The XMM-Newton serendipitous survey - III. The AXIS X-ray source counts and angular clustering

Context. Recent results have revised upwards the total X-ray background (XRB) intensity below similar to 10 keV, therefore an accurate determination of the source counts is needed. There are also contradictory results on the clustering of X-ray selected sources. Aims. We have studied the X-ray source counts in four energy bands: soft (0.5-2 keV), hard (2-10 keV), XID (0.5-4.5 keV) and ultra-hard (4.5-7.5 keV) in order to evaluate the contribution of sources at different fluxes to the X-ray background. We have also studied the angular clustering of X-ray sources in those bands. Methods. AXIS ( An XMM-Newton International Survey) is a survey of 36 high Galactic latitude XMM-Newton observations covering 4.8 deg(2) in the Northern sky and containing 1433 serendipitous X-ray sources detected with 5-s significance. This survey has similar depth to the XMM- Newton catalogues and therefore serves as a pathfinder to explore their possibilities. We have combined this survey with shallower and deeper surveys, and fitted the source counts with a Maximum Likelihood technique. Using only AXIS sources we have studied the angular correlation using a novel robust technique. Results. Our source counts results are compatible with most previous samples in the soft, XID, ultra-hard and hard bands. We have improved on previous results in the hard band. The fractions of the XRB resolved in the surveys used in this work are 87%, 85%, 60% and 25% in the soft, hard, XID and ultra-hard bands, respectively. Extrapolation of our source counts to zero flux is not sufficient to saturate the XRB intensity. Only galaxies and/or absorbed AGN could contribute the remaining unresolved XRB intensity. Our results are compatible, within the errors, with recent revisions of the XRB intensity in the soft and hard bands. The maximum fractional contribution to the XRB comes from fluxes within about a decade of the break in the source counts (similar to 10(-14) cgs), reaching similar to 50% of the total in the soft and hard bands. Angular clustering (widely distributed over the sky and not confined to a few deep fields) is detected at 99-99.9% significance in the soft and XID bands, with no detection in the hard and ultra-hard band (probably due to the smaller number of sources). We cannot confirm the detection of significantly stronger clustering in the hard-spectrum hard sources. Conclusions. Medium depth surveys such as AXIS are essential to determine the evolution of the X-ray emission in the Universe below 10 keV.