The effect of stellar multiplicity on protoplanetary discs. A NIR survey of the Lupus star forming region
Alice Zurlo, Lucas A. Cieza, Megan Ansdell, Valentin Christiaens, Sebastián Pérez, Josh Lovell, Dino Mesa, Jonathan P. Williams, Camilo Gonzalez-Ruilova, Rosamaria Carraro, Dary Ruíz-Rodríguez, Mark Wyatt
MMNRAS , 1–11 (2020) Preprint 26 November 2020 Compiled using MNRAS L A TEX style file v3.0
The effect of stellar multiplicity on protoplanetary discs:A NIR survey of the Lupus star forming region.
Alice Zurlo , (cid:63) , Lucas A. Cieza , Megan Ansdell , Valentin Christiaens ,Sebastián Pérez , Josh Lovell , Dino Mesa , Jonathan P. Williams ,Camilo Gonzalez-Ruilova , Rosamaria Carraro , Dary Ruíz-Rodríguez , Mark Wyatt Núcleo de Astronomía, Facultad de Ingeniería, Universidad Diego Portales, Av. Ejercito 441, Santiago, Chile Escuela de Ingeniería Industrial, Facultad de Ingeniería y Ciencias, Universidad Diego Portales, Av. Ejercito 441, Santiago, Chile National Aeronautics and Space Administration Headquarters, 300 E Street SW, Washington, DC 20546, USA School of Physics and Astronomy, Monash University, Clayton VIC 3800, Australia Departamento de Física, Universidad de Santiago de Chile, Av. Ecuador 3493, Estación Central, Santiago, Chile Institute of Astronomy, University of Cambridge, Madingley Road, Cambridge CB3 0HA, UK INAF - Osservatorio Astronomico di Padova, Vicolo dell’Osservatorio 5, 35122, Padova, Italy Institute for Astronomy, University of Hawaii at Manoa, Honolulu, HI, 96822, USA Instituto de Física y Astronomía, Universidad de Valparaíso, Gran Bretaña 1111, Playa Ancha, Valparaíso, Chile Chester F. Carlson Center for Imaging Science, School of Physics & Astronomy, and Laboratory for Multiwavelength Astrophysics,Rochester Institute of Technology, 54 Lomb Memorial Drive, Rochester NY 14623 USA
ABSTRACT
We present results from a near-infrared (NIR) adaptive optics (AO) survey of pre-main-sequence stars in the Lupus Molecular Cloud with VLT/NACO to identify (sub)stellar com-panions down to ∼
20 au separation and investigate the effects of multiplicity on circumstellardisc properties. We observe for the first time in the NIR with AO a total of 47 targets andcomplement our observations with archival data for another 58 objects previously observedwith the same instrument. All 105 targets have millimetre ALMA data available, which provideconstraints on disc masses and sizes. We identify a total of 13 multiple systems, including11 doubles and 2 triples. In agreement with previous studies, we find that the most massive(M dust >
50 M ⊕ ) and largest ( R dust >
70 au) discs are only seen around stars lacking visualcompanions (with separations of 20–4800 au) and that primaries tend to host more massivediscs than secondaries. However, as recently shown in a very similar study of >
200 PMS starsin the Ophiuchus Molecular Cloud, the distribution of disc masses and sizes are similar forsingle and multiple systems for M dust < M ⊕ and radii R dust <
70 au. Such discs correspondto ∼ >
20 au mostly affect discs inthe upper 10 % of the disc mass and size distributions. Key words:
Instrumentation: Adaptive optics, Instrumentation: interferometers, Binaries:visual, Protoplanetary discs, Methods: statistical, Planets and satellites: physical evolution
Recent Atacama Large Millimetre/sub-millimetre Array (ALMA)surveys in nearby ( d < pc) molecular clouds such as Lupusand Ophiuchus (Ansdell et al. 2016; Cieza et al. 2019) have shownthat protoplanetary discs exhibit a wide range of masses ( < ⊕ ) and sizes (r < (cid:63) E-mail: [email protected] (SED) class (Williams et al. 2019), stellar mass (Andrews et al.2013; Pascucci et al. 2016; Barenfeld et al. 2016), age (Ansdellet al. 2018; Ruíz-Rodríguez et al. 2018), and multiplicity (Harriset al. 2012; Cox et al. 2017; Zurlo et al. 2020). Disentangling theeffects of each variable is difficult and requires large samples andmulti-wavelength observations.In general, studies have found that discs in binary systemswith medium separations ( ∼ < © a r X i v : . [ a s t r o - ph . E P ] N ov A. Zurlo et al. wide companions (with separations >
200 au) have little impact ondisc properties (Harris et al. 2012; Cox et al. 2017). These effectsthat stellar companions have on disc properties can be understood interms of tidal truncation. In the case of a circumbinary disc, the inneredge location varies from 1.8–2.6 times the separation, dependingon the eccentricity of the orbit (e.g. Artymowicz & Lubow 1994).For a system with a stellar companion external to the circumstellardisc, truncation occurs at 0.3–0.5 times the physical separation ofthe binary companion (see, for example, Papaloizou & Pringle 1977;Artymowicz & Lubow 1994). A smaller disc also implies a shorterviscous dissipation timescale and previous infrared studies haveshown that discs around medium-separation binaries do in fact haveshorter lifetimes than those around single stars (Cieza et al. 2009;Kuruwita et al. 2018). In addition, Rosotti & Clarke (2018) foundthat secondary components of close binaries clear first due to theshorter viscous time scale associated with the smaller size of thedisc, while for wide binaries the difference in photoevaporation ratepermits a longer lifetime than tight systems.Nevertheless, the extent to which stellar multiplicity affectsdisc properties is not yet fully established. For instance, Cox et al.(2017) found significant differences in the disc sizes and massesseen around binary systems with respect to those of single stars.However, as noted by Zurlo et al. (2020), their sample was selectedbased on the presence of 70 µ m Spitzer excesses, which introducesstrong biases against small and faint discs. Since the vast majorityof discs in Ophiuchus are very small (r <
