X-Ray Spectroscopy of Centaurus X-3 with ASCA over an Eclipse

Abstract

The X-ray binary pulsar Cen X-3 was observed with the ASCA satellite over an entire binary eclipse. With unprecedented energy resolution at the iron K-energy band (FWHM ∼ 2%), three iron emission lines at 6.4 keV, 6.7 keV, and 6.97 keV were clearly resolved. At lower energies, lines from hydrogenic ions of Ne, Mg, Si, and S were also observed. Spectral variations over the eclipse were studied, in particular variations of the iron emission lines. Although the intensity of the 6.4 keV line, which is due to fluorescence from low ionized iron, decreased about one order of magnitude during the eclipse, the intensity drop of the 6.7 and 6.97 keV lines, which are respectively from Fe XXV and Fe XXVI, was at most a factor of three. This, confirming the conjecture by Nagase et al. (1992, AAA56.117.175), indicates that the 6.4 keV fluorescent line is emitted from the region close to the neutron star while the highly photoionized plasma which emits 6.7 and 6.97 keV lines is more extended than the size of the companion star (∼ 12 R⊙). The 6.4 keV line equivalent-width did not show a large variation over the eclipse, and was 75–145 eV. The dense Alfvén shell or an optically thick accretion disk is a likely candidate for reprocessing the fluorescent iron line. From the intensity ratio of the 6.7 and 6.97 keV lines during the eclipse, the ionization parameter of the photoionized plasma is estimated as ξ ∼ 103.4. The size and the density of the plasma are estimated as ∼ 16 R⊙ and ∼ 1.3 × 1011 cm−3, respectively. It is suggested that a lower ionization state is dominated in the pre-eclipse phase, since the 6.7 and 6.97 keV lines were not clearly seen and other K-lines from lighter elements were strong.

Delayed Iron Lines in Centaurus X‐3

“…Considering the equivalent width of the iron line, the optical depth is not much smaller than 0.1. Ebisawa et al (1996) reported that the optical depth was 0.52^0.10 from ASCA SIS data. Thus, we calculated the average delay of the iron band continuum from the original X-rays as…”

“…Although we do not know the geometry of the reprocessor, it may cover more than half of the sphere (i.e., 2n steradian) as seen from the neutron star (Ebisawa et al 1996), and our derived delay time is the averaged delay time during the observation. Thus, we assume the spherical shell for the reprocessor and calculate the distance between neutron star and the reprocessor by simply multiplying our derived delay time by light velocity.…”

“…Because the equivalent width of the iron line is D128 eV, the reprocessor must cover more than half of the sphere (e.g., Ebisawa et al 1996). On the other hand, the fact that the iron line emission is pulsed (Day et al 1993) indicates that the reprocessor does not cover the whole neutron star if the reprocessor is optically thin.…”

“…The energy of the absorption edge of the iron ion indicates that the ionization state is at most Fe IX (Ebisawa et al 1996). This requires the ionization parameter m \ L /(nr2) (Tarter, Tucker, & Salpeter 1969) must be 1È102 (Kallman & McCray 1982).…”

“…Because this line emission is not observed during eclipse, it must originate near to the neutron star (Nagase et al 1992 ;Ebisawa et al 1996) implying that dense matter exists near to the neutron star. Furthermore, it must cover a large solid 1 Present address : Department of Physics, College of Science, Rikkyo University, 3-34-1 Nishi-Ikebukuro, Toshima-ku, Tokyo 171-8501, Japan.…”