have declared them to be bonafide moving group members and whose memberships have notbeen put in doubt afterwards. In general, these objects havecoherent kinematics (with reliable distance and radial velocitydetermination) and youth features (coronal X-ray andchromospheric Hα emission, lithium in absorption and, in somecases, debris discs).

  • 2.

    Moving-group member candidates for which there is no definitive confirmation of true membership.

  • 3.

    Dubious moving group member candidates that have also been proposed as belonging to other young moving groups of similar kinematics, or even to the field. We include them in our work for completeness.

2.2. Proper motion companion candidates

For this search, we made extensive use of virtual observatory tools. We used the comprehensive PPMXL proper motion catalogue (Roeser et al. 2010), the Aladin sky atlas (Bonnarel et al. 2000), and the Starlink Tables Infrastructure Library Tool Set (STILTS; Taylor 2006) to look for common proper-motion companions to the 185 β Pictoris stars in Table A.1. The PPMXL catalogue is complete down to the visual magnitude V ≈ 20 mag and has typical individual mean errors of the proper motions between 4 and 10 mas/a, approximately. We applied the following selection criteria in our search.

  • We looked for companion candidates in a circular area of angular radius ρ = s/d (in arcsec) centred on each sample star, where s is the maximum projected physical separation, fixed at s = 105 au, and d (in pc) is the heliocentric distance shown in Table A.1. At the given distances, the search radii varied between over 4 deg for the closest β Pictoris stars (e.g., YZ CMi AB at 5.96±0.08 pc) and 12 to 23 arcmin for the most distant ones (e.g., LP 58–170 at 140±40 pc and V4046 Sgr AB and C at 73±18 pc). The median search radius was 44 arcmin.

  • We discarded from the survey 24 stars with total PPMXL proper motions μ < 50 mas/a (19) or no proper motions at all (5). Therefore, we looked for companions of 161 β Pictoris stars. Stars slower than μ = 50 mas/a were not considered at this step because of the large number of potential candidates with relative uncertainties of 10%–30% in proper motion that would fall in the surveyed area and pass the filter. As proper motion companion candidates, we classified only the PPMXL sources with a 2MASS counterpart for which the values of μαcosδ and μδ lie within 10% of those of the primary target (see Fig. 1).

    thumbnail Fig. 1

    Representative proper-motion diagram of all PPMXL sources brighter than J = 15.5 mag in a 30 arcmin-radius circular area centred on LP 648–20. The red square box in the bottom left indicates the proper-motion search area around LP 648–20, marked with a filled circle. The open circle corresponds to the bright, young G5 V star EX Cet.

    Open with DEXTER

  • We retained objects brighter than J = 15.5 mag. In general, fainter sources in the near-infrared also have very faint magnitudes in the optical, close to the limit of the USNO-B1 (Monet et al. 2003) digitisations of BJ, RF, and IN photographic plates, which were used by PPMXL. This faintness translates into large astrometric errors in the PPMXL proper motions. Keeping relatively bright sources assures the quality of the compiled astro-photometric measurements (see below), although prevents detecting fainter and, thus, low-mass β Pictoris members in, perhaps, the substellar domain.

Once we had a preliminary list of candidates, we inspected all of them visually with Aladin and the images and data of various all-sky surveys (Palomar Observatory Sky Survey I and II; 2MASS; SDSS-DR9, Ahn et al. 2012; WISE, Cutri et al. 2012, 2014; CMC14 and CMC15, Evans et al. 2002). In particular, we checked that the companion candidates have a unique and reliable entry in the PPMXL catalogue (e.g., at least four astrometric detections, no other PPMXL source at less than 2 arcsec, smooth variation of the magnitudes from BJ, through RJ, IN, J, H, Ks, to WISE W1–4). In this step, we discarded a number of preliminary companion candidates with erroneous PPMXL proper motions (i.e., with incorrect USNO-B1 matches) owing to close visual multiplicity or source confusion in very crowded fields at low Galactic latitudes. Some of the mistaken sources were identified around α Cir, 1RXS J171502.4–333344, V4046 Sgr, 1RXS J184956.1–013402, which have | b |< 7 deg, and, especially, V343 Nor, which is at less than 2 deg of the Galactic plane and, besides this, towards the Galactic centre. After this visual pre-cleaning, we obtained a list of 92 proper motion companion candidates to 65 β Pictoris stars.