15 au, Cieza et al. 2019)and low-mass (M dust (cid:46) ⊕ , Williams et al. 2019), the effects thatvisual binaries (with separations >
20 au) have on the sizes andmasses of the general disc population are weak (Zurlo et al. 2020).In order to investigate disc properties as a function of multi-plicity, we present here a near-infrared (NIR) adaptive optics (AO)survey with VLT/NACO for a sample of 125 young stellar objects(YSOs) in the Lupus star-forming region, which have previouslybeen observed by ALMA at ∼ . (cid:48)(cid:48) ∼ . (cid:48)(cid:48) The aim of this NIR survey is to study the multiplicity of YSOs inthe Lupus star-forming region and the effects on circumstellar discproperties. The NIR-Lupus sample presented in this paper consistsof the YSOs more massive than brown dwarfs in the Lupus I-IV clouds that have been identified as either Class II or Class IIIsources. The complete sample includes 125 objects: 95 Class IIsources that were observed with ALMA and presented in Ansdellet al. (2016, 2018) and 30 Class III objects that were observed withALMA and presented in Lovell et al. (2020).Among the 125 objects in our NIR-Lupus sample, we foundNACO observations for 58 of them in the European Southern Ob-servatory (ESO) archive. A total of 47 objects have never been These truncation radii rely on co-planarity of the discs with the binaryorbit and could change if more complex scenarios are considered. observed in the NIR with AO and are bright enough to be observedwith NACO, with a K magnitude of 12.6 for the faintest target.These 47 objects were included in our NACO target list, along withthe stars in our sample with poor-quality archival data and all theknown binaries that we were allowed to re-observe with NACO.There are 20 objects in our sample that are too faint for NACOobservations; these objects are listed in Table A1 and are excludedfrom our statistical analysis.The literature covering multiplicity in Lupus is currently lim-ited. Some visual binaries were presented in Reipurth & Zinnecker(1993): Sz 65 (the companion to Sz 66), Sz 68, Sz 81, Sz 88,and Sz 108. Sz 65 was also identified as a binary by Merín et al.(2008), while Sz 69, presented in the same paper as a binary, ap-pears as a single star in our survey. The discs in the system ofSz 68 (also known as HT Lup) have been resolved by ALMA (aspart of DSHARP, Andrews et al. 2018; Kurtovic et al. 2018) andin polarized scattered light with SPHERE (as part of DARTTS-S,Garufi et al. 2020). Sz 74 was also identified as a binary system byWoitas et al. (2001) and Sz 123 is composed of two close M-typestar binaries as reported by Alcalá et al. (2014). To study the multiplicity in Lupus, we observed the YSOs in oursample with VLT/NACO, using methods similar to the alreadycompleted NIR-ODISEA survey of the Ophiuchus molecular cloud(Zurlo et al. 2020). We refer the reader to Zurlo et al. (2020) fordetails on the observing strategy. In short, we used very quick (0.16–0.17 s) L (cid:48) band filter exposures with the “star-hopping” technique tominimize overheads and maximize the number of observed objects.We took 2 exposures for each target with the jittering technique torecord a quasi-simultaneous sky background. As for NIR-ODISEA,we chose the AO configuration N90C10 (where 90% of the star lightgoes to the AO system NAOS and the remaining 10% goes to thedetector). The observations were carried out in visitor mode duringthe nights of 6, 20, 22, and 23 of August 2019 (for a total of 1.2observing nights) under ESO program 0103.C-0466 (PI: Zurlo).The observing conditions were generally favourable, with seeingfrom 0 . (cid:48)(cid:48) . (cid:48)(cid:48) . (cid:48)(cid:48) ∼ > σ signals using the Daophot algorithm via the photutils package in the astropy library (Astropy Collaborationet al. 2018). All of these automatic detections are inspected visuallyfor confirmation in the individual images (in very few cases someartifacts were selected by the algorithm). The minimum separationprobed by our observations is related to the AO diffraction limit ofthe instrument in L (cid:48) band. This is 0 . (cid:48)(cid:48)
1, i.e. ∼
18 au at 150 pc. Binarieswith separations closer than 20 au would not be resolved in this
MNRAS , 1–11 (2020) tellar multiplicity in Lupus. study. A conservative error of half a pixel is given on the measuredseparation of the companions, which corresponds approximately tothe width of the point spread function (PSF) divided by the minimumsignal-to-noise ratio (SNR) of our observations. Note that an offsetof up to 2 degrees on the North orientation may be present, asreported in the NACO manual, version P103.The detected multiple systems are shown in Figures 1 and2. The individual values of all the contrast curves are shown inFigure 3. Contrast limits are calculated from the azimuthal medianof the rms noise, following Zurlo et al. (2014). Two triple systemsand 11 binaries are detected in the final NACO images. One of thetwo triple systems is Sz 123, reported as close binary by Alcaláet al. (2014), which shows a faint tertiary in the NW. This detectionhas the lowest local SNR of ∼
