Astron. Astrophys. 344, 857-867 (1999)

1. Introduction

It has long been recognized that the hard X-ray spectra of Seyfert galaxies (at E 20 keV, photo-electric absorption becomes inefficient) are a powerful tool for extracting information on the intrinsic source emission properties (i.e. luminosity and spectral shape). Such measurements are fundamental for the understanding of the emission mechanisms operating in these objects and, since they allow a direct comparison between different classes of AGNs (e.g. Seyfert 1 versus Seyfert 2 galaxies), for testing unified models (Antonucci 1993). At lower energies ( 0.1-20 keV), one can obtain information on column densities, ionization and abundances of the surrounding matter. This issue is important in testing the geometry and in particular the existence of molecular tori with 10²⁴ cm^-2 around Seyfert galaxies whose presence is essential for AGN unified models and synthesis models of the X-ray background.

In the few OSSE data available for Seyfert galaxies, the steep high-energy spectra, the evidence of a high-energy cutoff (e.g. Maisack et al. 1993, Zdziarski et al. 1995, Grandi et al. 1998), and the absence of an annihilation line favour thermal Comptonization models (Haardt 1997) in contrast with the predictions from non-thermal pair models (see Svensson 1994 for a review). More precise measurements by BeppoSAX have confirmed the presence of high-energy cutoffs in some Seyfert 1 galaxies (Piro et al. 1999) though only lower-limits have been found in others (Perola et al. 1999). To date, however, precise measurements are still lacking for Seyfert 2 galaxies with only a few attempted (Zdziarski et al. 1995, Bassani et al. 1995, Weaver et al. 1998), and, as such, fundamental to be performed.

Mkn 3 (z=0.0135) is one of the small sample of Seyfert 2 galaxies which show broad emission lines in polarized light suggesting the presence of a "hidden Seyfert 1 nucleus" (e.g., Miller & Goodrich 1990 and Tran 1995). Other evidence in favor of heavy obscuration are: the discovery of a biconical extended narrow line region (Pogge & De Robertis 1993), the low flux of ionizing photons inferred from the directly observed UV continuum compared to the ionizing photons required to produce the observed H emission (Haniff et al. 1988, Wilson et al. 1988), the low / ratio (0.14; Bassani et al. 1998) and the observation of heavy X-ray obscuration. Indeed, Ginga data have shown that the spectrum of Mkn 3 is flat ( 1.3), absorbed by a column density of 6 10²³ cm^-2 and has a strong iron line with equivalent width (EW) 500 eV (Awaki et al. 1991). Measurements at low X-ray energies revealed the presence of a soft excess and also indicated the presence of prominent soft X-ray emission lines (Kruper et al. 1990, Turner et al. 1993, Iwasawa et al. 1994 (I94 hereinafter)). The ASCA observation revealed that the Fe line emission decreased by a factor of 3 in response to a flux decline by a factor of 6 (I94). ASCA also resolved the iron line in a 6.4 keV component of EW 900 eV and a 6.7 keV one of EW 190 eV (I94). The analysis of the same data set alone, or in combination with measurements from other instruments, indicated the difficulty encountered in interpreting unequivocally the 2-10 keV spectrum (Turner et al. 1997a,b and Griffiths et al. 1998, G98 hereinafter). It also highlighted the need for a broad-band coverage to better interpret the observed emission. At high energies, Mkn 3 was detected by OSSE for the first time in March 1994 at a flux level of 2.8 10^{- 11} erg cm^-2 s^-1 in the 50-150 keV energy band (Johnson, private communication).

Here we present the observation of Mkn 3 by BeppoSAX. The results obtained highlight the potentialities of using BeppoSAX in broad-band X-ray spectroscopy studies of active galactic nuclei, in particular in the case of highly absorbed sources like Seyfert 2 galaxies. Throughout the analysis, we use = 50 km s^-1 Mpc^-1 and =0.

© European Southern Observatory (ESO) 1999

Online publication: March 29, 1999
helpdesk.link@springer.de