Evidence for X-ray synchrotron emission from simultaneous mid-IR to X-ray observations of a strong Sgr A* flare

Abstract

This paper reports measurements of Sgr A* made with NACO in L' -band (3.80 um), Ks-band (2.12 um) and H-band (1.66 um) and with VISIR in N-band (11.88 um) at the ESO VLT, as well as with XMM-Newton at X-ray (2-10 keV) wavelengths. On 4 April, 2007, a very bright flare was observed from Sgr A* simultaneously at L'-band and X-ray wavelengths. No emission was detected using VISIR. The resulting SED has a blue slope (beta > 0 for nuL_nu ~ nu^beta, consistent with nuL_nu ~ nu^0.4) between 12 micron and 3.8 micron. For the first time our high quality data allow a detailed comparison of infrared and X-ray light curves with a resolution of a few minutes. The IR and X-ray flares are simultaneous to within 3 minutes. However the IR flare lasts significantly longer than the X-ray flare (both before and after the X-ray peak) and prominent substructures in the 3.8 micron light curve are clearly not seen in the X-ray data. From the shortest timescale variations in the L'-band lightcurve we find that the flaring region must be no more than 1.2 R_S in size. The high X-ray to infrared flux ratio, blue nuL_nu slope MIR to L' -band, and the soft nuL_nu spectral index of the X-ray flare together place strong constraints on possible flare emission mechanisms. We find that it is quantitatively difficult to explain this bright X-ray flare with inverse Compton processes. A synchrotron emission scenario from an electron distribution with a cooling break is a more viable scenario.