6. The properties of the companionsources are listed in Table 1.
We found 58 objects in our Lupus sample in the ESO archive, allobserved with NACO but through different observing programs us-ing different techniques and filters. We reduced all the raw archivalimages with the same technique described in Section 3.1 and in-cluded any objects with low-quality or inconclusive data in ourNACO target list for an additional epoch of observation. Four sys-tems are multiple: Sz 68 (triple system, observed under program073.C-0379), Sz 81 (095.C-0610), RXJ16038 (085.C-0012), andIRAS 16051-3820 (097.C-0572). All of them but Sz 68 were ob-served in K s band with the jittering technique; Sz 68 was observedwith the NB_1.64 filter. The final images of the archival data areshown in Figure 1 with a different color map to distinguish themfrom our new NACO observations. Contrast curves are shown inFigure 3. Two new binaries are found, while Sz 68 is discovered tobe a triple system, not a binary as presented in Reipurth & Zinnecker(1993). Note that all the multiple systems reported in the literaturewere re-observed in this survey, or reduced from archival data. Theproperties of the systems are listed in Table 1. The ALMA data for the Class II sources in our Lupus sample weretaken during Cycle 3 in Band 6 under program 2015.1.00222.S (PI:Williams). The complete ALMA dataset of all 95 Class II sources inLupus was presented in Ansdell et al. (2018). The median 1.3 mmcontinuum rms was 0.10 mJy beam − and the typical beam sizewas . (cid:48)(cid:48) ( ∼
40 au at 150 pc). Only 24 of the Class II sources wereundetected by ALMA, and these typically had spectral types of M3or later. The candidate secondary sources reported in Ansdell et al.(2018) but detected only in the ALMA data are not considered asmultiple systems in this analysis, as they do not have NIR counter-parts. These systems are: Sz 88, which presents a marginally de-tected elongated PSF in ALMA, J16070384-3911113, which is toofaint to be observed with NACO, and J16073773-3921388, whichappears as a single star in the NACO image. The properties of thesesystems as reported by Ansdell et al. (2018) are listed in Table 2.The ALMA data for the Class III objects in our Lupus samplewere taken during Cycle 8 in Band 7 under program 2018.1.00437.S(PI: Wyatt) and are presented in Lovell et al. (2020). This Class IIIALMA survey had a moderate resolution of ∼ . (cid:48)(cid:48) ∼ µ Jy at 860 µ m, but detected only 4systems. None of the NIR binaries reported in this work are detectedin the ALMA data. We identify a total of 16 multiple systems: 3 triple systems and 13binaries. Three of these detected companions (Sz 108B, Sz 116 or2MASS J16094258-3919407B, and Sz 123C) are at different dis-tances than their primaries, thus we consider them not physicallybound (see below and Table 3). Some of these systems were alreadyknown as multiples from the literature (see Section 2). The NIR im-ages, over-plotted with the ALMA contours, are shown in Figures 1and 2. Four systems are resolved with ALMA, two close binariesare unresolved with ALMA, and 9 systems do not show millime-tre flux in at least one of the components. Each multiple system islabelled as resolved (Res), unresolved (UN), not detected (ND), ornot bound (N.B.) in Table 1. A summary of these 3 categories ofmultiple system is presented in Table 4, together with the numberof single stars. For each multiple system we measure the separationand the position angle of the companions, along with the ratio ofNIR flux. The millimetre flux ratio is also reported from Ansdellet al. (2018). Most of the systems have
Gaia
Data Release 2 (DR2)parallax measurements (Bailer-Jones et al. 2018), which we reportin Table 1 with all the other information mentioned above for theprimaries. Systems for which
Gaia detected the secondary com-ponent are listed in Table 3; the systems that present discrepanciesin the distances of the components are assumed as not physicallybound.For the other companion candidates that remain unconfirmedas bound, we estimate their probability of being background objects.We first consider a homogeneous spatial Poisson point process tocalculate the probability P ( n = | λ, B ) of not having any backgroundobject that is of equal or greater brightness to that of the companioncandidate in an area B for a number density λ . For each companioncandidate, we set B to the area of a circle with a radius equalto the candidate separation, and estimate the number density ofbackground stars that are of equal or greater brightness in that areaof the sky using the TRILEGAL model of the Galaxy (Girardiet al. 2012). We then consider the probability of having at leastone background star by chance P ( n > | λ, B ) = 1 - P ( n = | λ, B ) tobe a reasonable proxy for the probability of the candidate being abackground object, as in Reggiani et al. (2018) and Ubeira-Gabelliniet al. (2020). The calculated probabilities are reported in the lastcolumn of Table 1. For all stellar candidates, the probabilities areapproximately null owing to their large fluxes.Sz 88 presents an elongated shape in the millimetre, which maybe attributed to an edge-on disc or a close companion. The closemillimetre blob is only marginally detected in the ALMA data atthe ∼ σ level, and no signal is detected at the same position in theNIR. The nature of the close millimetre blob should therefore be fur-ther investigated with deeper high-contrast and/or high-resolutionobservations. Sz 123AB is composed of two close M-type stars(Alcalá et al. 2014) and we detect a third faint object in the NorthWest. This faint companion is detected by Gaia at a distance of1497.0 ± Jup . The candidate planet is located at ∼ ∼ MNRAS , 1–11 (2020)
A. Zurlo et al. -1.00.01.02.03.04.05.06.07.0-3.0-2.0-1.00.01.02.03.04.0
Sz 65 -0.50.00.5-0.50.00.5
Zoom -0.50.00.51.01.5-1.0-0.5 -1.0-0.50.00.51.0-1.0-0.5
Sz 74 -1.00.01.00.0
Sz 81 -0.50.0-0.50.00.5
RXJ16038
Sz 88 -4.0-3.0-2.0-1.00.01.02.0-1.00.01.02.03.04.05.0
IRAS 16051-3820 -1.00.01.02.0-1.00.01.0
V856 Sco -2.0-1.00.01.02.03.0-1.00.01.02.03.04.05.0
Sz 108 -0.50.00.5-0.50.00.51.0
IRAS 16054-3857
Figure 1.