Next, we performed a 10 arcsec-radius cross-match on our initial 185-star sample with the Washington Double Star catalogue (WDS – Mason et al. 2001, 2015). We got 163 positive cross-matches in 55 WDS systems. Of the cross-matches, 136 corresponded to close physical binaries not resolved by 2MASS nor PPMXL (ρ ≲ 2.5 arcsec) or to wider multiple systems, but with large magnitude differences measured with powerful adaptive optics systems (e.g., Lafrenière et al. 2007; Chauvin et al. 2010). The list of WDS systems unresolved or unidentified in our search are shown in Table A.2, which provides the star name, WDS discovery name (for resolved pairs) or reference (for spectroscopic binaries), multiplicity status (physical, visual – non-common proper motion –, single/double-line spectroscopic binaries), angular separation (interval of ρ when several visual companions are tabulated), position angle (θ), and WDS identifier (only for resolved pairs). For the physical and visual pairs in Table A.2, ρ and θ correspond to the latest epoch listed by WDS.

Thanks to the cross-match with WDS, we were able to add another 15 previously known secondaries detected by 2MASS to our list of 92 proper motion companion candidates. They did not pass our filters above because they have PPMXL proper motions that deviate more than 10% from those of the “primary”, probably because of relative orbital motion, erroneous measurements in right ascension and/or declination, proper motions with 1σ lower limits below the 50 mas/a boundary, or no proper motions at all.

thumbnail Fig. 2

PPMXL vs. adopted proper-motion diagrams in right ascension (left panel) and declination (right panel). The red dashed and dotted lines mark the one-to-one relationship and the 10% error area above and below it, respectively.

Open with DEXTER

Finally, we also added to our list the companion candidates of three additional pairs of β Pictoris stars that were not in WDS and that were not detected because of the reasons explained above: [SLS2012] PYC J02017+0117N & S (ρ ~ 10 arcsec and equal brightness) and TYC 112–917–1 & 2E 1249 AB (μ 41 mas/a), which have the same predicted or measured distances and radial velocities (Schlieder et al. 2012a; Elliott et al. 2014) and are quite obvious proper motion companion candidates in Aladin, and V4046 Sgr AB and C, which was presented by Kastner et al. (2011).

In Table A.3, we list the 110 (92+15+3) proper motion companion candidates that passed on to the next analysis stage.

2.3. Astro-photometric follow-up

2.3.1. Astrometry

We performed a dedicated astro-photometric follow-up of the 110 companion candidates in two steps. In the first one, we confirmed true common proper motion of the pairs with a precise astrometric study. This step was necessary because PPMXL used the astro-photometric USNO-B1 catalogue as input, which is known to be affected by systematics at the fainter optical magnitudes, especially when dealing with high proper motion stars.

Of the 184 objects in Table A.3 (74 primaries and 110 companion candidates), 55 had reliable proper motions measured by Hipparcos (TYC, Høg et al. 2000; HIP2, van Leeuwen 2007). For the remaining 129 objects, we measured precise proper motions from public data in virtual observatory catalogues as in Caballero (2010, 2012). In particular, we used astrometric epochs from the following catalogues: AC2000.2 (Urban et al. 1998), USNO-A2 (Monet 1998), GSC2.3 (Lasker et al. 2008), DENIS (Epchtein et al. 1997), CMC14 and CMC15, 2MASS, SDSS, and WISE. To maximise the number of astrometric epochs, N, and time baseline, Δt, of the follow-up, which translates into reducing the uncertainty in proper motion, we also used the SuperCOSMOS digitisations of Palomar Observatory Sky Survey photographic plates, especially for the faintest objects (Hambly et al. 2001; cf., Caballero 2012). The addition of SuperCOSMOS data allowed us to get at least four accurate astrometric epochs spread over a minimum of 11.5 a for all targets except for one star (2MASS J05113065–2155189, N = 3). The average number of astrometric epochs and time baseline were five and 34 a, respectively. In the extreme case of TYC 4571–1414–1, we measured its proper motion with eight astrometric epochs spread over almost 115 a. Table A.3 lists the 2MASS coordinates of the 184 “primaries” and companion candidates, and their PPMXL and adopted proper motions. For the adopted proper motions that do not come from TYC or HIP2, Table A.3 also provides the time baseline and number of epochs used in our astrometric follow-up. Except for partially resolved close binaries (e.g., AT Mic AB) or faint sources (r 16 mag), we were able to measure proper motions with typical uncertainties of 1 mas/a or less, which are comparable to or even better than TYC or HIP2.

