Companions to sdB binaries−degenerate or non-degenerate?

We report the results of an on-going programme to collect lightcurves for EHB binaries with orbital periods P ? 1 d. Degenerate and non-degenerate companions are clearly distinguished by the presence or absence of a reflection effect in these binaries. The amplitude of the reflection effect can be combined with other data to provide an estimate of the mass of the companion star. We find that the fraction of sdB binaries in our sample which have non-degenerate companions is 0.08±0.06 (1?σ error). These non-degenerate companions have very low masses (? 0.1M⊙). This property is not predicted by existing population synthesis models.     

Companions of Post-Common Envelope sdB Binaries

Preliminary results are presented from two ongoing complementary surveys intended to investigate the nature and characteristics of the optically invisible secondaries in post-common envelope subdwarf B (sdB) binary stars. We obtain precise radial velocities to derive periods and minimum companion masses for bright field sdB stars. These data are combined with light curves to search for eclipses, reflection effects, or ellipsoidal variations. We emphasize the importance of using complete unbiased samples, without which it will not be possible to understand the details of the multiple processes that produce these stars. It remains true that all known secondary companions in short-period sdB binaries are nearly invisible, thus they must be either low mass main sequence (MS) stars or compact objects, e.g., white dwarfs. In our small, nearly-complete sample, white dwarf secondaries outnumber MS secondaries by about a factor of five. Known MS masses in short-period sdB binaries are all surprisingly low, indicating a possible bimodal mass distribution for all MS secondaries in sdB binaries.     

The HYPER-MUCHFUSS project—target selection and analysis

The HYPER-MUCHFUSS project targets a population of high velocity subluminous B stars to discover either close binaries with massive unseen companions or hyper-velocity stars. Our starting point is the enormous database of SDSS. We preselected sdO/B candidates by colour and classified them by visual inspection of their spectra. We measured the radial velocity from the coadded SDSS spectra, which serves as first epoch measurement. Stars with high Galactic rest-frame velocities were selected and second epoch observations were obtained starting in 2007 at several sites. For the brighter targets we also included the SDSS individual spectra as additional information. In the course of our survey we observed 88 out of 265 stars from our target list. We discovered 39 HVS candidates as well as 49 close binaries. In addition we analysed all single spectra of sdBs from SDSS and found 120 close binaries. For the targets with constant RVs we performed a proper motion analysis with the highest possible accuracy from the available digitised photographic plates. Together with the analysed spectra and the calculation of the spectroscopic distance, we calculated complete trajectories and deduced the origins of these stars. Targets with high RV variability on short timescales were selected for follow-up. Numerical simulations based on the period and companion mass distribution of the known sdB binary sample were carried out to optimise the target selection and single out candidate binaries with massive companions. The follow-up campaign using WHT/ISIS and CAHA-3.5m/TWIN started in 2009.     

Radial velocity curves of nine sdB binary stars

Radial velocity curves have been measured for nine sdB stars from high resolution optical spectra and are found to be sinusoidal indicating that they are binary stars with circular orbits. Their periods range from ≈12 h to more than 8 days. The companions are invisible in optical light. Minimum companion masses are derived from the mass functions, assuming the mass of the sdB primary to be half a solar mass. We argue that the companions to UVO 1735+22, Feige 108, HD 188112, and HD 171858 are white dwarfs, since their (statistically) most likely mass is above ≈0.7M _⊙. Five of the systems have not been investigated before, for the others our results agree with previously published ones.     

Using 2MASS to detect main sequence companions to subdwarf B stars

We explore the binary fraction of subdwarf B (sdB) stars by using the Two Micron All Sky Survey to search for main sequence companions. We have convolved Kurucz models with appropriate filter bandpasses to examine how various combinations separate single sdB stars from sdB+MS binaries. We notice that the 2MASS magnitude limits greatly increase the fraction of inferred sdB+MS binaries and examine a magnitude-limited sample appropriate for single sdB stars.     

Photometry of four binary subdwarf B stars and the nature of their unseen companion stars

We present light curves of four binary subdwarf B stars (sdB), Ton 245, Feige 11, PG 1432+159 and PG 1017−086. We also present new spectroscopic data for PG 1017−086 from which we derive its orbital period,   P =0.073 d  , and the mass function,   f _m=0.0010±0.0002 M_⊙.  This is the shortest period for an sdB binary measured to date. The values of P and f _m for the other sdB binaries have been published elsewhere. We are able to exclude the possibility that the unseen companion stars to Ton 245, Feige 11 and PG 1432+159 are main-sequence stars or subgiant stars from the absence of a sinusoidal signal, which would be caused by the irradiation of such a companion star, i.e. they show no reflection effect. The unseen companion stars in these binaries are likely to be white dwarf stars. In contrast, the reflection effect in PG 1017−086 is clearly seen. The lack of eclipses in this binary combined with other data suggests that the companion is a low-mass M-dwarf or, perhaps, a brown dwarf.     