A gallery including all the detected multiple systems of the Lupus sample. For each object, the NACO image is shown with a logarithmic colourstretch. Images shown with a distinct green colour map correspond to archival data. When detected, the millimetre ALMA counterpart 1.3 mm emission isshown in white contour levels, ranging from 5 times the RMS noise (normally the RMS noise is 0.15-0.2 mJy in each map) to 90% the peak emission. None ofthe Class III objects are detected in the mm. The ALMA synthesized beam is shown in the bottom left corner. North is up, East is left, scale in arcsec.MNRAS , 1–11 (2020) tellar multiplicity in Lupus. -1.00.01.02.0-1.00.01.02.0 -10.0-9.0-8.0-7.0-6.0-5.0-4.0-3.0-2.0-1.00.01.0-6.0-5.0-4.0-3.0-2.0-1.00.01.02.03.04.05.0 -2.0-1.00.01.0-1.00.01.02.0 Sz 123 -0.50.00.5-0.5
V* V1097 Sco
Figure 2.
Same as Figure 1.
Table 1.
Properties of the companions detected by the NIR-Lupus survey. Note that the RA and DEC coordinates are reported for the primary star. Objectsmarked with a (cid:72) are in the planetary mass regime, in the range 8-12 M
Jup . The distance marked with ∗ is not derived from Gaia
DR2 (Bailer-Jones et al. 2018)as the others, but assumed to be 150 pc, the mean distance for the Lupus star forming region. The uncertainty on the separation is 2 au for all the targets, andon the angle is 2 degrees, as explained in Section 3.1.Name RA DELoC Class dist Sep PA NIR mm Cat. a P bkg b New or(pc) (au) (deg) f f f f (%) LiteratureSz65B (Sz66) 234.8656 -34.7715 II 155 983 97 0.46 0.21 Res 0.1 L2MASS J15430624-3920194B 235.7760 -39.3387 III 165 24 -16 0.99 - ND < − NSz68B c c < − NSz74B 237.0217 -35.2648 II 150 ∗
49 -5 0.33 0.03 UN < − LSz81B d < − LRXJ16038B d < − NSz88B 241.7524 -39.0389 II 158 245 34 0.1 - ND 0.1 LIRAS 16051-3820B d < − NSz108B 242.1786 -39.1042 II 169 683 23 0.31 - N.B. e < − N2MASS J16094258-3919407B 242.4274 -39.3280 III 147 245 26 0.6 - N.B. e < − N2MASS J16095628-3859518B 242.4845 -38.9978 II 157 90 12 0.93 - ND < − N (cid:72) (cid:72) Sz123C 242.7149 -38.8873 II 162 292 -53 0.01 - N.B. e < − N a RES: resolved, UN: unresolved, ND: non-detected discs with ALMA. See Table 4 and Section 4 for more details. b Probability of being a background object. We indicate known bound companions with a dash sign. c Archive NB_1.64 filter data d Archive K-band data e Not bound, see Table 3.MNRAS000
49 -5 0.33 0.03 UN < − LSz81B d < − LRXJ16038B d < − NSz88B 241.7524 -39.0389 II 158 245 34 0.1 - ND 0.1 LIRAS 16051-3820B d < − NSz108B 242.1786 -39.1042 II 169 683 23 0.31 - N.B. e < − N2MASS J16094258-3919407B 242.4274 -39.3280 III 147 245 26 0.6 - N.B. e < − N2MASS J16095628-3859518B 242.4845 -38.9978 II 157 90 12 0.93 - ND < − N (cid:72) (cid:72) Sz123C 242.7149 -38.8873 II 162 292 -53 0.01 - N.B. e < − N a RES: resolved, UN: unresolved, ND: non-detected discs with ALMA. See Table 4 and Section 4 for more details. b Probability of being a background object. We indicate known bound companions with a dash sign. c Archive NB_1.64 filter data d Archive K-band data e Not bound, see Table 3.MNRAS000 , 1–11 (2020)
A. Zurlo et al.
Table 2.
Properties of the companions in Lupus detected by ALMA but not observed or detected in the NIR. These candidate secondaries were assumed ascontaminants in our analysis. Name RA DEC Sep (au) PA (deg) mm flux ratioSz88B? 241.7524 -39.0389 54 213 -J16070384-3911113B? 241.7649 -39.1867 568 268 0.39J16073773-3921388B? 241.9065 -39.3609 344 267 0.87 * c o n t r a s t c u r v e IRAS 16051-3820J16095628-3859518J16094258-3919407Sz 65V856 ScoSz 68J15392828-3446180Sz 74 Sz 108IRAS 16054-3857Sz 123J16130240-4004329J15464121-3618472J15430624-3920194RXJ16038Sz 88
Figure 3.