We show a comparison of the original PPMXL proper motion values and the ones adopted by us in Fig. 2. While the values of proper motions in right ascension provided by PPMXL have in general good agreement with our adopted values, many PPMXL proper motions in declination have greater absolute values.

With the new data, we made a second, more precise, astrometric filtering and discarded 43 visual companions with adopted proper motions that deviate more than 10% from the proper motion of the system. This step of the follow-up thus left 67 physical companion candidates for the second step. In general, the astrometrically rejected objects are distant background stars with fake high proper motions in the PPMXL catalogue, which are located close to bright stars and were reported previously as companion candidates or which are located at large angular separations to our primary targets and have by chance similar, but not identical, proper motions. Of the rejected stars, four were catalogued companion candidates of c Eri, α Cir, CD–24 16238, and AF Psc, and one was a faint companion candidate to [SLS2012] PYC J10175+5542 (Schlieder et al. 2012a) that had been reported by W. J. Luyten (LDS 2851, WDS 10176+5542).

thumbnail Fig. 3

Optical-to-near-infrared colour-magnitude diagrams of our β Pictoris stars and common proper-motion companion candidates. In both panels, blue filled circles mark bona fide moving group members with membership flag 1 in Table A.1 (see Sect. 2.1), and open circles indicate other member candidates with flags 2 and 3. Black filled squares denote companions previously known in the literature that had not been reported as belonging to β Pictoris. Red filled stars mark our eight new proper motion companions. Typical error bars are shown in the bottom left corner. The black solid line is the average β Pictoris sequence computed with all candidate members as in the left panel (flags 1, 2, and 3). Left panel: MB vs. BKs diagram. The dash-dotted lines are the β Pictoris sequence shifted by ± 3σ. Grey times (×) and crosses (+) indicate discarded companion candidates from astrometry and photometry, respectively; squared symbols mark companion candidates in the WDS. Right panel: MV vs. VKs diagram. The blue dotted line is the sequence with only bona fide members (flag 1). Red dashed and dash-dotted lines are the 20 and 100 Ma isochrones from Baraffe et al. (2015). Green long dashed line is the 20 Ma isochrone from Siess et al. (2000) plotted only at highest masses. For clarity, we do not draw the discarded companion candidates. Some remarkable stars do not have V photometry.

Open with DEXTER

2.3.2. Photometry

In the second step of the follow-up, we studied the membership in the β Pictoris moving group of the 67 companion candidates that passed the previous astrometric filter with the help of colour-magnitude diagrams and theoretical isochrones.

First, we compiled B, V, r, J, H, Ks, and W1–4 magnitudes for all the sources investigated in this work. While infrared JHKsW1–4 photometry came in all cases from 2MASS and WISE, the origin of the optical BVr photometry was diverse. When available, we collected BVr photometry from UCAC4 (Zacharias et al. 2013). If not available, we got it from a number of sources: Tycho-2 (B and V, after transformation from BT and VT magnitudes), USNO-B1 (B, after average of two photographic BJ magnitudes), AC2000.2 (only one star), SPM4 (Girard et al. 2011V, only one star), CMC 15 (r), SDSS-DR9 (r), or the literature (Voges et al. 1999; Bakos et al. 2002; Torres et al. 2006; Beichman et al. 2010; Smart 2013).

