Photon Energy (keV)

Fig. 17.11 RXTE PCA spectra of five historical SNRs with hard nonthermal tails probably produced by synchrotron emission from TeV electrons. Each spectrum is characterized by a power law with spectral index of approximately 2 [114]

below the limit for producing X-ray synchrotron emission [13,168]. In contrast, the hard X-ray emission from the thin outer rim is likely to be synchrotron radiation from accelerated electrons. Vink and Laming use the width of the rim to estimate that the magnetic field near the shock front has a strength of 80-160 |G and maximum electron energies of —40-57 TeV [168]. The magnetic field is weak compared with the average, but strong compared to the expectation for shocked ISM, suggesting either the progenitor wind had a high field or post shock field amplification has occurred.

The hard X-ray properties of RCW 86, thought to be the remnant of SN 185, contrast sharply with those of the other historical remnants. In the northeastern quadrant, images reveal a nearly continuous shock structure. Oddly, one segment is dominated by thermal emission and an adjacent one by nonthermal [166]. In the southwestern quadrant, where the remnant is interacting with an ISM structure, a hard emission region is located well behind the shock front, and is correlated with K line emission from low ionization states of Fe [132]. Rho et al. use Chandra data to show that the hard emission is associated with the reverse shock, and that it is likely synchrotron emission and not nonthermal bremsstrahlung. The presence of this emission requires the reverse shock to accelerate electrons to order 50 TeV.

The second group of remnants comprises evolved remnants with shell-like morphology, dominated by nonthermal emission. In addition to SN 1006, prominent members of this group are two remnants discovered in the ROSAT All-Sky Survey, RX J1713.7-3946 and RX J0852.0-4622. A possible fourth member is RXJ 04591+5147. RX J1713.7-3946 shows no evidence whatever for a thermal component [24], and the evidence in RX J0852.0-4622 is marginal [149]. These remnants share other important properties. They have low radio surface brightness, which correlates strongly with the X-ray morphology. TeV y-ray emission from RX J1713.7-3946 (Fig. 17.12) and RX J0852.0-4622 by HESS shows a dramatic correlation with the X-ray surface brightness [1,2].2 The combination of X-ray and radio properties can be explained by expansion in a very low-density medium, in which little interstellar material is swept up. Consequently, the forward shock is not substantially decelerated and a strong reverse shock never forms. The combination of high forward shock velocity and low density results in a long synchrotron lifetime of the X-ray emitting electrons. Interestingly, both RX J1713.7-3946 and RX J0852.0-4649 have stellar remnants unlike the Ia remnant SN 1006, suggesting that the evolution into this class is not dependent on the type of the explosion.

In both groups of remnants, the similarity between the nonthermal X-ray features and the radio morphology indicates that the same emission process is responsible for both. The X-ray flux is substantially below that expected from an extrapolation of the radio spectrum, however. Thus if one spectrum fits both the radio and X-ray emitting particles, it must steepen between the radio and X-ray bands. There are ample mechanisms for producing a turnover: particle losses due to diffusion, the limitations on maximum particle energy from the age of the SNR, and synchrotron

2 The detection of TeV emission from SN 1006, in contrast, is uncertain, with a claimed CANGA-ROO detection at odds with a more sensitive HESS upper limit [3,156]

Fig. 17.12 In the nonthermal X-ray emission dominated remnant RX J1713.7-3946, there is a strong correlation between the TeV y-ray (colors) and X-ray (contours) morphology [1]. The TeV y-ray image is from HESS; the X-ray image is from ASCA

losses by the particles (as the more energetic particles radiate energy faster). Current hard spectra do not allow discrimination among the different mechanisms, but assuming any of them yields approximately the same turnover energy of the particle spectrum. For a sample of bright, young SNRs, the turnover photon energy is typically a few tenths of keV, which for reasonable magnetic fields corresponds to a turnover energy in the electron spectrum of a few TeV [130]. While the maximum electron energy falls far short of the "knee" in the cosmic ray spectrum generally associated with the highest energy Galactic cosmic rays, it is not clear how this energy relates to the maximum energy of protons, the dominant cosmic ray particle.

Indirect evidence for hadronic cosmic ray acceleration in SNR shocks in young remnants comes from comparison between theoretical models of shock structure and sensitive X-ray measurements, enabled by XMM-Newton and Chandra. Hydro-dynamic modeling indicates that the channeling of shock energy into particle acceleration modifies shock structure in three observable ways: first, the postshock compression ratio can become higher than four; second, the distance between the forward shock, the contact discontinuity, and the reverse shock becomes compressed; and, third, the proton temperature is suppressed below would be expected from the shock velocity [27].

For the SMC remnant 1E 0102.2-7219, Hughes et al. found a significant discrepancy between the electron temperature inferred from spectral fitting of Chandra ACIS data (0.4-1.0 keV) and that inferred from their measurement of the shock velocity, assuming any plausible degree of electron heating (2.5-45 keV) [58]. From this they argued that a significant fraction of the shock energy is channeled elsewhere than heating the postshock ions and electrons. They concluded that this channel is the acceleration of cosmic rays. A similar analysis yielded the same conclusion for Tycho [61].

Also for Tycho, Warren et al. used Chandra data to measure the average distances between the forward and reverse shocks and the contact discontinuity [171]. The distance between the forward shock and contact discontinuity is smaller than hydro-dynamic models predict, and can be explained if acceleration of ions is occurring in the forward shock. The nonthermal spectrum arising in the forward shock supports this conclusion. In contrast, both the distance between the contact discontinuity and the reverse shock and the thermal spectrum indicate that particle acceleration is not occurring in the reverse shock.

There is no firm evidence of hard, nonthermal emission from shock-accelerated electrons in older, thermal emission dominated SNRs. While neither Chandra nor XMM-Newton has fully mapped any of the prominent evolved Galactic remnants, the carefully studied regions of the shells of remnants such as Puppis A and the Cygnus Loop show no excess hard emission.

0 0

Post a comment