Contrast curves extracted from images shown in Figure 1 andFigure 2, calculated from azimuthally-averaged profiles centered at the pri-mary’s location. The curves flatten at distances greater than 5 (cid:48)(cid:48) . We notethat the structures displayed by some curves beyond 2 (cid:48)(cid:48) are due to detectorartifacts that affect the averaged profile, or due to reaching the edge of thedetector. Labels are provided for all curves.
Table 3.
Gaia distances for the detected systems. Objects assumed as notphysically bound are marked with a *.Name
Gaia distance (pc)Sz 65 155.3 ± ± ± ± ± ± ± ± ± ± ± ± ± ± As for NIR-ODISEA for the Ophiuchus molecular cloud, wecompare the masses of the Lupus discs in the different multiplic-ity categories. The dust mass for each disc is taken from the lastcolumn of Table 1 in Ansdell et al. (2018) for the Class II objects,and from Table 1 of Lovell et al. (2020) for the Class III ob-jects. Only four companions are detected in the millimetre: Sz 68,Sz 81A, Sz 65, and V856 Sco. The values of their masses are ob-tained from the millimetre fluxes listed in Table 4 of Ansdell et al.(2018), according to the same procedure described by the authors Single objectsResolved primariesResolved companionsUnresolved binariesND of component(s)
Figure 4.
Histogram of the total mass of the dust of the discs measured inthe ALMA data. for the primary components. Figure 4 shows a histogram of the discmasses in Lupus, color-coded by multiplicity category. In Figure 5,we show cumulative distribution functions of disc mass, accountingfor non-detections using the python package lifelines (Davidson-Pilon 2019), a Kaplan-Meier (KM) product estimator, as in Zurloet al. (2020); for clarity, we present separately the distributions forthe dust mass per system (left) and per component (right). We alsocompare the dust mass cumulative distributions for the primary vscompanion components (Figure 6) and with unresolved and unde-tected components (Figure 7). The cumulative distribution of thedisc sizes for single and multiple objects, expressed in semi-majoraxis of the disc, is displayed in Figure 8. The disc size for the ClassII objects is used when available in Ansdell et al. (2018) (Table 3),otherwise we assume the disc is unresolved. Concerning the ClassIII objects, 3 discs were unresolved, and only one was marginallyresolved (0 . (cid:48)(cid:48)
7, Lovell et al. 2020). The dust radii are measured fromthe 1.33 mm continuum images using a curve-of-growth method,in which successively larger photometric elliptical apertures are ap-plied until the measured flux is 90% of the total flux. For the detectedmultiples we show, respectively, the total mass (Figure 9, left panel)and the semi-major axis (Figure 9, right panel) of the primary discvs the separation of the companion, to display any dependence ofthe mass and size of the discs with the separation of the componentsin the system. Finally, in Figure 10 we show the distribution of thespectral type of the primary stars in single and multiple systems.
Even though the NIR-Lupus sample (105 objects, 20 discs are notobservable with NACO) is smaller than that of Ophiuchus (236objects, presented in Zurlo et al. 2020), the same trends are found.
MNRAS , 1–11 (2020) tellar multiplicity in Lupus. Table 4.
Summary of the four categories in which the population of NIR-Lupus is divided. Note that there are 2 triple systems, so the total number of systemsare 105. Category N. of objects DescriptionSingle objects 92 Single class II and class III objects in LupusResolved discs (Res) 4 Resolved disc companions in the mmUnresolved multiple discs (UN) 2 Binary systems where the components are unresolved in the mmND of component(s) (ND) 9 Only the primary is detected in ALMA, or none of the components is detected C u m u l a t i v e d i s t r i bu t i on Single objectsMultiple systems 0 50 100 150 200Average dust mass per disc (M )0.00.20.40.60.81.0 C u m u l a t i v e d i s t r i bu t i on Single objectsMultiple systems
Figure 5.
Cumulative distribution of the total mass of the dust ( left ) and of the average mass of the dust around each star ( right ) in the systems as measured inthe ALMA data, single vs multiple systems are shown. C u m u l a t i v e d i s t r i bu t i on Primary starsCompanions
Figure 6.
Cumulative distribution of the total mass of the dust in the pri-maries vs the companions as measured in the ALMA data.
The fraction of visual multiple systems (with separations of 20–4800 au) are similar in both samples, with 12 % (13/105) in Lupusand 18 % (43/236) in Ophiuchus. These fractions are significantlylower than in other star-forming regions where the multiplicitycensuses are more complete. The Taurus region, for example, showsa multiplicity frequency of ∼ C u m u l a t i v e d i s t r i bu t i on Unresolved binariesND of component(s)
Figure 7.