We were not able to compile optical BVr photometry for all our targets. Since we were able to compile magnitudes for more stars in the B band, we applied our photometric filtering using the reddest near-infrared band, Ks, and the bluest optical one, B. Actually, we failed to find reliable B photometry for only six stars: five low-mass stars or brown-dwarf candidates with spectral types at the M/L boundary, of which four are from Gagné et al. (2014, 2015), one is the known companion of L 186–67 A (see below), and the sixth one is a star close to the bright primary V343 Nor A. The reddest and faintest object in our sample with B and Ks photometry is 2MASS J06085283-2753583 (M8.5 V, Luhman et al. 2009; BKs = 9.1 mag). The use of WISE photometry did not provide significant improvement over the use of 2MASS Ks.

We performed our photometric filtering in a recursive scheme:

  • First we computed the B-band absolute magnitude MB with the heliocentric distances in Table A.1 and built the MB vs. BKs diagram in left-hand panel of Fig. 3. Any new companion candidate would have to be located at the same distance as the target star.

  • We defined an average β Pictoris sequence with all members and candidates in Table A.1 (flags 1, 2, and 3). All sources that did not pass the astrometric filter lie in the locus of background stars in the colour-magnitude diagram.

  • We checked the reliability of our average sequence by comparing it with the latest evolutionary models by Baraffe et al. (2015). Since BT-Settl does not provide MB magnitudes, we used MV ones instead. We built the MV vs. VKs diagram in right-hand panel of Fig. 3 and plotted the corresponding BT-Settl 20 and 100 Ma isochrones. The acceptable match between our average MV vs. VKs sequence and the 20 Ma isochrone encouraged us to use our average MB vs. BKs sequence for the photometric filter.

  • We discarded stars with absolute magnitudes MB and colours BKs inconsistent with the β Pictoris sequence. Most discarded stars lie outside the ± 3σ area around the sequence. The systematic error introduced by mixing different photometric systems for the blue magnitude seems to be smaller than the intrinsic scatter in the β Pictoris sequence, mostly due to uncertainties in distance.

Of the previous 67 stars, 31 did not pass the photometric filter. One known system, L 186–67 Aa, Ab, B, could not be studied photometrically because of the lack of reliable data in the optical, but the short angular separation between components and the large common proper motions ensured that it is a physical system.

3. Results and discussion

3.1. Known and new common proper motion pairs

From the initial list of 184 common proper motion companion candidates to β Pictoris stars in Sect. 2.2, only 36 targets passed the two filters of our astro-photometric follow-up in Sect. 2.3. Our final list of confirmed common-proper motion systems in the β Pictoris moving group, as shown in Table A.5, consists of

  • Eighteen known systems in which the two stars had been reported previously to belong to the moving group. All of them except one are listed by WDS; the exception is the wide system formed by V4046 Sgr AB and V4046 Sgr C, which was proposed and investigated for the first time by Kastner et al. (2011). Some of the 17 WDS systems have been known for decades, such as five pairs in the W. J. Luyten’s Double Star catalogue or the HD 14082 AB pair, which was resolved for the first time by F. G. W. Struve in 1821.

  • Ten known systems in which only one star had been reported previously as belonging to the moving group. Of the ten stars that had not been reported as belonging to β Pictoris (i.e., not listed in Table A.1), four displayed significant X-ray emission in ROSAT observations (Voges et al. 1999; Riaz et al. 2006; Kaplan et al. 2006; Haakonsen & Rutledge 2009) and two showed intense Hα emission at the chromospheric/accretion boundary for their spectral types (pEW(Hα) –12 to –16 Å – Reid et al. 1995; Riaz et al. 2006). Since there is one star that is an Hαand X-ray emitter (2MASS J00193931+1951050), half of the ten new stars have known signposts of youth, which supports membership in β Pictoris. Besides this, another one has a similar radial velocity to the primary in the system (CD–44 753 A and B – Kordopatis et al. 2013). For the other four new young stars, there are only photometric data available (and, in the case of 2MASS J07293670+3554531, mass and spectral type derived from photometry – Pickles & Depagne 2010; Janson et al. 2012).

