Ss 433: the Second Wolf-rayet X-ray Binary ?

We present mid-infrared spectrophotometric observations of SS 433 with ISOPHOT. The HeI+HeII lines in both spectra of SS 433 and of the Wolf-Rayet star WR147, a wind-colliding WN8+BO5 binary system, closely match. The 2.5-12 μm continuum radiation is due to an expanding wind free-free emission in an intermediate case between optically thick and optically thin regimes. The inferred mass loss rate evaluation gives ∼ 10 M⊙.yr . Our results are consistent with a Wolf-Rayet-like companion to the compact object in SS 433. A similar study for CygnusX-3 confirms the Wolf-Rayetlike nature of its companion, although with a later WN type than previously suggested. 1. The mysterious SS 433 system SS 433 has been discovered as a source of strong Hα lines, and as a variable radio and X-ray source [1]. It is an X-ray binary with a ∼ 13 days period, producing bipolar relativistic (0.26 c) jets undergoing a precession movement in ∼ 162.5 days, so covering a cone with an opening angle of 19.8 and an axis of ∼ 78.8 with the line of sight. This precession has indeed been observed in radio, at arcsec [2] and milliarcsec scales [3]. So SS 433 was actually the first microquasar discovered. In spite of intensive studies, no line has been associated to the companion star yet. It is generally accepted but has never been proved that SS 433 is a neutron star + massive star binary system. 2. Mid-IR observations of SS 433 2.1. Comparison with Wolf-Rayet stars spectra SS 433 has been observed with ISOPHOT, in the spectral mode at 2.5-5 μm (∼ 0.04 μm resolution) and 6-12 μm (∼ 0.1 μm resolution), and in the photometric mode at 12, 25 and 60 μm, in November 1996 and April 1997. We looked for WolfRayet (WR) star observations in the ISO archives, as it was the suspected type for SS 433 companion star. Several WRs were observed with ISOSWS, and so we rebined the SWS spectra to the spectral resolution of ISOPHOT. As a lot of hydrogen is seen in SS 433 optical spectra, we chose to compare it to WRs of nitrogen type (WN) which are in the first stage of WR evolution. Figure 1 (left) shows the observed spectra of the four corresponding WRs, classified from the not very evolved (late type WNL) to the evolved (early type WNE) ones: WR147 ↔ WN8+B0.5 (WNL), WR78 ↔ WN7h (WNL), WR136 ↔ WN6b (WNE) and WR134 ↔ WN6 (WNE). Comparing on figure 1 these spectra to the one of SS 433, it is clear that SS 433 is closest to WR147, a late type WR. 2 Fuchs, Koch-Miramond & Ábrahám Figure 1. Left: Comparison of SS 433 observed spectra with that of Wolf-Rayet stars (of WN type). Right: Comparison of WR147 dereddened spectrum with the ones of SS 433 taken at different precession phases (Φp) and orbital phases (φ). Lines are identified according to [5]. Figure 1 (right) also shows the comparison of WR147 spectrum with that of SS 433 taken at different phases. SS 433 and WR147 spectra were dereddened with AV = 8 and 11.2 mag respectively, using the Lutz et al. [4] law. The lines, identified according to [5], are strong and weak H+HeI blended lines (because of ISOPHOT spectral resolution) and weak HeII lines. No metallic line was observed. The SS 433 line spectrum is clearly WR-like. We note that WR147 is known as a colliding wind binary system which may explain its resemblance to SS 433. 2.2. Continuum Figure 2 shows the continuum spectra of SS 433, which flux density (Fν) is well fitted by power laws in the 2-12 μm range, corresponding to free-free emission: optically thin for Fν ∝ λ , and in the intermediate regime between optically thin and thick for Fν ∝ λ . In far-IR, an additional component, as a blackbody due to dust surrounding the system, is needed. The 0.6 spectral index is characteristic of an ionized homogeneous wind with a spherical expansion at a constant velocity [7], although it is also valid for a more complex geometry as long as it stays thick [6]. So SS 433 continuum spectrum corresponds to the one emitted by a standard O or WR wind. 2.3. Mass loss rate We calculated the mass loss rate of this free-free emitting wind, following the Wright & Barlow [7] formula (8). With a distance D=3.5 kpc, a Gaunt factor g∼ 1, a Fν =1000mJy flux at 4μm (7.5×10 13 Hz), and for a WN-type wind where SS 433: the second Wolf-Rayet X-ray binary ? 3 the mean atomic weight per nucleon μ = 1.5, the number of free electrons per nucleon γ = 1, the mean ionic charge Z=1 and the velocity v∞ = 1000 km.s , we find: Ṁ = 1.0× 10 M⊙.yr . However, the recent WN mass-loss rate estimates [8] show that this value has to be lowered by a factor 2 or 3 due to clumping in the wind. This is in good agreement with the mass transfer rate estimated by van den Heuvel et al. [9] assuming a normal homogeneous WR wind, or with the recent mass transfer values obtained from simulations of SS 433 evolution by King et al. [10]. For WR147, Ṁ ≃ 1.5 − 3.7 × 10 M⊙.yr −1 and the WN mean is ∼ 3× 10 M⊙.yr , so our ∼ 10 M⊙.yr −1 mass loss evaluation for SS 433 is compatible with a strong WNL wind. Figure 2. Left: Continuum spectra of SS 433. Right: Continuum spectra of CygnusX-3. What happens to the material contained in this wind is unknown since only ∼ 10 M⊙.yr −1 is ejected into the jets [11]. It may be accreted then ejected via the L2 point [12], or close to the compact object [10]. It may then form dust seen in far-IR at a distance > 20 a.u. This material is probably responsible for the equatorial outflows observed with the VLBA [13][14] at ∼ 100 a.u. from SS433.