Cumulative distribution of the total mass of the dust in the systemswhere the tight binaries are not resolved by ALMA vs the ones where thebinaries are resolvable in the mm, but only the primary are detected. to be homogeneously distributed by spectral type, as shown in Fig-ure 10, but the fraction of multiple systems is lower for late-typeobjects, as expected. In Figure 11 we show the histograms that in-clude also the Ophiuchus sample, which follows the same trend.Stellar multiplicity surveys show that the frequency of multiple sys-tems strongly depends on the mass of the primary (e.g., Duquennoy& Mayor 1991; Fischer & Marcy 1992; Burgasser et al. 2003). In-deed, solar-type stars have a high probability of having companions( ∼ MNRAS000
Cumulative distribution of the total mass of the dust in the systemswhere the tight binaries are not resolved by ALMA vs the ones where thebinaries are resolvable in the mm, but only the primary are detected. to be homogeneously distributed by spectral type, as shown in Fig-ure 10, but the fraction of multiple systems is lower for late-typeobjects, as expected. In Figure 11 we show the histograms that in-clude also the Ophiuchus sample, which follows the same trend.Stellar multiplicity surveys show that the frequency of multiple sys-tems strongly depends on the mass of the primary (e.g., Duquennoy& Mayor 1991; Fischer & Marcy 1992; Burgasser et al. 2003). In-deed, solar-type stars have a high probability of having companions( ∼ MNRAS000 , 1–11 (2020)
A. Zurlo et al. C u m u l a t i v e d i s t r i bu t i on Single objectsMultiple systems
Figure 8.
Cumulative distribution of the projected semi-major axis of thediscs as measured in the ALMA data. spectral types) show a significantly lower fraction of 8% (Duchêne& Kraus 2013).The distributions of dust disc masses for single and multiplesystems are very similar up to M dust (cid:39)
50 M ⊕ (Figure 5). This resultis the same if we include the Ophiuchus discs, as shown in Figure 12(left panel). A Kolmogorov-Smirnov test comparing the distribu-tions of singles versus multiples gives a p-value of 40% for Lupusand 51% for Lupus and Ophiuchus combined. These high p-valuessuggest that the distributions are nearly identical. The correspondingcumulative probability distribution for each category is displayedin Figure 12 (right panel). The larger sample size strengthens thestatistical significance. Discs around single stars may have massivediscs up to M dust (cid:39)
200 M ⊕ , as derived from the same cumulativedistribution. Note that the most massive discs of both Lupus andOphiuchus have M dust (cid:39)
200 M ⊕ .Only 4 multiple systems in Lupus have all the componentsdetected also by ALMA. For these systems, the discs around theprimaries are more massive than the discs around the secondaries(Figure 6). This result is confirmed in the larger combined sampleof Lupus and Ophiuchus: in Figure 13 we show the cumulativedistributions of the disc masses for primaries and companions. Discsaround the primary stars are indeed systematically more massivethan the ones around the secondary stars. This result is likely to bedue to the combination of two factors: the strong dependence ofdisc mass on stellar mass (e.g., Andrews et al. 2013) and the factthat secondaries are expected to have smaller truncation radii. Thisresult is also in agreement with the predictions of Rosotti & Clarke(2018) that discs around secondary stars have shorter lifetimes thandiscs around primaries.As for the mass, the disc size is also influenced by the effects ofmultiplicity: the semi-major axis is smaller around discs in multiplesystems, as seen in Taurus (Manara et al. 2019). From the completesamples of all the discs of Lupus and Ophiuchus we can constrain themaximum size for multiple-system discs to be ∼
80 au, while discsaround single stars can be more than 3 times larger. The semi-majoraxis cumulative distribution is presented in Figure 14.In Figure 15 (left panel), we show total disc dust mass as afunction of semi-major axis (defined as the radius containing 90 % ofthe flux) for single and multiple systems. The figure shows, not onlythat single systems span a larger range in disc sizes and masses, but also that singles have a wider range of masses for a given disc size.This finding suggests that disc around binary systems might havedifferent surface density profiles. Deriving accurate surface densityprofiles requires radiative transfer modeling, which is outside thescope of this paper. However, at 1.3 mm the brightness profile iswell correlated with the distribution of mass as a function of radius.In Figure 15 (right panel), we show the normalized brightnessprofiles for discs in single and multiple systems, deprojected usingthe inclination and position angle from imfit (from CASA, Mc-Mullin et al. 2007), for individual sources in Ophiuchus and Lupus.Using the data from both star-forming regions, we obtain the radialprofiles for sources where the peak SNR is > × rms and the imfitgaussian coincides with the signal peak, thereby excluding transi-tion discs as well as poorly resolved sources. The figure includes147 radial profiles between single and binary stars for both samples.Discs in multiple systems are clearly more compact than single stardiscs. The same finding is presented in Manara et al. (2019) for theTaurus region.Even if the influence of multiplicity is evident on the mass andsize of the protoplanetary discs, Desidera & Barbieri (2007) foundthat the properties of exoplanets around single and wide binaries aresimilar, with the exception of a possible effect on the eccentricity ofthe planet orbits. For tight binary systems ( <
40 au) the disc life timeis expected to be so short that planets would have to form within1 Myr (Kraus et al. 2012; Cheetham et al. 2015). The fraction ofbinary systems in Lupus and Ophiuchus is only 4% from the NIR-Lupus and NIR-ODISEA surveys, even if they are not sensitive tovery short-period binaries due to the NACO spatial resolution (0 . (cid:48)(cid:48) ∼
20 au projected separation at 150 pc.