  • Eight new common proper motion systems with β Pictoris stars. In reality, there are WDS entries for two β Pictoris pairs that were presented for the first time by Alonso-Floriano et al. (2011): EX Cet A, B (CAB 3) and HD 173167 A, B (CAB 8). Although the results from this preliminary publication have already been used by other authors (Shkolnik et al. 2012; Eisenbeiss et al. 2013; Bowler et al. 2015), we consider their discovery as part of this work. Moór et al. (2013) “rediscovered” the pair HD 173167 A, B, although they did not report ρ or θ. The optical spectra of these two stars and of TYC 112–917–1 and 2E 1249 AB in the new pair WDS 05200+0613 display intense Li i λ6707.8 Å line in absorption for their spectral type (Alcalá et al. 2000; Torres et al. 2006), which supports their extremely young age. Six of these stars are reported here as new member candidates in the β Pictoris moving group.

On some occasions we use the term “pair” to refer to multiple systems that contain only two components resolvable from the ground with standard imaging (i.e., no adaptive optics or lucky imaging) and spectroscopic devices. Most of our systems are such pairs. However, Table A.5 lists 12 triple and two quadruple hierarchical systems that contain one or two close pairs unresolved by public catalogues (Table A.2). The two quadruple systems are MV Vir Aa, Ab, B, C and HD 199143 AB, CD (for which the close components were resolved first by Jayawardhana & Brandeker 2001). The latter has an “A. Tokovinin” WDS entry dated after 2011, but the wide multiplicity was previously reported by Alonso-Floriano et al. (2011) and, especially, Zuckerman et al. (2001b).

The existence of 14 triples and quadruples in a list of 36 multiples provides a high-order-multiple ratio of about 1:3, which is unexpectedly high. Law et al. (2010) found a similar ratio of about 1:2 for wide M-dwarf binaries of the field and suggest that some of the binaries with large separations are actually triple and quadruple systems. (Actually, Caballero 2007 and Burgasser et al. 2007 pointed it out before.) The increment of the high-order-multiple fraction for the widest systems is supported by the work of Reipurth & Mikkola (2012), who used N-body simulations of the dynamical evolution of triple systems to suggest that loosely bound triple systems might appear to be very wide binaries. However, recent dedicated surveys for multiplicity of F, G, K (Tokovinin et al. 2014b; Elliott et al. 2015) and M dwarfs in the field (Cortés-Contreras et al. 2014) have found lower ratios of about 1:10. Although our sample comprises a wide range of masses and separations, it is not large enough to do an appropriate comparison with the previously mentioned works. Another explanation might be an observational effect of a biased sample in which surveys for nearby young stars are naturally slanted towards detecting intrinsically bright binaries and multiple stars (Malmquist bias), and active spectroscopic binaries (very close separations enhances stellar activity). The discovery of new moving groups members only based on astrometry, as in this survey, may help to alleviate this observational bias.

In Table A.5 we list our WDS identifiers in italics if they are not included in the WDS catalogue at the time of writing these lines (i.e., V4046 Sgr AB,C and six of the eight new pairs). In total, in this survey we propose 16 new stellar members of the β Pictoris moving group: six in new pairs and ten in known systems with only one reported young star. One of the new β Pictoris stars in a new pair is HD 173167 A, which was discovered by Alonso-Floriano et al. (2011) and classified afterwards as a moving group member by Moór et al. (2013). These values represent an increase of 9% in the total number of reported β Pictoris stars and of almost 30% in the number of wide proper motion systems in the moving group. We ran the on-line BANYAN tool1 (Malo et al. 2013) on the 16 new proposed members of β Pictoris and calculated approximate membership probabilities (Table 2). We used the distances of the systems provided in Table A.5, our proper motion measurements in Table A.3, and radial velocities from the literature (for those objects without radial velocity measurements, we assumed the values of their companions). Although only seven of the 16 pairs showed high-probability memberships to β Pictoris (see Table 2), these results should be used with caution because most of the new candidates lack accurate distances or radial velocities.

