NGC 1052 is the first "type 2" LINER, where broad lines were discovered in spectropolarimetric measurement (Barth et al. 1999). One may therefore expect that its X-ray spectrum resembles that of Seyfert 2 galaxies, where the luminous nuclear emission is photoelectrically absorbed by matter with a substantial column density. The observation performed with BeppoSAX demonstrates that this is indeed the case, thanks to the first measure of the broadband (0.1-100 keV) X-ray spectrum of this object.
Even more intriguing is the nature of the accretion flow in this AGN. The nuclear spectrum is rather flat (), as the ASCA observation had already suggested (G99; Weaver et al. 1999). The best-fit value is inconsistent at 2 level with the distribution of spectral indices observed in Seyfert 1s (; ; Nandra et al. 1997; Reynolds 1997) and radio-quiet quasars (; ; Reeves & Turner 2000). It is closer to the distribution of spectral indices in radio-loud quasars (; ; Reeves & Turner 2000). However, NGC 1052 radio-loudness, according to the Wilkes & Elvis (1987) definition, is 1.1, therefore pointing to a radio-quiet or, at most, a borderline object. There are some Seyfert 2s, whose ASCA spectral indices (Turner et al. 1997) are as flat as those measured in NGC 1052. Recently, Malaguti et al. (1999) pointed out that at least in one case (NGC 2110) this could be due to an incorrect modeling of a complex photoelectric absorber in the ASCA bandpass. To verify this hypothesis, we have extracted from the BeppoSAX archive the PDS spectra of the "flat" Seyfert 2s of the Turner et al. (1997) sample. In Fig. 3 we compare the versus 2-10 keV (extrapolated) flux iso- contours for NGC 1052 and the "flat" Seyfert 2s (the results for NGC 5252 are not shown, because the PDS detection is too weak). In all cases the intrinsic spectral index as measured by the PDS is significantly steeper than in ASCA. This suggests that the nature of the accretion flow in the nucleus of NGC 1052 is qualitatively different from that normally at work in bright nearby AGN.
The NGC 1052 high-energy spectrum is can be also well approximated by a thermal bremsstrahlung with keV, in agreement with the expectations of the ADAF scenario. No estimate of the nuclear black hole mass is available for NGC 1052. If we use the bulge B magnitude (Ho et al. 1997) versus black hole mass relation of Magorrian et al. (1998), we obtain . The bolometric luminosity of the ADAF component extrapolated from the BeppoSAX measurement is erg s-1, suggesting an accretion rate . This value is significantly lower than the critical , below which the onset of the ADAF regime would occur (Narayan & Yi 1995).
Critical diagnostic tools for the existence of an ADAF are the radio-to-X-ray Spectral Energy Distribution (SED) and luminosity ratio (Di Matteo et al. 2000). In Fig. 4 we show the NGC 1052 radio/X-ray SED. The radio data points are taken from the NED archive. We observe a good qualitative agreement between the shape of the SED in NGC 1052 and in the "candidate-ADAF" elliptical galaxies of the Allen et al. (2000) sample. In Table 2 we compare the ratios between the powers at 6 keV and 5 GHz in these galaxies and in NGC 1052. In the latter this ratio is well within the (rather narrow) range observed in the former objects, confirming the qualitative similarity. For comparison, the value of this ratio is typically in radio-quiet quasars (Elvis et al. 1994)
It is also interesting to compare the X-ray with the IR photometry of Becklin et al. (1982). Their 1.2µm flux in the innermost 2" (300 pc) - 26 mJy - corresponds to a ratio of about 15. Such a high value is typical of Seyfert 2 galaxies (Mass-Hesse et al. 1995). The IR emission is therefore likely to be dominated by heated dust, totally masking the ADAF IR intrinsic emission and therefore preventing us from checking the presence of a wind in the ADAF scenario (Di Matteo et al. 2000).
The luminosity of the broad component of is erg s-1 (Ho et al. 1997; Barth et al. 1999). An ADAF cannot provide the corresponding amount of (mainly UV) ionizing photons. This points to a flow with a small sonic radius , consistent with models where the viscosity parameter is ( a few ; Narayan et al. 1997). Outside this radius the flow would smoothly connect to a standard thin accretion disk.
The BeppoSAX data alone cannot rule out the possibility that a substantial contribution to the hard X-ray emission comes from a nuclear relativistic jet. Chandra will have enough spatial resolution to image X-ray structures on scales as small as 100 pc. However, it is to be noted that the ROSAT HRI images of NGC 1052 unveiled only a soft X-ray extended emission on a much wider scale (15 kpc) than that of the radio structures (G99), therefore probably due to the emission of diffuse intergalactic gas. The contribution of this component to the hard X-ray emission is most likely negligible.
© European Southern Observatory (ESO) 2000
Online publication: January 29, 2001