We have conducted a NIR AO imaging survey of all the Class II andIII objects of the Lupus star-forming region. We complemented ourVLT/NACO L (cid:48) band observations of 47 stars with archival data tocover almost all the 125 circumstellar discs of Lupus. Only 20 faintobjects were not included in this analysis. Our study is sensitive tovisual binaries with projected separations between 20–4800 au —spectroscopic, tight, and large separation binaries are not includedin our statistics. In total we report the detection of 11 potentiallybound binary systems and 2 triple systems, where the primarieshave different spectral type spanning from A to L. The frequencyof visual companion multiplicity in Lupus in the range 20-4800au is then 12% (13/105). Among the companions, one object isin the planetary regime (12 M
Jup ), but we cannot exclude it as abackground contaminant with the available data.To strengthen the statistical analysis, we combined this Lupussurvey with its twin in Ophiuchus (Zurlo et al. 2020). The resultsthat were obtained with the combined 341 (105+236) PMS starsare listed below. The following statistics are applicable to visualcompanions in the range of projected separations ∼ > ∼ > dust = 50 M ⊕ . Higher massdiscs (up to ∼
200 M ⊕ ) are only found around single stars.(iii) Discs around primary stars are systematically more massivethan the ones around their companions. MNRAS , 1–11 (2020) tellar multiplicity in Lupus. Separation (au)010203040506070 T o t a l du s t m a ss ( M ) Resolved binariesUnresolved binariesND of component(s) Separation (au)01020304050607080 S e m i - m a j o r a x i s ( au ) Resolved binariesUnresolved binariesND of component(s)
Figure 9.
Total mass of the system in dust ( left ) and semi-major axis of the discs as measured in the ALMA data ( right ) versus projected separation in betweeneach component of the multiple system. Unresolved discs are indicated with arrows.
A7 K2 K5 K6 K7 M0 M1 M2 M3 M4 M5 M605101520 SingleMultiple
Figure 10.
Histogram of the spectral type of the single stars vs multiplesystems. (iv) Discs around single stars can have semi-major axes up to300 au, while discs around multiple systems are 3 times smaller.(v) Discs around stars in multiple systems have flux distributionsthat tend to be much more compact than those seen for single stars.This suggests significant differences in their surface density profiles.(vi) The visual binaries identified in our AO IR survey haveseparations >
20 au and mostly affect discs properties in the 10 % upper end of the mass and size distributions. The incidence of tightvisual binaries (with separations of 10-40 au) that are likely toprevent the formation of planets is rather low (4 %), possibly due tothe limited angular resolution of our survey. ACKNOWLEDGEMENTS
The authors are grateful to the anonymous referee for theircomments that improved the quality of the manuscript. A.Z.acknowledges support from the FONDECYT
Iniciación en inves-
A G K M L10 SingleMultiple
Figure 11.
Histogram of the spectral type of the single stars vs multiplesystems of all the primary stars of Lupus and Ophiuchus. The sample isdominated by M-type stars, but multiple systems have been identified acrossthe entire spectral range. tigación project number 11190837. L.C. acknowledges supportfrom FONDECYT Regular number 1171246. V.C. is gratefulfor Australian Research Council funding via DP180104235. S.P.acknowledges support from FONDECYT grant number 1191934and the Joint Committee of ESO and the Government of Chile.J.P.W. acknowledges support from the National Science Foundationunder grant AST-1907486. R.C. acknowledges financial supportfrom CONICYT Doctorado Nacional N ◦ MNRAS000
Histogram of the spectral type of the single stars vs multiplesystems of all the primary stars of Lupus and Ophiuchus. The sample isdominated by M-type stars, but multiple systems have been identified acrossthe entire spectral range. tigación project number 11190837. L.C. acknowledges supportfrom FONDECYT Regular number 1171246. V.C. is gratefulfor Australian Research Council funding via DP180104235. S.P.acknowledges support from FONDECYT grant number 1191934and the Joint Committee of ESO and the Government of Chile.J.P.W. acknowledges support from the National Science Foundationunder grant AST-1907486. R.C. acknowledges financial supportfrom CONICYT Doctorado Nacional N ◦ MNRAS000 , 1–11 (2020) A. Zurlo et al. C u m u l a t i v e d i s t r i bu t i on Single (Lup+Oph)Multiple (Lup+Oph) 0 50 100 150 200Total dust mass (M )0.00.20.40.60.81.0 C u m u l a t i v e d i s t r i bu t i on Single objectsResolved A starResolved B starUnresolved binariesND of component(s)
Figure 12.
Left : Cumulative distribution of the average mass of the dust around each star in the systems for the two combined samples of Lupus and Ophiuchus,single vs multiple systems are shown. The distributions are very similar up to ∼
50 M ⊕ , but discs around single stars can reach dust masses as large as 200 M ⊕ . Right : Cumulative distribution of the total mass of the dust for the different categories of systems C u m u l a t i v e d i s t r i bu t i on Primary (Lup+Oph)Companions (Lup+Oph)
Figure 13.
Cumulative distribution of the total mass of the dust in theprimaries vs the companions as measured for both Lupus and Ophiuchussamples. Primary stars have more massive discs than their secondaries dueto the strong dependence of disc mass on stellar mass and the fact that theirdiscs have larger truncation radii.
DATA AVAILABILITY
The data underlying this article are available in the ESO archive athttps://archive.eso.org/eso/eso_archive_main.html, and can be ac-cessed with program numbers 073.C-0379, 085.C-0012, 095.C-0610, 097.C-0572, 0103.C-0466.
