4. Origin of the X-ray emission
To shed light onto the possible emission mechanism(s) giving rise to the jet's X-ray emission, the above data were combined with results from our radio-to-optical studies at resolution (Meisenheimer et al. 1997, Neumann et al. 1997b, Röser & Meisenheimer 1991) to investigate the run of the continuum across the electromagnetic spectrum for the different regions along the jet. We plot the spectra of knots A, B, C, D, and H in Fig. 9, where the flux was multiplied by different factors to disentangle the individual spectra in the graph. The continua are model spectra as described by Heavens & Meisenheimer (1987) and Meisenheimer & Heavens (1986). They result from Fermi acceleration of particles in a strong shock and include the radiation losses in a finite down-stream region. The exponential cut-off reflects the maximum energy obtained by the particles and is well established for knots B to H also by our recent HST WFPC2 data at 300 nm (Jester et al., in preparation). It is evident that in general the X-ray flux level is not a continuation of the radio-optical synchrotron cut-off continuum. Therefore different emission mechanisms have to be discussed for the individual knots.
4.1. Synchrotron radiation from the jet
Only for knot A does the extrapolation of the radio-to-optical continuum approximately meet the X-ray flux level. Extrapolation of the 6cm flux with the low frequency spectral index of knot A ( into the ROSAT range predicts an X-ray flux of 32 nJy, only a factor of roughly two above the observed level. Whereas in Meisenheimer et al. (1996) a lower limit to the cut-off frequency for knot A could be set at about Hz, we now have to increase this value by about a factor of 50 (see Fig. 9). According to standard synchrotron theory we can therefore place a new lower limit to the maximum energy of the relativistic particles in knot A of
where we have used the minimum energy field of 67 nT. For the other knots the primary synchrotron component producing the observed radio-to-optical continuum exhibits an exponential cut-off in the optical/UV-range (Meisenheimer et al. 1996). To maintain the synchrotron scenario also for these regions a second particle population with higher maximum energy has to be invoked 3. These populations are indicated in Fig. 9 connecting to the ROSAT HRI points with an assumed spectral index of , the low-frequency average for knots A to D. The required density of relativistic particles in these hypothetical components decreases outwards along the jet. It is a factor of 5 below the density of the particles producing the observed radio-to-optical continuum for knot B and a factor of 15 and 200 below that in knot C and D, respectively. The small fraction of relativistic particles make it very hard to detect them at e.g.optical wavelengths. However, if this second population is confined to some centres of very effective acceleration on sub-arcsecond scales, we expect to see deviations from the cut-off spectrum in the optical spectral index map derived from our HST R- and U-band data at a resolution of (Jester et al. 2000, in preparation).
4.2. Synchrotron Self-Compton emission (SSC) from the jet
Prime sources for inverse Compton emission are compact regions with high radio photon densities in the jet, for which the hot spot H is the most likely candidate. The size of the emission region and the radio flux originating from it determine the amount of synchrotron-self-Compton emission. We have calculated the inverse Compton emission along the lines given by Blumenthal & Tucker (1970) as follows: The most compact component certainly is the hot spot's acceleration region, where the optically radiating particles are produced. Its contribution to the synchrotron spectrum can be inferred from the low-frequency power-law and the high-frequency cut-off. The intermediate range with the break in the continuum (see Fig. 9) is dominated by the superposition of the down-stream regions, where the relativistic particles already have lost part of their energy. Connecting the cut-off part smoothly with a power-law of index -0.60 we estimate the 5 GHz-flux from the acceleration region itself to be about 0.1 Jy. From the most recent analysis of the hot spot's spectrum by Meisenheimer et al. (1997) we infer a thickness of the emission region close to the internal shock (Mach disk) of 1.4 pc, the width perpendicular to the jet (from our best-resolved radio map) is taken to be 2.2 kpc (circular cross-section assumed). For a minimum energy field of 35 nT (assumed constant over the entire hot spot in the model) we set the density in relativistic particles in this volume to reproduce the above estimated 6cm flux of 0.1 Jy. Integrating this synchrotron emission over the range 10 MHz to GHz (corresponding to Lorentz-factors of 100 to ) produces an inverse Compton flux of about 1 nJy, to within factors of 2 to 4 what is observed (see Table 2). As this is only a rough estimate to check the order of magnitude the discrepancy could well be removed by fine-tuning the parameter assumed (filling factor, geometry).
For the other knots synchrotron-self-Compton emission fails to meet the observed level by orders of magnitude, e.g. for knot A we expect an inverse Compton X-ray flux of only 0.01 nJy. As the photon density of the microwave background radiation is one order of magnitude less than the synchrotron photon density in all knots, its photons cannot account for the X-ray flux via inverse Compton scattering either.
4.3. Thermal bremsstrahlung from the jet
If future high resolution observations fail to reveal locations of Hz in knots B, C, and D, there remains the bremsstrahlung emission from a thermal plasma as a last resort to explain the jet's X-ray emission outside knot A and the hot spot. A plasma at 108 K and with an electron density of 1 cm-3 spread over the volume of a typical jet knot ( from our HST images, Jester et al. 2000, in preparation) would produce an X-ray flux in the ROSAT window corresponding to 0.01 nJy. Even with this unrealistically high electron density this is orders of magnitude below the observed X-ray flux level. Furthermore any sufficiently dense thermal plasma would result in total depolarisation of the jet's radio emission and in a substantial rotation measure. Both are not observed (Conway et al. 1993). So thermal bremsstrahlung is highly unlikely to account for the observed X-ray flux.
© European Southern Observatory (ESO) 2000
Online publication: July 27, 2000