The results of binary surveys among T Tauri stars in the T associations Taurus-Auriga and Chamaeleon, and in the OB association Scorpius-Centaurus are summarized, and implications on our understanding on the formation of binary and multiple systems are discussed.

We present the results of a high-angular resolution spectro-imaging survey of a sample of isolated Herbig Ae-Be (HAeBe) stars aimed at searching for close companions. The fraction of wide binaries is found to be significantly higher in HAeBe stars than in solar-mass field dwarfs, and suggests a companion star fraction possibly larger than 1. The spectral energy distribution of the primaries and secondaries is derived independently from multi-wavelength resolved photometry of the systems. It is found that young low-mass companions usually do not exhibit infrared excesses, which suggests that the lifetime of their circumstellar disk might have been shortened by the influence of the massive primary.

We give here an overview of the current state of our survey for pre-main sequence spectroscopic binaries. Up to now we have taken 739 spectra of 250 pre-main sequence stars. We find that 8% of the stars show significant radial velocity variations, and are thus most likely spectroscopic binaries. In addition to the targets showing radial velocity variations, 6% of the targets are double-lined spectroscopic binaries i.e., the total fraction of spectroscopic binaries is expected to be about 14%. All short-period SB2s are monitored photometrically in order to search for eclipses. An eclipsing SB2 would allow the direct measurement of the masses of both stellar components. Measurements of the stellar masses together with determinations of the stellar radii are a crucial test of evolutionary tracks of pre-main sequence stars.

We summarise the results of recent optical and near-infrared imaging studies of the binary fraction among young low-mass stars in the dense Orion Trapezium Cluster. Over the separation range ∼ 30–500 AU and within the observational errors, there appears to be no excess of binary systems in the cluster relative to the main sequence field star population. Over the separation range ∼ 1000–5000 AU, the cluster is deficient in binaries relative to the field. Both results are in contrast to those found for the more distributed population of young stars in the Taurus-Auriga dark clouds, which is overabundant in binaries by roughly a factor of two. We briefly discuss possible origins for this difference and observational tests which may distinguish between them, and the implications these results have for our understanding of the typical environment where most young stars are born.

With known properties such as age, distance and stellar density, young clusters provide ideal samples for binary star studies. We summarize recent results from high-resolution multiplicity surveys of IC 348, α Per, the Pleiades, the Hyades, and Praesepe. The statistics of the resolved companions are used to address binary star formation and evolution in clusters. Over the ranges of ages, densities and stellar masses covered by the clusters, it is found that the binary fraction is correlated with stellar density rather than age, and that both the binary fraction and binary mass ratio distribution depend upon mass. The impact of companions on X-ray emission and stellar rotation is also discussed.

We have analysed the binarity and multiplicity characteristics of 120 O-type stars in 22 very young open clusters and found marked differences between the “rich” (N ≥ 6 O-type stars and primaries) and “poor” (N = 1) clusters. In the rich clusters, the binary frequencies vary between 14% (1 SB among 7 stars) and 80% (8 SBs among 10 stars). Multiple systems seem not to be frequent and stars are spread all over the cluster area. In poor clusters, the binary frequency of the O-type objects is nearly 100%, with orbital periods around 3 days. Several binaries are also eclipsing. Additional companions are always present. They form either hierarchical multiple stars or trapezium systems. These massive multiple systems are generally found close to the cluster center, although there are exceptions.

The preponderance of binary systems in all known stellar populations makes them exciting dynamical agents for research on topics as varied as star formation, star-cluster dynamics and the interiors of young and old stars. Today we know that the Galactic-field binary population is probably a dynamically evolved version of the Taurus–Auriga pre-main sequence population, and that the initial distributions of binary-star orbital elements are probably universal. Furthermore, N-body calculations tentatively suggest that OB stars form in energetic binaries near cluster cores, and that binaries with ‘forbidden’ orbital elements that are produced in stellar encounters, may turn out to be very useful windows into stellar interiors, potentially allowing tests of pre-main sequence evolution theory as well as of models of main-sequence stars.

