Left: The X-ray luminosity function for late-type galaxies at z~0.1. This is estimated using data from an ongoing serendipitous survey that identifies normal galaxies on public XMM-Newton and Chandra fields (look below for details). Right: X-ray source counts for late-type galaxies combining the deepest pencil beam surveys currently available with our wide-angle shallow serendipitous sample. Comparison with models that assume different rates of evolution for the low-z X-ray luminosity function on the left are also plotted. These suggest evolution for the spiral population of the form ~(1+z)³.

Until recently, our knowledge on the X-ray properties of `normal' galaxies (i.e. those dominated by emission of stellar origin) was essentially confined to the local Universe only (about 30Mpc) via targeted observations of a few X-ray bright systems. This situation changed drastically with the advent of the XMM-Newton and Chandra missions. These sensitive X-ray observatories provided, for the first time, X-ray selected normal galaxy samples at cosmologically interesting redshifts (z=0.1-1).

This development opened the way to X-ray evolution studies of the dominant population in the Universe: normal galaxies. In this respect the X-ray wavelengths are a unique diagnostic, providing information complementary to other wavebands (e.g. optical, IR). They are the only tool available to directly probe the hot gas and the X-ray binary population in galaxies and to explore how these components evolve with redshift. To address this issue however, one requires a wide redshift baseline and therefore a combination of deep pencil-beam samples (probing high-z systems) and wide-angle shallow surveys to provide an anchor point at low-z.

Although deep samples are available (e.g. Chandra-Deep Fields), there is a pressing need for a complete low-z sample. We are in the processes of addressing this issue by identifying galaxies using:

• public XMM-Newton and Chandra fields that overlap with the Sloan Digital Sky Survey (SDSS) to exploit the superb and uniform 5-band optical photometry and spectroscopy available in this area (Georgakakis et al. 2004a; Georgantopoulos et al. 2005)
• the XMM-Newton Serendipitous Source Catalogue (1XMM) produced by the XMM-Newton Survey Science Center (SSC). This comprises about 30,000 unique X-ray sources detected in about 500 XMM-Newton pointings. We are combining this unique database with photographic plate material to identify candidate galaxies which we then follow with spectroscopic observations (Georgakakis et al. 2006).

The surveys above cover about 15deg² and probe galaxies at z~0.1. Using this unique sample we have estimated, for the first time, the X-ray luminosity function of early and late-type galaxies separately at z~0.1 (see figure above; Georgantopoulos et al. 2005; Georgakakis et al. 2006). We have further combined our serendipitous wide-angle shallow samples with the Chandra pencil-beam ultra-deep fields (e.g. Alexander et al. 2003), probing galaxies out z~1, to construct the X-ray source counts and to constrain the X-ray evolution of spirals and ellipticals separately. We find that the former evolve as ~(1+z)³ to z~1 while the latter show no evolution to these redshifts (see figure; Georgakakis et al. 2006).