6. High-Mass X-Ray Binaries

In HMXB the donor star is more massive than $\sim$ 2M_$\odot$ (Hayakawa (1985)). The orbital separation is generally wider than in a typical LMXB because the companion is larger. The companion is generally highly evolved and dominates the optical emission. Tidal distortion of the companion can cause ellipsoidal variations in the optical light curve. These occur partly because the observer will see different parts of the distorted companion at different angles as it rotates. The optical light curve may also be affected by X-ray heating and light from an accretion disk. By modeling the ellipsoidal variations it is possible to determine how close the companion comes to filling its Roche Lobe (e.g. van Paradijs and McClintock (1995) and references therein). The companion which is of spectral type O or B has a strong stellar wind. If the wind is not highly ionized it may be driven by the absorption of UV photons from the companion. This is called a line-driven wind. If the wind material is highly ionized it will be transparent to UV radiation and another driving mechanism such as X-ray heating may be involved. Accretion can also take place by Roche lobe overflow which is favored by the low density of the donor. The dominant ionization mechanism is photoionization due to the X-ray luminosity which may be as high as $\sim$ 10³⁸ erg s^-1. Because the HMXB have heavier companions they tend to be younger than LMXB. The HMXB are population I objects and are concentrated in the Galactic plane. LMXB, in contrast, tend to be population II objects and their spatial distribution shows no preference for the Galactic plane, although they tend to be concentrated towards the Galactic center. The HMXB also tend to have stronger magnetic fields than the LMXB. It is possible that this is due to accretion-induced decay of the neutron star's magnetic field. However the issue of magnetic field decay is not settled (Bhattacharya and Srinivasan (1995)). For example, the LMXB pulsar Her X-1 has a field of 3 x 10¹² G (Voges et al. (1982)) and is believed to be > 10⁸ years old while X2259+586 which is a young pulsar in a supernova remnant seems to have a field of  5 x 10¹¹ G (Koyama et al. (1986)). Comparison of nine X-ray pulsars whose fields are known from cyclotron line observations (Makishima et al. (1992)) indicates that the magnetic field does not decay on the timescale of the lifetime of massive X-ray binaries ($\sim$ 10⁷ years). A consequence of the lower average magnetic fields of LMXB is that the accretion disk extends closer to the accreting star where the radiation flux is high. This can photoionize and heat the inner accretion disk creating an ADC which can modify the spectrum (e.g. Kallman (1993)).

The properties of the known accreting X-ray pulsars (AXP) are summarized by Bildsten et al. (1997). Table 2.6 shows the properties of the known HMXB with supergiant companions that are also pulsars (Bildsten et al. (1997)). The compact component in these systems is a neutron star with a high magnetic field ( $\sim$ 10¹² G). The pulsed emission comes from the channeling of ionized accreting matter onto the polar caps of the rotating neutron star. These systems are distinct from the rotation-powered pulsars. They have much longer spin periods (on the order of 1- 10² s). At these periods the dipole radiation from an isolated pulsar would be undetectable. Such long periods cannot be attained by the spin-down power of a rotation-powered pulsar. The AXP must have been spun down by some other mechanism. The most likely mechanism is accretion. Davidson and Ostriker (1973) showed that a fast-spinning neutron star would not be able to accrete matter because of the centrifugal barrier in the effective potential. This is known as the propeller effect (Illarionov and Sunyaev (1975)). The rotating magnetosphere would eject any infalling matter from the system, resulting in a braking torque on the pulsar. For a pulsar with spin period P₀ the efficiency of this braking exceeds that of dipole-radiation braking by a factor (r_c/r_m)³ where r_c = P₀c/2$\pi$ is the light-cylinder radius and r_m is the magnetosphere radius (see Section 5.4).

Some HMXB have detectable radio emission which is mostly due to synchrotron radiation in lobes of matter ejected from the system. However, in contrast to the rotation-powered pulsars none of the pulsar systems in Table 2.6 is a radio source. In fact, none of the LMXB that are known to be accreting pulsars has detectable radio emission (Hjellming and Han (1995)). This is not surprising because if the early-type companion's wind is significantly ionized it will obscure radio emission from the source's inner regions. The refractive index of the medium for radio waves of frequency $\nu$ is

$$\sqrt{1-\left({\nu_p\over \nu}\right)^2}$$ (10)

where

$$\nu_{p}^{} \sqrt{4\pi N_e e^2\over m_e}$$ (11)

is the plasma frequency. Close to the source where N_e is high, n will be imaginary and the radio emission will be attenuated. Thus any detectable radio emission will come from regions far enough from the source so that $\nu_{p}^{}$ < $\nu$. The minimum distance for which this condition is met is between 10¹² and 10¹⁵ cm (Hjellming (1978)). If matter is ejected from the system it may have detectable synchrotron emission in radio. An example of this is the non-pulsating HMXB SS433 which has radio jets (e.g. Hjellming and Han (1995)). X-ray binaries that have radio jets are known as microquasars. These are believed to be binary systems where the primary is a neutron star or black hole which is accreting at super-Eddington rates, causing the formation of jets. The name microquasar comes from the similarity of these jets to those ejected by the accreting black holes at the centers of active galactic nuclei.

Table 1: Accretion-powered pulsars in high-mass supergiant and giant systems (Bildsten et al. (1997) and references therein)

Name	lII	bII	Pspin (s)	Porb (d )	Companion
Cen X-3	292.1	+0.3	4.82	2.09	V779 Cen (06-8f)
SMC X-1	300.4	-43.6	0.717	3.89	Sk16O (BO I)
RX J0648.1-4419	253.7	-19.1	13.18	1.54	HD 49798 (06p)
LMC X-4	276.3	-32.5	13.5	1.408	Sk-Ph (07 III-V)
OAO 1657-415	344.4	+0.3	37.7	10.4	(B0-6Iab)
Vela X-1	263.1	+3.9	283	8.96	HD77581 (BO.5Ib)
lE 1145-614	295.5	-0.0	297	5.648	V830 Cen (B2Iae)
4U 1907+09	43.7	+0.5	438	8.38	(B I)
4U 1538-52	327.4	+2.1	530	3.73	QV Nor (BO1ab)
GX 301-2	300.1	-0.0	681	41.5	Wray 977 (B1.5Ia)