Atmospheric Parameters and Abundances of sdB Stars

We summarize recent results of quantitative spectral analyses using NLTE and metal line-blanketed LTE model atmospheres. Temperatures and gravities derived for hundreds of sdB stars are now available and allow us to investigate systematic uncertainties of T _e ff, log g scales and to test the theory of stellar evolution and pulsations. Surface abundance patterns of about two dozen sdB stars are surprisingly homogenous. In particular the iron abundance is almost solar for most sdBs. We highlight one iron-deficient and three super metal-rich sdBs, a challenge to diffusion theory. sdB stars are slowly rotating stars unless they are in close binary systems, which is hard to understand if the sdB stars were formed in merger events. The only exception is the pulsator PG 1605+072 rotating at vsin i= 39 km/s. Signatures of stellar winds from sdB stars may have been found.     

Orbital periods of the binary sdB stars PG 0940+068 and PG 1247+554

We have used the radial velocity variations of two sdB stars previously reported to be binaries to establish their orbital periods. They are PG 0940+068 ( P =8.33 d) and PG 1247+554 ( P =0.599 d). The minimum masses of the unseen companions, assuming a mass of 0.5 M_⊙ for the sdB stars, are 0.090±0.003 M. for PG 1247+554 and 0.63±0.02 M_⊙ for PG 0940+068. The nature of the companions is not constrained further by our data.     

A photometric study of a new short period eclipsing sdB binary NSVS 07826147

NSVS?07826147 and NSVS?04818255 were suspected as possible eclipsing sdB binary systems by Kelley and Shaw. In order to investigate the short period eclipsing sdB binaries, we have listed them as our observing targets and began to monitor it from March 2009. Till now, we have obtained four complete CCD light curves and 16 high precise times of light minimum of NSVS?07826147. All light curves show strong reflection and a very narrow eclipse, which implies that NSVS?07826147 should be a new short period eclipsing sdB binary. For the system NSVS?04818255, no eclipse character was found in the light curves according to our observations. However, we found that a star near NSVS?04818255 is a close binary system with a period of about 0.32498 days. Up to now, only seven short period eclipsing sdB binary systems have been found, including NSVS?07826147. With the new precise epochs obtained by us, the new period of this system is derived as 0.16177046(5) days, which can be used to predict the epochs of light minimum. Furthermore, the light curves of NSVS?07826147 are analyzed using the Wilson–Van Hamme code and the testing photometric solutions are presented and discussed.     

Very low mass stars in binaries: a theoretical look

We have undertaken a series of hydrodynamic + N ‐body simulations in order to explore the binary properties of young stars. We find that multiple stars are a natural outcome of collapsing turbulent flows, with a high incidence of N > 2 multiples, specially among the higher mass objects. We find a positive correlation of multiplicity with primary mass and a companion frequency that decreases with age, during the first few Myr after formation. Binary brown dwarfs are rarely formed, in conflict with observations. Brown dwarfs as companions are predominantly found orbiting binaries or triples at large separations. The paucity of ultra low mass and low mass ratio binaries has been investigated further, and we tentatively conclude that their formation is intricately related to an appropriate selection of initial conditions and an accurate modelling of disc accretion and evolution. (© 2005 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Population synthesis of Galactic subdwarf B stars

I briefly review the method of population synthesis of binary stars and discuss the preliminary results of a study of the Galactic population of subdwarf B stars. In particular I focus on the formation of (apparently) single sdB stars and their relation to (apparently) single helium-core white dwarfs. I discuss the merits of mergers of two helium white dwarfs and interactions with sub-stellar companions for explaining these single objects. A preliminary conclusion is that the current observations suggest both mechanisms may contribute, but that the helium white dwarfs are likely formed in majority from interactions with sub-stellar companions.     

Metal abundances of subdwarf B stars—the extended sample

The formation of sdBs as well as the chemical composition of their atmospheres is still puzzling. While helium and other light elements are depleted relative to solar values, heavy elements are highly enriched. Diffusion processes in the hot, radiative atmosphere of these stars are the most likely explanation. Although several attempts have been made, it has not yet been possible to model all the observed features of sdB atmospheres. A drawback of most prior studies was the small sample size. We present a detailed abundance analysis of 139 sdBs. A general trend of enrichment was found with increasing temperature for most of the heavier elements. The lighter elements like carbon, oxygen and nitrogen are depleted and less affected by temperature. Although there is considerable scatter from one star to the other, the general abundance patterns in most sdBs are similar. An interplay between gravitational settling, radiative levitation and weak winds is most likely responsible. About 3% of the analysed stars show an enrichment in carbon and helium, which cannot be explained in the framework of diffusion alone. Nuclear processed material must have been transported to the surface somehow.     