REFERENCES
Alcalá J. M., et al., 2014, A&A, 561, A2 C u m u l a t i v e d i s t r i bu t i on Single (Lup+Oph)Multiple (Lup+Oph)
Figure 14.
Cumulative distribution of the projected semi-major axis of thediscs in Lupus and Ophiuchus. The distributions are very similar up to ∼ , 1–11 (2020) tellar multiplicity in Lupus. Semi-major axis (au)050100150200250 T o t a l du s t m a ss ( M ) Single (Lup+Oph)Multiple (Lup+Oph) N o r m a li s ed f l u x Single (Lup+Oph)Multiple (Lup+Oph)
Figure 15.
Left : Dust mass vs semi-major axis of the discs of Lupus and Ophiuchus. Discs around single stars span a larger range in disc sizes and masses.
Right : Deprojected radial profiles of normalized brightness for discs with peak
S NR > in Ophiuchus and Lupus. Discs in multiple systems tend to be muchmore compact than those around single stars.Davidson-Pilon C., 2019, The Journal of Open Source Software, 4, 1317Desidera S., Barbieri M., 2007, A&A, 462, 345Duchêne G., Kraus A., 2013, ARA&A, 51, 269Duquennoy A., Mayor M., 1991, A&A, 500, 337Fischer D. A., Marcy G. W., 1992, ApJ, 396, 178Garufi A., et al., 2020, A&A, 633, A82Girardi L., et al., 2012, Astrophysics and Space Science Proceedings, 26,165Harris R. J., Andrews S. M., Wilner D. J., Kraus A. L., 2012, ApJ, 751, 115Kraus A. L., Ireland M. J., Martinache F., Hillenbrand L. A., 2011, ApJ,731, 8Kraus A. L., Ireland M. J., Hillenbrand L. A., Martinache F., 2012, ApJ,745, 19Kurtovic N. T., et al., 2018, ApJ, 869, L44Kuruwita R. L., Ireland M., Rizzuto A., Bento J., Federrath C., 2018, MN-RAS, 480, 5099Lovell J. B., et al., 2020, MNRAS,Manara C. F., et al., 2019, A&A, 628, A95McMullin J. P., Waters B., Schiebel D., Young W., Golap K., 2007, inShaw R. A., Hill F., Bell D. J., eds, Astronomical Society of the PacificConference Series Vol. 376, Astronomical Data Analysis Software andSystems XVI. p. 127Merín B., et al., 2008, The Astrophysical Journal Supplement Series, 177,551Papaloizou J., Pringle J. E., 1977, MNRAS, 181, 441Pascucci I., et al., 2016, ApJ, 831, 125Raghavan D., et al., 2010, ApJS, 190, 1Reggiani M., et al., 2018, A&A, 611, A74Reipurth B., Zinnecker H., 1993, A&A, 278, 81Rosotti G. P., Clarke C. J., 2018, MNRAS, 473, 5630Ruíz-Rodríguez D., et al., 2018, MNRAS, 478, 3674Ubeira-Gabellini M. G., Christiaens V., Lodato G., Ancker M. v. d., FedeleD., Manara C. F., Price D. J., 2020, ApJ, 890, L8Williams J. P., Cieza L., Hales A., Ansdell M., Ruiz-Rodriguez D., CasassusS., Perez S., Zurlo A., 2019, ApJ, 875, L9Woitas J., Köhler, R. Leinert, Ch. 2001, A&A, 369, 249Zurlo A., et al., 2014, A&A, 572, A85Zurlo A., et al., 2020, Monthly Notices of the Royal Astronomical Society,496, 5089 Table A1.
List of the not observable Lupus objects in the NIR with NACO,then excluded in the statistical analysis.Name RA DECSSTc2dJ154301.3-340915 15:43:01.29 -34:09:15.40SSTc2dJ154302.3-344406 15:43:02.29 -34:44:06.20SSTc2dJ160115.6-415235 16:01:15.55 -41:52:35.30SSTc2dJ160703.9-391112 16:07:03.85 -39:11:11.60SSTc2dJ160714.0-385238 16:07:14.00 -38:52:37.90SSTc2dJ160754.7-391545 16:07:54.75 -39:15:44.60SSTc2dJ160801.7-391231 16:08:01.75 -39:12:31.60SSTc2dJ160815.0-385715 16:08:14.96 -38:57:14.50SSTc2dJ160828.1-391310 16:08:28.10 -39:13:10.00SSTc2dJ160831.1-385600 16:08:31.10 -38:56:00.00SSTc2dJ160851.4-390530 16:08:51.43 -39:05:30.40SSTc2dJ160858.3-390749 16:08:58.30 -39:07:49.40SSTc2dJ160916.4-390444 16:09:16.43 -39:04:43.70SSTc2dJ160920.3-390402 16:09:20.30 -39:04:01.60SSTc2dJ160923.2-390407 16:09:23.15 -39:04:07.40SSTc2dJ160934.2-391513 16:09:34.18 -39:15:12.70SSTc2dJ160939.3-390432 16:09:39.29 -39:04:31.80SSTc2dJ161013.1-384617 16:10:13.06 -38:46:16.80SSTc2dJ161027.4-390230 16:10:27.43 -39:02:30.20SSTc2dJ161204.5-380959 16:12:04.48 -38:09:59.00
APPENDIX A: NOT OBSERVABLE LUPUS OBJECTS
This paper has been typeset from a TEX/L A TEX file prepared by the author.MNRAS000