None of the new reported wide systems have parallax measurements for both components. However, the location of the 16 stars (eight primaries and eight secondaries) in the colour-magnitude diagrams suggests that both components are located at similar distances. Definitive parallactic confirmation of common distance will have to wait until early 2017 with the second Gaia release. In the meantime, we can infer the true physical binding of the systems with the computation of the reduced gravitational binding energy.

Table 2

Membership probabilities for the 16 new β Pictoris candidates using the BANYAN on-line tool.

3.2. Projected separations and binding energies

In Table A.5 we list the angular separations, ρ, and position angles, θ, at the 2MASS epoch of observation of the 36 wide pairs in the β Pictoris moving group. Angular separations vary from 8.2 arcsec for BD–21 1074A, Ba, Bb to about 1.3 deg for the triple system AU Mic–AT Mic AB (Luyten 1941; Caballero 2009). Among our new pairs, ρ varies from 10.6 arcsec to 24.5 arcmin.

To distinguish between true very wide physical binaries and co-moving pairs of “single” stars that belong to the same kinematic group, we computed the reduced gravitational binding energies, (Caballero 2009), of the 36 systems. With the angular separations and distances, we obtained the projected physical separations, s, which vary from merely 100–120 au for the known pairs WDS 08228–5727 (L 186–67 Aa, Ab, B) and WDS 10596+2527 (HD 95174 AB) to about 7 × 104 au (0.34 pc) for the new pair WDS 08290+1125. Given the uncertainties in the distance (Table A.1), we provide only two significant figures for s.

We derived masses 1 and 2 from J-band absolute magnitudes MJ and the Baraffe et al. (2015) or Siess et al. (2000) evolutionary models at 20 Ma for solar metallicity and the appropriate mass intervals. When available, we gathered masses of single early type stars and close binaries from the literature (e.g., Strassmeier & Rice 2000; Neuhäuser et al. 2002; Caballero 2009; Donati et al. 2011; Janson et al. 2012; Elliott et al. 2015; Montet et al. 2015) or suitable information that allowed us to make a precise derivation (e.g., magnitude differences from adaptive optics or lucky imaging, mass ratios from spectroscopic monitoring – Chauvin et al. 2010; Neuhäuser et al. 2011; Janson et al. 2012; Messina et al. 2014; Bowler et al. 2015; Elliott et al. 2015). Masses range approximately from 2.4 M for η Tel A to well below the substellar boundary for L 186–67 B with a broad maximum of the distribution at 0.5–1.0 M. Derived masses reasonably match those expected from spectral types, when available. For the sake of completeness, we also list spectral types compiled from a number of sources in Table A.5 (Riaz et al. 2006; Reid et al. 2007; Pickles & Depagne 2010; Caballero 2012; Janson et al. 2012; Kraus et al. 2014; Messina et al. 2014; Rodríguez et al. 2014; Mason et al. 2015; I. Gallardo & M. Gómez Garrido, priv. comm.; SIMBAD).

The greatest absolute value of reduced binding energy among the 36 systems in Table A.5, of  J, corresponds to the strongly bound pair HD 95174 AB, which is not only the tightest one, but also contains two stars of ~0.8 M. On the other hand, there are two very fragile system candidates with binding energies of 0.57–2.7 1033 J, almost one order of magnitude lower than that of the Luyten’s system AU Mic+AT Mic AB, which lies at the boundary between very wide binaries and couples of single stars that are co-moving within the same stellar kinematic group (Caballero 2010; see the title of this series of papers). As a result, it is likely that the components in the two new fragile system candidates WDS 08290+1125 and WDS 23317–0245, which includes the flaring star AF Psc (Bond 1976; Kraus et al. 2014; Ramsay & Doyle 2014), originated in the same parental cloud and were ejected at the same time, in the same direction, and at the same velocity, but they are not physically bound. The six other new pairs have binding energies between 13 and 1400 × 1033 J and may survive the eventual disruption by the Galactic gravitational potential for some billion years (Weinberg et al. 1987; Close et al. 2007). In any case, detecting features of youth in the spectra of WDS 08290+1125 A and B and the wide M6.0 V companion candidate to AF Psc (Reid et al. 2007) would shed light on their actual membership in the β Pictoris moving group.

