A&A 578, A107 (2015) Astronomy DOI: 10.1051/0004-6361/201424823 & c ESO 2015 Astrophysics The donor star of the X-ray pulsar X1908+075, S. Martínez-Núñez1, A. Sander2, A. Gímenez-García1,3, A. Gónzalez-Galán2, J. M. Torrejón1,3, C. Gónzalez-Fernández4 , and W.-R. Hamann2 1 X-ray Astronomy Group, Departamento de Física, Ingeniería de Sistemas y Teoría de la Señal, University of Alicante, PO Box 99, 03080 Alicante, Spain e-mail: [email protected] 2 Institut für Physik und Astronomie, Universität Potsdam, Karl-Liebknecht-Str. 24/25, 14476 Potsdam, Germany 3 Instituto Universitario de Física Aplicada a las Ciencias y las Tecnologías, University of Alicante, PO Box 99, 03080 Alicante, Spain 4 Institute of Astronomy, University of Cambridge, Madingly Road, Cambridge, CB3 0HA, UK Received 17 August 2014 / Accepted 21 April 2015 ABSTRACT High-mass X-ray binaries consist of a massive donor star and a compact object. While several of those systems have been well studied in X-rays, little is known for most of the donor stars as they are often heavily obscured in the optical and ultraviolet regime. There is an opportunity to observe them at infrared wavelengths, however. The goal of this study is to obtain the stellar and wind parameters of the donor star in the X1908+075 high-mass X-ray binary system with a stellar atmosphere model to check whether previous studies from X-ray observations and spectral morphology lead to a sufficient description of the donor star. We obtained H-andK-band spectra of X1908+075 and analysed them with the Potsdam Wolf-Rayet (PoWR) model atmosphere code. For the first time, we calculated = ± = +6 = ± a stellar atmosphere model for the donor star, whose main parameters are: Mspec 15 6 M, T∗ 23−3 kK, log geff 3.0 0.2 and log L/L = 4.81 ± 0.25. The obtained parameters point towards an early B-type (B0–B3) star, probably in a supergiant phase. Moreover we determined a more accurate distance to the system of 4.85 ± 0.50 kpc than the previously reported value. Key words. binaries: close – stars: individual: X1908+075 – stars: massive – stars: winds, outflows – X-rays: binaries 1. Introduction This study concludes that the system contains a highly magnetized neutron star orbiting in the wind of a massive com- High-mass X-ray binaries (HMXBs) are composed of a compact panion star (mass function of (6.07 ± 0.35) M) with an orbital star orbiting a donor star from which there is an accretion of ma- period of 4.400±0.001 days. Thus, they identify X1908+075 as a terial (see Chaty 2013, for a review). There is a broad literature HMXB. Moreover, this study found changes in the optical depth of the X-ray studies of these types of systems compared to the along the line of sight to the compact object correlated with the few detailed studies performed about the donor stars using stel- orbital phase. These changes could be causing the observed soft lar atmosphere models (e.g. Clark et al. 2002; González-Galán X-ray modulation. et al. 2014). The physical characteristics of the donor stars har- Levine et al. (2004) estimate a mass of the donor star in bouring HMXBs, however, are essential for a complete picture the range of 9–31 M and an upper limit on its size of about of the physical processes occurring in these binary systems. + + 22 R, using the mass function and the orbital inclination angle. The binary system X1908 075, also known as 4U 1909 07, Previous values were derived from modelling the orbital phase- was first discovered in 1978 by Forman et al. (1978) with the dependence of the X-ray flux. They also infer a wind mass-loss Uhuru satellite. The source has been observed in surveys car- −6 −1 rate for the donor star of 4 × 10 M yr , which is larger than ried out with OSO7, Ariel, HEAO-1, EXOSAT, and INTEGRAL the expected theoretical value according to Vink et al. (2000). satellites, among others. It is a persistent X-ray source that shows Given the high rate combined with the estimated mass and ra- fluctuations of 10% in the soft X-rays, 2–12 keV energy range dius, Levine et al. (2004) concluded that the system might be a (Levine et al. 2004). Wolf-Rayet star with a neutron star companion that could evolve The binary system is a highly absorbed and faint pulsar that and become a black hole-neutron star system in 104 to 105 yrs. shows strong X-ray pulsations at a period of 605 s (Levine et al. + Morel & Grosdidier (2005) analysed near-infrared observa- 2004). The location of X1908 075 in the pulse–orbital period tions of stars in, or