Are Jets in Symbiotic Stars Driven by Magnetic Fields?
Total Page:16
File Type:pdf, Size:1020Kb
A&A 432, L17–L20 (2005) Astronomy DOI: 10.1051/0004-6361:200500012 & c ESO 2005 Astrophysics Editor Are jets in symbiotic stars driven by magnetic fields? the M. Stute and M. Camenzind to Landessternwarte Heidelberg, Königsstuhl, 69117 Heidelberg, Germany e-mail: [email protected] Letter Received 25 November 2004 / Accepted 26 January 2005 Abstract. We compare two scenarios to launch jets – formation by MHD processes or formation by thermal pressure in the boundary layer (BL) – with respect to their compatibility with observational data of jets in symbiotic stars, especially in the well studied jet source MWC 560. Finally, we discuss points of further research to be done. Key words. ISM: jets and outflows – binaries: symbiotic 1. Introduction lobe overflow, not due to wind accretion. The short timescales in these systems make them the best understood accreting sys- ff Although jets are ubiquitous phenomena in many di erent as- tems. Remarkably, they show no jet emission. Further related trophysical objects as young stellar objects where they are objects are supersoft X-ray sources, in which the temperature driven by protostars, symbiotic stars (white dwarfs), X-ray bi- and pressure in the boundary layer (BL) of the WD are in naries (neutron stars and stellar mass black holes) and active the correct range to maintain steady nuclear burning on the galactic nuclei (supermassive black holes), their formation is WD surface. relatively unclear. In jet formation models presented so far, the magnetic The mass loss rate of the jet is found to be connected to the field seems to play a key role. The first analytical work mass accretion rate of the underlying disc found in most objects studying magneto-centrifugalacceleration along magnetic field (e.g. Livio 1997). Therefore the necessary components seem to lines threading an accretion disc was done by Blandford & be well known and common to all objects. A more careful in- Payne (1982). They have shown the braking of matter in az- vestigation of one specific class of objects should promise new imuthal direction inside the disc and their acceleration above insights also for the mechanisms in the other classes. From the the disc surface by the poloidal magnetic field components. observational point of view, one needs observations with a high Toroidal components of the magnetic field then collimate the spatial resolution and kinematic informations from regions as flow. Numerous semi-analytic models extended the work of near as possible to the jet source. These aspects make the class Blandford & Payne (1982), either restricted to self-similar so- of symbiotic stars ideal testbeds. lutions and their geometric limitations (e.g. Pudritz & Norman Symbiotic stars are interacting binaries with orbital periods 1986; Vlahakis & Tsinganos 1998, 1999; Ferreira & Casse in the range of years. These systems show outbursts similar to 2004) or with non-self-similar solutions (e.g. Camenzind 1990; classical novae. The stellar component is a cool red giant (RG), Pelletier & Pudritz 1992; Breitmoser & Camenzind 2000). the hot component a white dwarf (WD) with temperatures of Another approach is to use time-dependant numerical 50 000–200 000 K. Both stars show mass loss through super- MHD simulations to investigate the formation and collima- sonic winds. Wind material from the RG is captured by the tion of jets. In most models, however, a polytropic equilib- WD to form an accretion disc. The accretion then causes ther- rium accretion disc was regarded as a boundary condition monuclear explosions of the WD surface leading to an increase (e.g. Krasnopolsky et al. 1999, 2004; Anderson et al. 2004; in luminosity followed by jet emission. Jets are detected in 10 Goodson et al. 1999). The magnetic feedback on the disc struc- out of almost 200 symbiotic stars (Brocksopp et al. 2004) and ture is therefore not calculated self-consistently. Only in recent this process was directly observed in CH Cygni (Taylor et al. years were the first simulations including the accretion disc 1986). Other famous systems are R Aquarii and MWC 560. self-consistently in the calculations of jet formation presented While the first two objects are seen at high inclinations, the jet (e.g. Casse & Keppens 2002, 2004; Kato et al. 2004). axis in MWC 560 is practically parallel to the line of sight. Due to the fact, that strong magnetic fields have Another class of accreting WDs are cataclysmic variables been detected so far only in one symbiotic system (CV) which are very close WD binaries with a low mass