3.3. Benchmark objects and probable members in other young moving groups

The 36 wide systems tabulated by us can help to constrain the actual membership of some controversial candidate members in β Pictoris:

  • WDS 01367–0645. Some authors have alsoclassified the primary of the system, EX Cet(G5 V), as a member of the Hercules-Lyraassociation (~100–200 Ma – Monteset al. 2001; López-Santiago et al.2006; Shkolnik et al. 2012; Eisenbeisset al. 2013).

  • WDS 02305–4342. The primary CD–44 753 A is also a member candidate of the Columba association (~15–50 Ma – Torres et al. 2006, 2008; Elliott et al. 2014; Malo et al. 2014a).

  • WDS 08228–5727. The membership of the primary L 186−67 Aa, Ab to β Pictoris is ambiguous (Malo et al. 2013, 2014a). The late-M common proper motion companion, L 186–67 B, whose physical binding in the system had been confirmed earlier (Bakos et al. 2002; Bergfors et al. 2010; Janson et al. 2012, 2014), would have a mass close to the deuterium-burning mass limit if it were 20 Ma old. If membership in β Pictoris were confirmed, the triple system would be a benchmark for very low-mass substellar astrophysics.

  • WDS 09361+3733. While there are no membership studies for the primary, the homonymous secondary HD 82939 Ba,Bb was listed not only as a β Pictoris star by Schlieder et al. (2012a,b), but also as a young field star by Malo et al. (2014b).

  • WDS 16172+7734. Schlieder et al. (2012a) listed the primary TYC 4571–1414–1 as a probable member of both β Pictoris and AB Doradus (~70 Ma) moving groups.

  • WDS 21214–6655. The primary star V390 Pav A has also been classified as a member of the Tucana-Horologium association (~30 Ma – Zuckerman et al. 2001a; Mamajek et al. 2004; Rojas et al. 2008).

If the six systems above were eventually discarded as true β Pictoris “pairs”, 30 systems would still remain for further investigation in the young moving group, of which six (20%) are reported here for the first time.

Certain systems in Table A.5 are also particularly important in the low-mass domain, because they can be used to test evolutionary models. Just to cite one example, the secondary of the pairs WDS 16172+7734 (presented here for the first time) and WDS 21105–2711 (Bergfors et al. 2010; Malo et al. 2013, 2014b) lie close to the substellar limit and, therefore, to the lithium depletion boundary. As a result, a high-resolution spectroscopic analysis of both primaries and secondaries could shed more light on the debated age of β Pictoris.

4. Conclusions

We searched through 35 previous publications and compiled an exhaustive list of 185 members and member candidates in the nearby, young (~20 Ma) β Pictoris moving group, around which we looked for common proper-motion companions at projected physical separations of up to 105 au. For that, we made extensive use of the Aladin and STILTS virtual observatory tools and numerous public all-sky catalogues (e.g., WDS, PPMXL, 2MASS).

Of the 184 initial common proper-motion companion candidates, 129 were the subject of a precise astrometric follow-up, by which we measured proper motions with typical uncertainties of only 1 mas/a, and 67 of a multi-band photometric study. Eventually, we discarded five previously reported pairs and retained 36 reliable pair candidates. Of them, 18 and 10 are known systems with both components or only one component classified as β Pictoris members, respectively, and eight are new pairs in the moving group. We also report 16 new star and brown dwarf candidates in β Pictoris for the first time. These values represent an increase of 9% in the total number of reported objects in the moving group and of almost 30% in the number of wide proper motion systems.

We investigated the 36 pairs with available public information in detail. Among them, there are 12 triple and two quadruple systems, which points out to a greater incidence of high-order multiplicity in β Pictoris than in the field, possibly ascribed to a member list biased towards close binaries or an increment of the high-order, multiple fraction for very wide systems.