We review current observations of binary star populations with particular attention to what insight these populations can give us into the problem of how binary stars form. Significant progress has been made in the past few years, revealing variations as a function of site, primary star mass, and binary star separations. The variations in the binary star population with type of star formation site in comparison with the field, suggests that ∼30% of the field binaries formed in loose T associations and ∼70% formed in the dense progenitors of open clusters. Variations with mass and separation on the whole are well matched by the predictions of fragmentation followed by competitive accretion. However, there remains much work to be done on both the observational and theoretical end before a complete picture of binary star formation can be developed.

Current millimetre arrays are well suited to study the gas and dust distribution around young binary systems of separation ∼ 0.3–3″. In particular, they clearly demonstrate that disks around binary systems are tidally truncated. I review here the observed properties at mm wavelengths of the gas and dust around the best known multiple systems. I also discuss the limitations of current mm interferometers.

GG Tau is a textbook example of a binary system. The circumstellar material around GG Tau is divided in several distinct regions: 1) small, low mass, circumstellar disks, detected in the near-IR and mm domain, 2) a well defined ring, of inner radius 180 AU, detected in the mm domain and in scattered near-IR light, 3) a more extended, colder disk detected in the 13CO(2–1) and 13CO(1–0) lines. Recent observations of the 12CO(2–1) clearly show this extended disk, but also reveal a fourth component of the circumstellar material: (relatively) diffuse and hot gas in the tidally unstable region. Estimate of the gas content suggest this material may be feeding the inner disks at about 10−6 M⊙/yr.

We review recent imaging results on the circumstellar matter of young binaries obtained using the Hubble Space Telescope (HST) and by PUEO, a ground-based adaptive optics (AO) system operating at CFHT. In the area of circumbinary disks, new results for GG Tau help form a more complete picture of this prototype system. However, HST images of the UY Aur ring indicate a more complex system than first thought. A new example of a circumstellar disk in a multiple star system has been found in HV Tau, joining HK Tau in this category. Examples of envelopes, outflow cavities, and jets are shown for several young binaries. About half of the nearby young binaries in our survey possess some kind of local nebulosity at optical or near-IR wavelengths.

We present near-infrared images of the circumbinary disk surrounding the pre–main-sequence binary star, GG Tau A, obtained with NICMOS aboard the Hubble Space Telescope. These images have a SNR ∼25 times higher than previous ground-based measurements, which allows the spatially resolved disk to be explored in detail. The geometry of the ring implies that the circumbinary disk is not intrinsically circular, possibly due to interactions with the central binary star. Overall, the circumbinary disk is redder than the central binary, with the amount of red excess increasing with wavelength. Significant variations in color over the 13 arcsec2 covered by the disk are also observed, raising the possibility that disk inhomogeneities are present.

Analysis of a sample of giant Herbig-Haro flow sources shows that 79%–86% are binaries or higher order multiples. This represents the youngest sample of stars studied so far for binarity. A stellar dynamics jet hypothesis is proposed in which the dynamical decay of triple or multiple systems leads to giant outflow activity. Close triple approaches will cause serious perturbations and probably direct collisions among individual circumstellar disks, with a consequent burst of outflow activity, which can produce giant HH bow shocks. As one component is ejected, the two remaining stars and their small truncated disks form a closer bound pair with high eccentricity. Gas streams from a circumbinary disk feed the stars and this as well as other dynamical effects cause the binary orbit to shrink. As the stellar components gradually spiral towards each other, accretion and outflow becomes cyclic, modulated on an orbital time scale. The resulting HH flows can be read as a fossil record of the evolution of orbital motions of the newly formed binary as it shrinks from a typical separation of 100 AU or more to 10 AU or less. After a triple disintegration event, both components (star and close binary) leave their nascent envelope, and while one component becomes visible as a T Tauri star, the other will be obscured for a while by the envelope and will appear as a bright near-infrared object, thus explaining the socalled IRC binaries which are infrequently found in star forming regions.