Magnetic braking of Ap/Bp stars: application to compact black-hole X-ray binaries

We examine the proposal that the subset of neutron-star and black-hole X-ray binaries that form with Ap or Bp star companions will experience systemic angular-momentum losses due to magnetic braking, not otherwise operative with intermediate-mass companion stars. We suggest that for donor stars possessing the anomalously high magnetic fields associated with Ap and Bp stars, a magnetically coupled, irradiation-driven stellar wind can lead to substantial systemic loss of angular momentum. Hence, these systems, which would otherwise not be expected to experience 'magnetic braking', evolve to shorter orbital periods during mass transfer. In this paper, we detail how such a magnetic braking scenario operates. We apply it to a specific astrophysics problem involving the formation of compact black-hole binaries with low-mass donor stars. At present, it is not understood how these systems form, given that low-mass companion stars are not likely to provide sufficient gravitational potential to unbind the envelope of the massive progenitor of the black hole during a prior 'common-envelope' phase. On the other hand, intermediate-mass companions, such as Ap and Bp stars, could more readily eject the common envelope. However, in the absence of magnetic braking, such systems tend to evolve to long orbital periods. We show that, with the proposed magnetic braking properties afforded by Ap and Bp companions, such a scenario can lead to the formation of compact black-hole binaries with orbital periods, donor masses, lifetimes and production rates that are in accord with the observations. In spite of these successes, our models reveal a significant discrepancy between the calculated effective temperatures and the observed spectral types of the donor stars. Finally, we show that this temperature discrepancy would still exist for other scenarios invoking initially intermediate-mass donor stars, and this presents a substantial unresolved mystery.     

Low mass companions to white dwarfs

This paper summarizes the results of over 17 years of work searching for low mass stellar and substellar companions to more than 370 nearby white dwarfs. Roughly 60 low mass, unevolved companions were found and studied all together, with over 20 discovered in the last few years, including the first unambiguous brown dwarf companion to a white dwarf, GD 1400B. The resulting spectral type distributions for companions to white dwarfs and nearby cool field dwarfs are compared, and the implications for binary star formation are discussed. A brief analysis of GD 1400B, including new data, is also presented. (© 2005 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

The orbital period distribution of subdwarf-B binaries

We present the results of a 3.5 year long campaign to measure orbital periods of subdwarf-B (sdB) star binaries. We directly compare our observed orbital period distribution with that predicted by using binary population synthesis. Up to now, most of our systems seem to have been formed through two of the formation channels discussed by Han et al. (2003, MNRAS 341, p. 669), i.e. the first and the second common envelope ejection (CE) channels. At this point, thanks to the long baseline of our observations, we are starting to detect also very long orbital period systems. These have probably come from a complete different formation path, the first stable Roche Lobe overflow (RLOF) channel in which the first mass transfer phase is stable. This channel is expected to lead to the formation of very wide binaries with typical orbital periods ranging from 1 month to 1 year.     

Predicting the properties of binary stellar systems: the evolution of accreting protobinary systems

We investigate the formation of binary stellar systems. We consider a model where a 'seed' protobinary system forms, via fragmentation, within a collapsing molecular cloud core and evolves to its final mass by accreting material from an infalling gaseous envelope. This accretion alters the mass ratio and orbit of the binary, and is largely responsible for forming the circumstellar and/or circumbinary discs.
Given this model for binary formation, we predict the properties of binary systems and how they depend on the initial conditions within the molecular cloud core. We predict that there should be a continuous trend such that closer binaries are more likely to have equal-mass components and are more likely to have circumbinary discs than wider systems. Comparing our results with observations, we find that the observed mass-ratio distributions of binaries and the frequency of circumbinary discs as a function of separation are most easily reproduced if the progenitor molecular cloud cores have radial density profiles between uniform and 1/ r (e.g., Gaussian) with near-uniform rotation. This is in good agreement with the observed properties of pre-stellar cores. Conversely, we find that the observed properties of binaries cannot be reproduced if the cloud cores are in solid-body rotation and have initial density profiles which are strongly centrally condensed. Finally, in agreement with the radial-velocity searches for extrasolar planets, we find that it is very difficult to form a brown dwarf companion to a solar-type star with a separation ≲10 au, but that the frequency of brown dwarf companions should increase with larger separations or lower mass primaries.     

