First Results from the X-Ray and Optical Survey of the Chandra Deep Field South^*

We present our first results from 120 ks of X-ray observations obtained with the Advanced CCD Imaging Spectrometer on the Chandra X-Ray Observatory. The field of the two combined exposures is 0.096 deg² and the detection limit is to a S/N of 2 (corresponding to ~7 net counts). We reach a flux of 2 × 10^-16 erg s^-1 cm^-2 in the 0.5-2 keV soft band and 2 × 10^-15 erg s^-1 cm^-2 in the 2-10 keV hard band. Our combined sample has 144 soft sources and 91 hard sources, for a total of 159 sources. Fifteen sources are detected only in the hard band, and 68 only in the soft band. For the optical identification, we carried out a survey in VRI with the FORS-1 imaging spectrometer on the Antu telescope (UT-1 at VLT) complete to R ≤ 26. This data set was complemented with data from the ESO Imaging Survey (EIS) in the UBJK bands and the ESO Wide Field Imager Survey (WFI) in the B band. The positional accuracy of the X-ray detections is of the order of 1'' in the central 6'. Optical identifications are found for ≃90% of the sources. Optical spectra have been obtained for 12 objects. We obtain the cumulative spectra of the faint and bright X-ray sources in the sample and also the hardness ratios of individual sources. A power-law fit in the range 2-10 keV using the Galactic value of N_H ≃ 8 × 10¹⁹ cm^-2 yields a photon index of Γ = 1.70 ± 0.12 and 1.35 ± 0.20 (errors at 90% confidence level) for the bright and faint samples, respectively, showing a flattening of the spectrum at lower fluxes. Hardness ratio is given as a function of X-ray flux and confirms this result. The spectrum of our sources is approaching the spectrum of the X-ray background (XRB) in the hard band, which has an effective Γ = 1.4. Correlation function analysis for the angular distribution of the sources indicates that they are significantly clustered on scales as large as 100''. The scale dependence of the correlation function is a power law with index γ ~ 2, consistent with that of the galaxy distribution in the local universe. Consequently, the discrete sources detected by deep Chandra-pointed observations can be used as powerful tracers of the large-scale structure at high redshift. We discuss the log N- log S relationship and the discrete source contribution to the integrated X-ray sky flux. In the soft band, the sources detected in the field at fluxes below 10^-15 erg s^-1 cm^-2 contribute (4.0 ± 0.3) × 10^-12 erg cm^-2 s^-1 deg^-2 to the total XRB. The flux resolved in the hard band down to the flux limit of 2 × 10^-15 erg s^-1 cm^-2 contributes (1.05 ± 0.2) × 10^-11 erg cm^-2 s^-1 deg^-2. Once the contribution from the bright counts resolved by ASCA is included, the total resolved XRB amounts to 1.3 × 10^-11 erg cm^-2 s^-1 deg^-2, which is 60%-80% of the total measured background. This result confirms that the XRB is due to the integrated contribution of discrete sources, but shows that there is still a relevant fraction (at least 20%) of the hard XRB to be resolved at fluxes below 10^-15 erg s^-1 cm^-2. We discuss the X-ray flux versus R magnitude relation for the identified sources. We find that ≃10% of the sources in our sample are not immediately identifiable at R > 26. For these sources, S_X/S_opt > 15, whereas most of the ROSAT and Chandra sources have S_X/S_opt < 10. We have also found a population of objects with unusually low S_X/S_opt that are identified as galaxies. The R-K versus R color diagram shows that the Chandra sources continue the trend seen by ROSAT. For our 12 spectroscopically studied objects with redshifts, we observe four QSOs, five Seyfert 2 galaxies, one elliptical, and two interacting galaxies. We compare L_X versus z obtained with these measurements and show that Chandra is achieving the predicted sensitivity.