4 X-ray spectroscopy

The Beppo-SAX concentrator/spectrometer system consists of four separated concentrator mirrors, three of them covering the 1.6- 10 keV range (Medium Energy Concentrator Spectrometer, or MECS) and the fourth extending to lower energies down to 0.1 keV (Low Energy Concentrator Spectrometer, or LECS). The concentrators are designed to have a large effective area around the iron K$\alpha$ line complex: 150 and 50 cm² for MECS and LECS, at 6 keV. Also, Beppo-SAX is able to provide spatially resolved spectra: its energy and angular resolution are $\Delta E/E=8\%$ at 6 keV and $\theta_{{\rm FWHM}} \approx 40''$, respectively.

  

Table 3: 1SAXJ0027.1-1926

$$\begin{tabular}{ccccccc} \hline \hline \multicolumn{7}{c}{}\\... ... & -- & -- & 61 & 62.55 & 1.06 \\ \hline \hline \end{tabular}$$

  

Table 4: 1SAXJ0027.2-1930

$$\begin{tabular}{ccccccc} \hline \hline \multicolumn{7}{c}{}\\... ... & -- & -- & 19 & 22.05 & 1.30 \\ \hline \hline \end{tabular}$$

  

Table 5: 1SAXJ0027.0-1928

$$\begin{tabular}{ccccccc} \hline \hline \multicolumn{7}{c}{}\\... ...$ & -- & -- & 13 & 14.98 & 1.36 \\ \hline \hline \end{tabular}$$

In order to obtain the emission weighted spectral information of the three main sources in the A33 field, we have extracted the photons from circular regions drawn around each source (see Fig. 1). The extraction radius, smaller than the suggested 4 arcmin radius region since the sources are separated by a small angular distance, might introduce a systematic uncertainty. We have used the appropriate Ancillary Response File to correct for this effect. We fitted the source spectra using both a Raymond-Smith code (1977; hereafter RS) or a MEKAL code (Mewe et al. 1995) to model the thermal intracluster gas emissivity and a simple absorbed power-law, non-thermal model. Background spectra have been extracted from library blank-sky images in the same circular regions as the sources.

a) 1SAXJ0027.1-1926

The spectrum of the brightest source in the field was extracted, both for the LECS and the MECS instruments, from a circular region of 2 arcmin radius centered on the X-ray position of Table 1. The combined LECS-MECS spectrum is shown in Fig. 4: we do not observe any low energy excess absorption in the spectrum, thus we keep $N_{\rm H}$ fixed at the galactic value of 1.86 10²⁰ cm^-2 (Dickey & Lockmann 1990) relative to the source position.

The best fit spectral parameters for the MECS spectrum are listed in Table 3 together with their uncertainties at $68.3 \%$ (and $90 \%$ in parentheses) confidence level. We use 605 source photons in this spectral fit.

Within 2 arcmin from its center, the source has a flux of $F_{2-10~{\rm keV}} = (4.20\pm0.32)~10^{-13}$ erg s^-1 cm^-2, evaluated using the MEKAL best fit parameters. The other models give similar fluxes. This flux is also consistent, within the errors, with the flux of the X-ray source 1RXSJ002709.5-192616 in the ROSAT band.

The optical magnitude of the M-star if $m_V \approx 19$. Assuming that the X-ray flux of the M-star contributes to $50 \%$ of the total flux of 1SAXJ0027.1-1926, we obtain $F_{2-10}/F_{V} \approx 2.$ in the 2-10 keV band and $F_{0.3-3.5}/F_{V} \approx 6$ in the 0.3-3.5 energy band (assuming a thermal emission at T=1 keV). This ratio is almost one order of magnitude higher than the values of F_0.3-3.5/F_V for X-ray selected stars in the EMSS (see Fig. 1 in Maccacaro et al. 1988). This means that the contribution of the M-star to the X-ray flux of 1SAXJ0027.1-1926 should be $\ \raise -2.truept\hbox{\rlap{\hbox{$\sim$ }}\raise5.truept \hbox{$<$ }\ }8 \%$ to be consistent with the values of $F_{\rm X}/F_{V}$ for normal stars. If this is the case, then more than half of the X-ray emission of 1SAXJ0027.1-1926 is due to the AGN (listed as A in Table 2) at z=0.2274 with a luminosity $L_{2-10~{\rm keV}} \ \raise -2.truept\hbox{\rlap{\hbox{$\sim$ }}\raise5.truept \hbox{$<$ }\ }4.5~10^{43}$ erg s^-1. Otherwise, the source 1SAXJ0027.1-1926 should result from the blend of the AGN and of a different unknown X-ray source.