We measured angular separations and projected physical separations, compiled or derived masses for components in all systems, and computed reduced gravitational binding energies. Two of the new pair candidates could be unbound couples of single stars that are co-moving within β Pictoris, while at least one of the components in six (new and known) pairs have also been reported to belong to other young moving groups and associations (four in Hercules-Lyra, Columba, AB Doradus, Tucana-Horologium) or to the field (two). There are three pairs (one presented here) with masses of secondaries at or below the hydrogen-burning limit, and they can be used as benchmarks for upcoming age-dating works in β Pictoris. Our study provides a comprehensive analysis of the wide multiplicity in one of the closest and youngest moving groups known and, therefore, also serves as input to models of moving-group evolution and eventual dissipation by the Galactic gravitational field.


Acknowledgments

We thank the anonymous referee for the report. This research made use of the Washington Double Star Catalogue, maintained at the US Naval Observatory, the SIMBAD database and VizieR catalogue access tool, operated at Centre de Données astronomiques de Strasbourg, France, and the Spanish Virtual Observatory (http://svo.cab.inta-csic.es). Financial support was provided by the Universidad Complutense de Madrid, the Comunidad Autónoma de Madrid, and the Spanish Ministerios de Ciencia e Innovación and of Economía y Competitividad under grants AP2009-0187, AYA2011-24052, and AYA2011-30147-C03-02, and -03.

References

Appendix A: Long tables

Table A.1

Investigated β Pictoris members and member candidates.

Table A.2

Unresolved or unidentified systems.

Table A.3

Astrometry measurements.

Table A.4

Photometry measurements.

Table A.5

Common proper-motion companion candidates.

All Tables

Table 1

Sources of the β Pictoris stellar sample.

Table 2

Membership probabilities for the 16 new β Pictoris candidates using the BANYAN on-line tool.

Table A.1

Investigated β Pictoris members and member candidates.

Table A.2

Unresolved or unidentified systems.

Table A.3

Astrometry measurements.

Table A.4

Photometry measurements.

Table A.5

Common proper-motion companion candidates.

All Figures

thumbnail Fig. 1

Representative proper-motion diagram of all PPMXL sources brighter than J = 15.5 mag in a 30 arcmin-radius circular area centred on LP 648–20. The red square box in the bottom left indicates the proper-motion search area around LP 648–20, marked with a filled circle. The open circle corresponds to the bright, young G5 V star EX Cet.

Open with DEXTER
In the text
thumbnail Fig. 2

PPMXL vs. adopted proper-motion diagrams in right ascension (left panel) and declination (right panel). The red dashed and dotted lines mark the one-to-one relationship and the 10% error area above and below it, respectively.

Open with DEXTER
In the text
thumbnail Fig. 3

Optical-to-near-infrared colour-magnitude diagrams of our β Pictoris stars and common proper-motion companion candidates. In both panels, blue filled circles mark bona fide moving group members with membership flag 1 in Table A.1 (see Sect. 2.1), and open circles indicate other member candidates with flags 2 and 3. Black filled squares denote companions previously known in the literature that had not been reported as belonging to β Pictoris. Red filled stars mark our eight new proper motion companions. Typical error bars are shown in the bottom left corner. The black solid line is the average β Pictoris sequence computed with all candidate members as in the left panel (flags 1, 2, and 3). Left panel: MB vs. BKs diagram. The dash-dotted lines are the β Pictoris sequence shifted by ± 3σ. Grey times (×) and crosses (+) indicate discarded companion candidates from astrometry and photometry, respectively; squared symbols mark companion candidates in the WDS. Right panel: MV vs. VKs diagram. The blue dotted line is the sequence with only bona fide members (flag 1). Red dashed and dash-dotted lines are the 20 and 100 Ma isochrones from Baraffe et al. (2015). Green long dashed line is the 20 Ma isochrone from Siess et al. (2000) plotted only at highest masses. For clarity, we do not draw the discarded companion candidates. Some remarkable stars do not have V photometry.

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.