close to, the error box of HEAO-1/A3.They plane (Corbet 1986) clearly indicates that this is a wind-fed, suggest that the optical counterpart of the system might be a late HMXB. O-type supergiant at a distance of 7 ± 3 kpc. Levine et al. (2004) determined the orbital parameters of ∼ / In this paper, we present intermediate-resolution (R 2500) the system (see Table 1) with RXTE PCA pointed observations. infrared spectroscopic observations of the counterpart of the + Based on observations made with the William Herschel Telescope X-ray pulsar X1908 075 in the error box reported by the operated on the island of La Palma by the Isaac Newton Group in the Chandra position. For the first time, a non-LTE stellar model at- Spanish Observatorio del Roque de los Muchachos of the Instituto de mosphere analysis of X1908+075 is performed providing impor- Astrofísica de Canarias. tant physical parameters, such as the stellar temperature (T∗), the Appendix A is available in electronic form at terminal velocity of the wind (∞)andtheeffective surface grav- http://www.aanda.org ity (geff), for the donor star. We also calculate a more accurate Article published by EDP Sciences A107, page 1 of 9 A&A 578, A107 (2015) Table 1. Relevant orbital parameters derived by Levine et al. (2004). Parameter Orbital second-epoch values 12 ax sin i [cm] (1.43 ± 0.03) × 10 τ90 [MJD] 52 631.383 ± 0.013 Pspin ([s]) 604.684 ± 0.001 −1 −8 Pspin˙ [s s ] (1.22 ± 0.09) × 10 e 0.021 ± 0.039 Porb [days] 4.4007 ± 0.0009 f (M)[M]6.07± 0.35 value of the distance to the binary system and enclose the value of its inclination. We conclude that the optical counterpart is, in contrast to previous assumptions, an early B-type (B0–B3) star, probably in a supergitant phase. In Sect. 2 we describe the details of the observations, fol- lowed by a short overview of the stellar atmosphere models in Sect. 3. The results are shown in Sect. 4, discussed in Sect. 5, and conclusions are given in Sect. 6. Fig. 1. UKIDSS image of the X1908+075 region. North is up and east is left. The red circle indicates the Chandra position of the X-ray source. Size of the image: 1 × 1. 2. Observations / In July 2009 (proposal id: 135-WHT45 09A), we obtained code (PoWR). The PoWR code provides a model for a spher- K H two intermediate-resolution spectra ( and -band spectra at ical symmetric star with an expanding atmosphere by iteratively 55 018.97 MJD and 55 019.92 MJD respectively) of the opti- + solving the radiative transfer equation and the statistical equi- cal counterpart of X1908 075 using the Long-slit Intermediate librium equations in non-LTE. The code further includes energy Resolution Infrared Spectrograph (LIRIS) mounted on the 4.2 m conservation and treats both the wind and photosphere in a con- William Herschel Telescope (WHT), at the Observatorio del sistent scheme. Roque de los Muchachos (La Palma, Spain). The instrument is × The main aspects of the code are summarized in Gräfener equipped with a 1024 1024 pixel HAWAII detector. We took et al. (2002)andHamann & Gräfener (2003). This code has . advantage of the excellent seeing and made use of the 0 65 slit been applied to all kinds of stars that could be potential donors in K H in combination with the intermediate-resolution and pseu- wind-fed HMXB systems, such as O and B stars (e.g. Oskinova K − dogrisms. The configuration covers the 2055 2415 nm range, et al. 2011; Evans et al. 2012) and Wolf-Rayet (WR) stars (e.g. R ∼ giving a minimum resolving power 2500 at 2055 nm and Hamann et al. 2006; Sander et al. 2012). Wind inhomogeneities a slightly higher resolving power at longer wavelengths. The are considered in a so-called “micro-clumping” approach, as- H − configuration covers the 1520 1783 nm range, giving a min- suming that the wind is not smooth, but instead consisting of op- R ∼ imum 2500 at 1520 nm. tical thin cells with an increased density and a void interclump The position of the X-ray source was accurately determined medium (cf. Hamann & Koesterke 1998). using the High Energy Transmission Gratings (HETG) on board The “stellar radius” R∗ marks the lower boundary of the Chandra m = ± , as the intersection between grating orders 1and model atmosphere and is set at a Rosseland optical depth of the zero order image. The coordinates of the X-ray source are τ = 20, where we assume that the hydrostatic equation is ful- α = h . δ =+ ◦ . 19 10 48 2 and 07 35 51 8 , with an estimated un- filled. In the quasi-hydrostatic regime we use this equation to .