Diffuse X-ray emission in a deep Chandra image of the Galactic center

We examine the spectrum of diffuse emission detected in the 17' x 17' field around Sgr A* during 625 ks of Chandra observations. The spectrum exhibits He-like and H-like lines from Si, S, Ar, Ca, and Fe that are consistent with an origin in a two-temperature plasma, as well as a prominent low-ionization Fe Kalpha line. The cooler, kT approximate to 0.8 keV plasma differs in surface brightness across the image in the range (0.2-1.8) x 10(-13) ergs cm(-2) s(-1) arcmin(-2) (observed, 2-8 keV). This soft plasma is probably heated by supernovae, along with a small contribution from the winds of massive Wolf-Rayet (W-R) and O stars. The radiative cooling rate of the soft plasma within the inner 20 pc of the Galaxy could be balanced by 1% of the kinetic energy of one supernova every 3 x 10(5) yr. The hotter, kT approximate to 8 keV component is more spatially uniform, with a surface brightness of (1.5-2.6) x 10(-13) ergs cm(-2) s(-1) arcmin(-2) (observed, 2-8 keV). The intensity of the hard plasma is correlated with that of the soft, but they are probably only indirectly related, because neither supernova remnants nor W-R/O stars are observed to produce thermal plasma hotter than kT approximate to 3 keV. Moreover, a kT approximate to 8 keV plasma would be too hot to be bound to the Galactic center and therefore would form a slow wind or fountain of plasma. The energy required to sustain such a freely expanding plasma within the inner 20 pc of the Galaxy is similar to10(40) ergs s(-1). This corresponds to the entire kinetic energy of one supernova every 3000 yr, which is unreasonably high. However, alternative explanations for the kT approximate to 8 keV diffuse emission are equally unsatisfying. The hard X-rays are unlikely to result from undetected point sources, because no known population of stellar objects is numerous enough to account for the observed surface brightness. Neither is there evidence that nonthermal mechanisms for producing the hard emission are operating, as the expected shifts in the line energies and ratios from their collisional-equilibrium values are not observed. We are left to conclude either that there is a significant shortcoming in our understanding of the mechanisms that heat the interstellar medium or that a population of faint (<10(31) ergs s(-1)) hard X-ray sources that is a factor of 10 more numerous than cataclysmic variables remains to be discovered.