We have carried out high-resolution near-infrared imaging observations of a protostar L1551 IRS 5 with the Subaru Telescope. The jet structure of IRS5 is resolved into two independent jets for the first time from the ground. Successive near-infrared spectroscopy has revealed that the jet emission is dominated by [Fe II] lines in the J and H-bands. While the visual-extinction reaches more than 20 mag in the close vicinity of IRS 5, it decreases rapidly at ∼1″ from IRS 5 and remains constant around 7 mag at larger distances. The twisted structure and bright emission knots are intrinsic to the jets, not due to a spatial variation of the extinction.

Infrared companions are young stellar objects with unusual properties gravitationally bound to more or less typical T Tauri stars. As such they promise to be the source of information on either a particular phase in the development of young stars or on a particular mode of development. We discuss the observed properties of infrared companions as well as attempts to explain their physical status with the aim to see how much of solid conclusion has been obtained so far.

We present echelle spectroastrometric observations of the Z CMa system in the Hα region showing its feasibility and underlining the astrophysical importance of the information stored in the spatial width of the spectra.

The Hα profile dating from 12/1991 presents a “bullet” of blue-shifted emission superimposed on the broad absorption component. We interpret this material as an unresolved bullet of gas at the base of the jet. We show that the jet emission fills in the absorption component of the Hα profile and has a wide range of velocities.

The application of the lunar occupation method to studies of young stars is discussed. The statistics of results on binarity and multiplicity is reviewed.

Observations at millimeter and submillimeter wavelengths are sensitive to the total mass of circumstellar and circumbinary dust in a multiple system, and in some cases single-dish observations can help constrain the location of disk material in spite of their lack of spatial resolution. Young binary stars show a great diversity of disk properties, with a large part of the variation accounted for by binary separation. Many of the closest binaries (those with separations a less than a few AU) harbor massive circumbinary disks. Binaries with a ≳ 100 AU tend to have massive circumstellar disks. In both cases, the properties of these disks (as deduced from millimeter and infrared fluxes) are indistinguishable from those around single stars. In the intermediate separation range (10 ≲ a ≲ 100), however, while disks do exist in most binaries, they are strongly limited by the presence of stellar companion, with inferred dust masses of order an Earth mass. While comparison of sample properties is secure, calculating masses in individual systems is limited by the uncertainty in dust opacity and surface density distribution laws (as in single stars), with the additional complication of the uncertain disk geometry in the system.

We present the results of a study performed with the goal to investigate whether low-mass pre-main sequence binary stars are formed by multiple fragmentation or via stellar capture. If binaries form preferentially by fragmentation, we expect their disks to be co-planar. On the other hand, the capture scenario will lead to a random distribution of disk orientations. We performed near-infrared polarization measurements of 49 young visual binary stars in the K band with SOFI at the NTT. The near-infrared excess radiation of the targets mostly point to the presence of disks. For a major fraction of the sample, evidence for disks is also obvious from other features (outflows, jets, Herbig-Haro objects). We derived the disk orientation from the orientation of the polarization vector of both components of each binary. This statistical study allows to test which hypothesis (co-planarity, random orientation) is consistent with the observed distribution of polarimetric position angles. We find evidence that the disks are preferentially aligned.

Photospheric Doppler images for both stellar components of the double-lined spectroscopic binary V824 Ara reveal surface temperature inhomogeneities of up to 1800 K on both stars. The absolute brightness and the mass of the two stars suggest that they are very close to the main sequence – but not yet on the ZAMS – if an inclination of the orbital plane (and rotational axes) of 52° is adopted as suggested by the Doppler imaging. Since both stars are active, we solve for the inclinations of the rotation axes of both stellar components separately and emphasize that it can be used to test coplanarity predictions from binary star-formation theories. We find that the spin inclinations for the individual components of V824 Ara agree to within their uncertainties.