Abstract

We present a comparative analysis between X-ray binaries containing  an accreting neutron star (NS) and a white dwarf (WD) in terms of their spectral and timing properties during outbursts based on the {\it RXTE}, ASCA, $Suzaku$ and {\it Beppo}SAX observations. As representatives of such binaries, we used the NS binary 4U~1636--53 and the WD binary SS~Cygni.  We find that the X-ray broad-band energy spectra of both sources can be adequately reproduced by  a composition of a Comptonized component ({CompTB}) and a  {\it Gaussian} iron line component 

during the quiescent and flare states. Using  the electron temperature $T_e$ as a flare tracer,  

we find that the spectral index $\alpha$  is almost constant during X-ray outbursts (regardless of the activity state), but it turns out to be different for  a NS and a WD: 

$\alpha_{NS}=1.00\pm 0.02$ (4U~1636--53) and $\alpha_{WD}=0.85\pm 0.03$  (SS~Cygni), respectively.   We interpret this  index difference using the transition layer (TL) model and associate it to different  boundary conditions at the surfaces for a NS and a WD. 

 In fact, the obtained higher temperature of the NS surface provides its higher reflectivity than that for the WD surface. In addition, the lower temperature of the WD surface ensures that the WD surface absorbs X-rays  forming in the TL 

which leads to different spectral index values.  We also interpret the constant 

index-behavior for NS/WD binaries   as  in the case when the main contribution to the X-ray spectrum is made by the Comptonized component formed in the TL %layer 

between the accretion disk and the NS/WD surface. This effect of the index stability, now  well established for the  NS 4U~1636--53 at $\alpha=1$ and the  WD SS~Cyg at $\alpha=0.85$  was previously discovered in other low-mass X-ray NSs and WDs. The index constancy for a WD and its lower value compared to NSs is due to a lower surface  temperature  of a WD and points out the fundamental difference between NS and WD sources.

Statistics