The XMM-Newton Spectrum of 4U 1543-475 in the Low/Hard State and a Comment on Accretion Flow Constraints in this Regime

ATel #212; J. M. Miller (CfA), A. C. Fabian (IoA), W. H. G. Lewin (MIT)
on 4 Dec 2003; 21:01 UT
Credential Certification: Jon M. Miller (jmmiller@cfa.harvard.edu)

Tweet

We have analyzed an XMM-Newton/EPIC-pn spectrum of the Galactic black hole 4U 1543-475 in the low/hard state available in the public archive. The 19.5 ksec exposure started on 2002-08-18 UT 12:53:32. The detector was run in "small window" mode with the "thin" optical blocking filter. SAS 5.4.1 was used for data reduction tasks. Source events were extracted in a circle (radius = 24 arcsec) centered on the source position. Background events were extracted in an annulus. Standard event screening was applied according the recipes in the MPE XMM-Newton cookbook. Custom response matrices were generated. After binning to require 15 counts per bin, the spectrum was analyzed on the 0.3-10.0 keV range using XSPEC 11.2.

A simple absorbed power-law model (phabs*powerlaw) gives N_H = 3.8(2) E+21 atoms/cm^2, Gamma = 2.00(5), and K(pl) = 1.02(5) E-3, and chi-squared/dof = 695.1/658 (errors are 90% confidence). This corresponds to an unabsorbed flux of 5.74 E-12 erg/cm^2/s, or a luminosity of 3.9 E+34 erg/s for a distance of 7.5 kpc (Park et al. 2003).

Given that 16,500 counts were recorded (after subtracting the background), the low column density along the line of sight to 4U 1543-475, and the superb low-energy sensitivity of the EPIC-pn camera, this spectrum is among the very best ever obtained from a Galactic black hole at such a low luminosity. We have therefore calculated a number of limits and briefly explored what can and cannot be ruled-out in this regime. The 95% confidence upper limits on the strength of a narrow Fe K-alpha emission line (narrow meaning equivalent to, or less than, the instrument resolution) ranges between 140-260 eV in the 6.40-6.97 keV band. The 95% confidence upper limit on a relativistic Laor Fe K-alpha emission line with r_in fixed at 3 r_g is 200 eV (r_g = GM/c^2).

Adding a multi-color disk (MCD) blackbody component improves the fit at the 3 sigma level. With this model, we measure N_H = 5(1) E+21 atoms/cm^2, kT = 0.19(5) keV, K(MCD) = 130 +/- 120, Gamma = 1.98(7), and K(pl) = 1.0(1) E-3 for chi-squared/dof = 680.7/656. Clearly, the normalization of the MCD component is not well constrained. Taking the source inclination to be zero (note that any non-zero inclination would give larger color radii) and the distance to be 7.5 kpc, this normalization translates to an inner disk color radius of 9km. This is probably unphysical, since 1 r_g ~ 15 km for M_BH = 10 Msun. If we apply Shimura & Takahara's (1995) color conversion factor of 1.7, the inner radius becomes 25 km. Merloni, Fabian, & Ross (2000) have reported that a color conversion factor of f = 3 is possible when the corona dominates; this correction gives an inner disk radius of 77 km, which could correspond to the innermost stable circular orbit (see Park et al. 2003). Reflection from a neutral disk at the innermost stable circular orbit should generate an Fe K line with an equivalent width of approximately 180 eV (George & Fabian 1991), which is within the confidence limits calculated above. This model is a minor statistical improvement over a power-law and does not represent the detection of an inner accretion disk, but it clearly shows that an inner disk cannot be ruled-out statistically.

The fact that this high quality X-ray spectrum obtained in the L_X = 10^(34-35) erg/s range cannot rule-out an inner disk, means that the power-law models often fit to lower quality spectra obtained in this regime (and fit to the spectra of sources at lower-luminosities) cannot be used to infer that an inner disk is absent, as per the predictions of some advection-dominated or jet-dominated accretion flow models. Either family of models may be valid, but the inner geometries assumed in such models are not proven based on power-law X-ray spectra. [Indeed, jet models might be aided by an inner disk, as many models for jet formation depend on a disk.] As geometric constraints are more easily obtained through spectral models than by e.g. break frequencies in power density spectra (which lack a clear physical interpretation), this result indicates (a) that far longer X-ray exposures are needed and (b) that multi-wavelength observations are required to constrain accretion flow models at low X-ray luminosities.