5 X-ray-to-optical flux ratios

X-ray-to-optical flux ratios can yield important information on the nature of X-ray sources (Maccacaro et al. 1988). A value of $-1<\log (\frac{f_X}{f_R}) <1$ is a clear sign of AGN activity since normal galaxies and stars have usually lower X-ray-to-optical flux ratios, $\log (\frac{f_X}{f_R}) <-2$. In Fig. 6, we plot the X-ray flux in [0.5-2.0] (a) and [2.0-10] (b) keV bands as a function of the R magnitude for the 98 sources of the sample. A large fraction of the sources spans the typical X-ray-to-optical flux ratio of AGN. While in the soft band (Fig. 6a) the Type-2 AGN and the unidentified sources are confined at the lower fluxes of our sample, in the hard band (Fig. 6b), where the effect of the absorption is weaker, the range of fluxes covered by these sources is almost the same as that of the Type-1 AGN sample. We also note that $32 \%$ of the sources are confined in a region with $\log (\frac{f_X[2-10]}{f_R}) >1$. Among the sources with a high X-ray-to-optical flux ratio, $\sim$85% are heavily absorbed ( $\log(N_{\rm H}) > 21.5$) and $\sim$60% are EROs. Their optical classification is still largely incomplete due to their faintness: two are Type-1 AGN, four are Type-2 AGN, one is an extended source and 13 are unidentified.

At the current flux limit of our complete sample, the population of objects with very low X-ray-to-optical flux ratio is largely missing. Such a population was unveiled by the Chandra deep surveys (Giacconi et al. 2001; Hornschemeier et al. 2001) and found to comprise normal galaxies and low-luminosity AGN (with $L_{\rm X} < 10^{42}$ erg s^-1 in the [0.5-10] keV energy band).

There are examples in the literature of "X-ray bright'' objects ($L_{\rm X}$[2-10] >10⁴¹ erg s^-1) but without any obvious signature of nuclear activity in the optical spectra (Griffiths et al. 1995; Comastri et al. 2002 and references therein). A heavily obscured AGN is among the most likely explanations. In our sample there are two sources ($\char93 60$ and $\char93 92$) optically classified as normal galaxies (their optical spectra show emission lines that declare them as star-forming galaxies) which are however X-ray luminous: $L_{\rm X}[2{-}10]=3.7 \times 10^{41}$ and $L_{\rm X}[2{-}10]=1.9 \times 10^{42}$ erg s^-1respectively. Their X-ray spectra are clearly absorbed ( $N_{\rm H}=2^{+2.8}_{-1.6}\times10^{23}$ and $N_{\rm H}=9^{+3.2}_{-1.4}\times10^{20}$  ${\rm cm}^{-2}$) reinforcing the evidence that they contain an obscured AGN. More examples of this class of objects are expected at the completion of the optical identification of the newly detected XMM-Newton sources.

Figure 9: X-ray luminosity in the [2-10] keV rest-frame band versus redshift. Symbols are as in Fig. 2. The dashed line shows the current limit in the X-ray flux of our sample.

In Fig. 7, we plot the X-ray flux in [0.5-2.0] (a) and [2.0-10] (b) keV bands as a function of $K^\prime$ magnitudes. In these diagrams, there is a strong overlap between Type-2 AGN and unidentified sources and the Type-1 AGN population, more prounonced in Fig. 7b. The overlap in magnitudes is likely due mainly to a K-correction effect (see also Fig. 4b), whereas in the X-ray hard band the effect of the absorption is weaker (see also Fig. 6b).

Finally, in the soft band (Figs. 6a and 7a) the fraction of absorbed objects increases significantly as the flux decreases. This inevitably leads, in a flux limited sample, to a bias in the $N_{\rm H}$ distribution at high value of $N_{\rm H}$ (see Fig. 2).