2013MNRAS.428.3462H

We exploit XMM-Newton archival data in a study of the extended X-ray emission emanating from Galactic Centre (GC) region. XMM-Newton EPIC-pn and EPIC-MOS observations, with a total exposure time approaching 0.5 and 1 Ms, respectively, were used to create mosaicked images of a 100 pc x 100 pc region centred on Sgr A* in four bands covering the 2-10 keV energy range. We have also constructed a set of narrow-band images corresponding to the neutral iron fluorescence line (Fei Kα) at 6.4 keV and the K-shell lines at 6.7 and 6.9 keV from helium-like (Fexxv Kα) and hydrogenic (Fexxvi Lyα) iron ions. We use a combination of spatial and spectral information to decompose the GC X-ray emission into three distinct components. These comprise: first the emission from hard X-ray emitting unresolved point sources; secondly the reflected continuum and fluorescent line emission from dense molecular material and, thirdly, the soft diffuse emission from thermal plasma in the temperature range kT ~ 0.8-1.5 keV.

We show that the unresolved-source component accounts for the bulk of the 6.7- and 6.9-keV line emission and also makes a major contribution to both the 6.4-keV line emission and the 7.2-10 keV continuum flux. We fit the observed X-ray surface-brightness distribution with an empirical 2D model, which we then compare with a prediction based on an NIR-derived 3D mass model for the old stellar population in the GC. The X-ray surface brightness falls-off more rapidly with angular offset from Sgr A* than the mass-model prediction. One interpretation is that the 2-10 keV X-ray emissivity increases from ~ 5x1027ergs- 1M- 1☉ at 20 arcmin up to almost twice this value at 2 arcmin. Alternatively, some refinement of the mass model may be required, although it is unclear whether this applies to the Nuclear Stellar Cluster, the Nuclear Stellar Disc or a combination of both components.

The unresolved hard X-ray emitting source population, on the basis of spectral comparisons, is most likely dominated by magnetic cataclysmic variables, primarily intermediate polars. We use the X-ray observations to set constraints on the number density of such sources at angular offsets between 2 and 20 arcmin from Sgr A* (projected distances at the GC of 4.6-46 pc). Our analysis does not support the conjecture that a significant fraction of the hard X-ray emission from the GC originates in very hot ( ∼ 7.5 keV) diffuse thermal plasma.