2004MNRAS.347L...7B

In a recent paper, Lehnert & Bremer have photometrically selected a sample of Lyman break galaxies at z > 4.8 from a single VLT/FORS2 pointing and spectroscopically confirmed half of them to be at 4.8 < z < 5.8. To study the properties of such galaxies further, we have photometrically selected a similar sample (V_AB> 28, i_AB< 26.3, i_AB-z_AB> 0) from the Hubble Space Telescope Advanced Camera for Surveys (HST ACS) images of the Chandra Deep Field -South. This selection results in a sample of 44 sources from ∼150 arcmin². We find that such galaxies are often barely resolved in the ACS images, having half-light radii of 0.1-0.3 arcsec (<2 kpc). They show no difference in spatial clustering from sources selected by i_AB< 26.3, i_AB-z_AB> 0, which are generally galaxies of lower redshift. However, their distribution over the field is not uniform and their surface density varies considerably over areas comparable to a single 8-m or HST pointing. A reliable determination of the surface and volume densities of such galaxies requires a sky area considerably larger than the current ACS imaging of this field. No individual Lyman break candidate was detected to a 3σ limit of 6x10^–17 erg/s/cm at 0.5-5 keV by Chandra (a limiting luminosity of below 2x10⁴³ erg/s at z ∼ 5.3). By summing over all positions, we find that the mean source must be undetected at a level at least a factor of 4 times fainter than this. This rules out anything other than a weak active galactic nucleus (AGN) contribution to the emission from these objects; we conclude that luminous AGNs made little contribution to the final stages of re-ionization of the Universe.