main (Z Andromedae, Sokoloski & Bildsten 1999), the jet forma- sequence star as companion. The mass transfer is due to Roche tion by magneto-centrifugal forces exclusively seems to be Article published by EDP Sciences and available at http://www.aanda.org or http://dx.doi.org/10.1051/0004-6361:200500012 L18 M. Stute and M. Camenzind: Are jets in symbiotic stars driven by magnetic fields? insufficient. Radiative launching can be excluded due to too – or the magnetic luminosity (Camenzind 1997) small radiation fields. A new possibility to accelerate plasma 1 3 34 close to the central object was proposed involving SPLASHs Lmag = Ω∗ R∗ Bp Bϕ = 4.76 × 10 (SPatiotemporal Localized Accretion SHocks) in the BL (Soker 2 3 Editor Ω ∗ R∗ Bp Bϕ −1 & Regev 2003). Locally heated bubbles expand, merge and ac- × erg s . (3) celerate plasma to velocities larger than the local escape veloc- h−1 7 × 108 cm MG MG ity. Soker & Lasota (2004) applied this model to disk-accreting − − − the 9 2 1 white dwarfs to explain the absence of jets in CV. They have The flux in the UV band was measured as 10 erg cm s −6 −1 (Maran et al. 1991) and the UV luminosity – which is likely to found a critical accretion rate of ∼10 M yr below which to no jets should be present. This scenario was introduced only in be equal to the accretion power of the disk and the boundary analytic estimates. layer – is then In Sect. 2, we give general estimates based on observations 2 35 d −1 of the well studied jet in MWC 560. After that we calculate the LUV = 1.2 × 10 erg s . (4) Letter magnetic field near the white dwarf required for jet formation kpc ff by MHD processes in Sect. 3. In Sect. 4, we list di erent ways With a derived distance of 2.5 kpc, this suggests a minimal ac- with which a magnetic field in MWC 560 could be detected. −8 −1 cretion rate of 6.3 × 10 M yr and therefore an ejection Finally a discussion is given. efficiency of 14%. The minimal accretion rate is below the critical accretion 2. General estimates rate derived by Soker & Lasota (2004) by a factor of 16. As these authors claim an uncertainty of a factor of ∼10 and the As observed by Schmid et al. (2001) and as used to simulate remaining factor could naturally arise from our estimates, this the jet nozzle in Stute et al. (2005), the parameters of the jet in should not concern. MWC 560 are Using the observables M˙ , Ljet and P˙ 0, one can fix in prin- ciple the two free parameters χ and Γ of the model of Soker −1 – the velocity vjet = 1000 km s ; & Regev (2003). These parameters are the fractions of the 6 −3 – the number density njet = 5 × 10 cm , which is equal to a mass outflow rate due to SPLASHs to the accretion rate and −18 −3 mass density ρjet = 8.4 × 10 gcm ;and of the initial kinetic energy to the final kinetic energy inside – the jet radius Rjet = 1AU. a SPLASH, respectively. Fixing the model highly depends on estimating the right mass accretion rate which, however, is Using the equations rather difficult. ˙ = π 2 v Mjet Rjet mH njet jet 3. Required magnetic fields in the magnetic jet P˙0 = M˙ jet vjet formation scenario 1 2 1 2 3 Ljet = M˙ v = π R mH njet v , 2 jet 2 jet jet Let Ψ be a magnetic surface anchored at the inner radius of the this specifies accretion disk. Then this surface remains always inside the jet up to large distances. Using the constants of motion along this −9 −1 – the mass outflow rate M˙ jet = 9.33 × 10 M yr ; surface and the jet parameters, one can make detailed estimates 25 −1 −2 – the momentum discharge P˙0 = 5.93 × 10 gcm s ;and following from the scenario. 33 −1 – the kinetic jet luminosity Ljet = 2.93 × 10 erg s . As magneto-centrifugally driven jets have fast- magnetosonic Mach numbers Mfm ∼ 3 (e.g. Krasnopolsky This luminosity can be provided by several different et al. 2004), the total magnetic field in the jet should be mechanisms: 2 2 1 B , + Bϕ, = 4 πρjet vjet = 0.34 G ∼ Bϕ,jet, (5) – by the (insufficient) luminosity of the white dwarfs p jet jet 3 2 4 as the azimuthal component should be dominant inside the jet. LWD = 4 π R∗ σ T = ϕ 2 4 With the conservation of current I RB , the azimuthal field R∗ T = 3.5 × 1030 erg s−1;(1)near the jet source is then 7 × 108 cm 104 K ∼ . – the accretion power of the disk (and almost the same Bϕ,0 7 3kG (6) amount coming from a BL) of a rotating magnetized WD As Bϕ, B , , the total magnetic field should be by far 0 p 0 larger. An upper limit can be found using again Eq. (5), namely GM∗ M˙ M∗ = = . × 35 2 Lacc 1 19 10 Bp,jet 0.34 G, and the conservation of flux Ψ=Bp R , lead- R∗ M ing to magnetic fields in the range of −1 ˙ × R∗ M −1 8 −8 −1 erg s ;(2) 7 × 10 cm 10 M yr 7.3kG Bp,0 155 MG.