In X-ray band, the dominant emission from supernova remnants (SNRs) is the strong emission from the thermal plasma heated by the shock wave of the supernova explosion. However, the additional hard emission represented by a power-law exists in some SNRs. To study such a hard component, the observations sensitive to the higher energy with high energy resolution is effective. In the present work, we observed two shell-like SNRs --- SN 1006 and IC 443 --- with such detectors; SN 1006 was observed with the Ginga and ASCA satellites, and IC 443 was done with the GIS detector on-board the ASCA satellites.
From SN 1006, we detected the strong power-law emission of a photon index of ~3 up to 20 keV, and made clear that the emission is non-thermal origin and confined in two bright shells. The emission can be interpreted as the synchrotron emission from ~100 TeV electrons. The power-law component is confined in the both side of the spherical remnant. At the same time, we detected strong emission lines of highly ionized Ne, Mg, Si and S inside the shell, which had not been detected and is the evidence that the thermal plasma exists. Since the thermal emission was strongly contaminated by the leak of the power-law shells due to the loose point spread function of the X-ray mirror, we could not infer the abundances of the elements precisely; however, the inferred values of much larger than the solar abundances suggest that SN 1006 is still in the free expansion phase or the early stage of the adiabatic expansion phase.
From IC 443, we detected the incomplete semi-circular hard emission surrounding the center-filled thermal emission. The whole hard component in the ASCA field of view cannot explain the Ginga hard component reported by Wang et al. (1992); it suggests that the hard component has larger extent than the present field of view. The hard emission overlaps the EGRET point-source error contour (Thompson et al. 1995, Esposito et al. 1996), and the slope and the flux of the X-ray emission are consistent with the assumption that the X-rays and gamma-rays are emitted from the same electrons by synchrotron emission and the inverse-Compton mechanism, respectively. At the same time, we analyzed the thermal component using the non-equilibrium ionization plasma model part by part, and found that the systematic change of the interstellar absorption suggested by the previous observations does not present.
If the hard components come from the non-thermal electrons accelerated by the shock wave, protons should be accelerated up to the same energy; therefore, it can be interpreted as a support of the cosmic-ray acceleration in SNRs up to ~100 TeV in both free-expansion and adiabatic phases.