b) 1SAXJ0027.2-1930

The average spectrum of 1SAXJ0027.2-1930 was extracted from a circular region of 2 arcmin radius centered on the X-ray position of Table 1 (see also Fig. 5). In this region there is a clear excess of galaxies (see Fig. 3) which is $\sim 1.5$ arcmin away from the Abell catalog position of A33. Fitting the spectrum (which contains 140 source photons) with a RS thermal model with temperature, abundance and redshift as free parameters, the fit gives $\chi^2_{{\rm red}} = 1.22$. Fixing the value of $N_{\rm H}$ to the galactic value (1.86 10²⁰ cm^-2) we obtain an average temperature $T = 3.1\pm$ 0.9 keV and a redshift $z_{\rm X} = 0.72 \pm 0.04$. The abundance is only marginally constrained at Fe/H = 0.98 $\pm$ 0.71 of the solar value. However, the fit results are mainly due to a marginally significant spectral feature at $E \sim 4$ keV.

Therefore, we fixed the redshift of the X-ray source at z = 0.2409, as measured from the optical spectra (see Sect. 3), and we fitted the spectrum again, fixing the abundance to a value Fe/H = 0.3 solar. The results of the fit are shown in Table 4. Uncertainties in the temperature of 1SAXJ0027.2-1930 are given at $68.3 \%$ (and $90 \%$ in parentheses) confidence level. The low count rate of the source does not allow a more accurate description of the X-ray emission.

Assuming the MEKAL best fit parameters we obtain an integrated flux of $F_{2-10~{\rm keV}} = (2.4\pm0.3)~10^{-13}$ erg s^-1 cm^-2 in the 2 arcmin radius extraction region (which corresponds to a linear size of $\approx 1 ~h^{-1}_{50}$ Mpc). The other models give consistent fluxes. At the redshift of the cluster this flux corresponds to a luminosity $L_{2-10~{\rm keV}} = (7.7\pm0.9)~10^{43}~h^{-2}_{50}$ erg s^-1 and to a bolometric luminosity $L_{{\rm bol}}=(2.2\pm0.3)~10^{44}~h^{-2}_{50}$ erg s^-1.

c) 1SAXJ0027.0-1928

We extracted the spectrum of 1SAXJ0027.0-1928 from a circular region of 1arcmin radius centered on the X-ray position of Table 1 (see Fig. 6). Results of the fit are shown in Table 5 (uncertainties on the best fit values are given here at 68.3% confidence level. Note that the spectrum of this source contains 90 source photons). Assuming an absorbed power-law non-thermal model, we derived a flux of $F_{2-10~{\rm keV}} = (4.74~\pm ~0.8)~10^{-14}$ erg s^-1 cm^-2. There are two galaxies, a spiral (C) at the same z of A33, and an elliptical (D) in the region for which we took an optical spectrum.

The identification of the source is not certain at the moment. Assuming that the galaxy D at z = 0.2863 is the X-ray emitter, its X-ray luminosity would be $L_{2-10~{\rm keV}} = 1.9~10^{43}$ erg s^-1. Such an X-ray luminosity seems to be sensibly higher than the X-ray luminosity of a "normal'' galaxy. The possibility that the X-ray emission is due to a more distant, unidentified object cannot be excluded at present.