Close binary EHB stars from SPY

We present the results of a radial velocity (RV) survey of 46 subdwarf B (sdB) and 23 helium-rich subdwarf O (He-sdO) stars. We detected 18 (39%) new sdB binary systems, but only one (4%) He-sdO binary. Orbital parameters of nine sdB and sdO binaries, derived from follow-up spectroscopy, are presented. Our results are compared with evolutionary scenarios and previous observational investigations.     

The proper-motion signal of unresolved binaries in the Hipparcos catalogue

We present an investigation of the differences between quasi-instantaneous stellar proper motions from the Hipparcos catalogue and long-term proper motions determined by combining Hipparcos and the Astrographic Catalogue. Our study is based on a sample of about 12000 stars of visual magnitude from 7 to 10 in two declination zones on the northern and equatorial sky. The distribution of the proper-motion differences shows an excess of large deviations. This is caused by the influence of orbital motion of unresolved binary systems. The proper-motion deviations provide statistical evidence for 360 astrometric binaries in the investigated zones, corresponding to about 2400 such binaries in the entire Hipparcos catalogue, in addition to those already known. In order to check whether the observed deviations are compatible with standard assumptions on the basic parameters of binary stars, we model the impact of orbital motion on the observed proper motions in a Monte Carlo simulation. We show that the simulation yields an acceptable approximation of the observations, if a binary frequency between 70% and 100% is assumed, i.e.if most of the stars in the sample are assumed to have a companion. Thus Hipparcos astrometric binaries confirm that the frequency of non-single stars among field stars is very high. We also investigate the influence of the mass function for the secondary component on the result of the simulation. A constant mass function and mass functions with moderate increase towards low masses lead to results, which are compatible with the observed proper-motion effects. A high preponderance of very-low-mass or substellar companions as produced, for example, by a M^—1 power law is not in agreement with the frequency of proper-motion deviations in our sample of stars.     

Mass-loss predictions for Subdwarf B stars

We present the results of Monte Carlo mass-loss computations for hot low-mass stars, specifically for subdwarf B (sdB) stars. It is shown that the mass-loss rates on the Horizontal Branch (HB) computed from radiative line-driven wind models are not high enough to create sdB stars. We argue, however, that mass loss plays a role in the chemical abundance patterns observed both in field sdB stars, as well as in cluster HB stars. The derived mass loss recipe for these (extremely) hot HB stars may also be applied to other groups of hot low-mass stars, such as post-HB (AGB-manqué, UV-bright) stars, over a range in effective temperatures between ?10 000 and 50 000 K. Finally, we present preliminary spectral synthesis on the more luminous sdB stars for which emission cores in Hα have been detected (Heber, U., et al.: 2003, in:Stellar Atmosphere Modeling, ASP Conference Proceedings, p. 251). We find that these line profiles can indeed be interpreted as the presence of a stellar wind with mass loss on the order of 10^?11?M _⊙ yr ^?1.     

On the formation and evolution of black hole binaries

We present the results of a systematic study of the formation and evolution of binaries containing black holes and normal-star companions with a wide range of masses. We first reexamine the standard formation scenario for close black hole binaries, where the progenitor system, a binary with at least one massive component, experienced a common-envelope phase and where the spiral-in of the companion in the envelope of the massive star caused the ejection of the envelope. We estimate the formation rates for different companion masses and different assumptions about the common-envelope structure and other model parameters. We find that black hole binaries with intermediate- and high-mass secondaries can form for a wide range of assumptions, while black hole binaries with low-mass secondaries can only form with apparently unrealistic assumptions (in agreement with previous studies).
We then present detailed binary evolution sequences for black hole binaries with secondaries of 2 to 17 M_⊙ and demonstrate that in these systems the black hole can accrete appreciably even if accretion is Eddington-limited (up to 7 M_⊙ for an initial black hole mass of 10 M_⊙) and that the black holes can be spun up significantly in the process. We discuss the implications of these calculations for well-studied black hole binaries (in particular GRS 1915+105) and ultraluminous X-ray sources of which GRS 1915+105 appears to represent a typical Galactic counterpart. We also present a detailed evolutionary model for Cygnus X-1, a massive black hole binary, which suggests that at present the system is most likely in a wind mass-transfer phase following an earlier Roche-lobe overflow phase. Finally, we discuss how some of the assumptions in the standard model could be relaxed to allow the formation of low-mass, short-period black hole binaries, which appear to be very abundant in nature.