Intergalactic stars in z∼ 0.25 galaxy clusters: systematic properties from stacking of Sloan Digital Sky Survey imaging data

We analyse the spatial distribution and colour of the intracluster light (ICL) in 683 clusters of galaxies between z = 0.2 and 0.3, selected from ∼1500 deg 2 of the first data release of the Sloan Digital Sky Survey (SDSS-DR1). Surface photometry in the g-, r- and i-bands is conducted on stacked images of the clusters, after rescaling them to the same metric size and masking out resolved sources. We are able to trace the average surface brightness (SB) profile of the ICL out to 700 kpc, where it is less than 10 -4 of the mean surface brightness of the dark night sky. The ICL appears as a clear surface brightness excess with respect to an inner R 1/4 profile which characterizes the mean profile of the brightest cluster galaxies (BCGs). The surface brightness of the ICL ranges from 27.5 mag arcsec -2 at 100 kpc to ∼32 mag arcsec -2 at 700 kpc in the observed r-band. This corresponds to a value of SB in the range 26.5-31 in the rest-frame g-band. We find that, on average, the ICL contributes only a small fraction of the total optical emission in a cluster. Within a fixed metric aperture of 500 kpc, this fraction is 10.9 ± 5.0 per cent for our clusters. A further 21.9 ± 3.0 per cent is contributed on average by the BCG itself. The radial distribution of the ICL is more centrally concentrated than that of the cluster galaxies, but the colours of the two components are identical within the statistical uncertainties. In the mean the ICL is aligned with and more flattened than the BCG itself. This alignment is substantially stronger than that of the cluster light as a whole. We find the surface brightness of the ICL to correlate both with BCG luminosity and with cluster richness, while the fraction of the total light in the ICL is almost independent of these quantities. These results support the idea that the ICL is produced by stripping and disruption of galaxies as they pass through the central regions of clusters. Our measurements of the diffuse light also constrain the faint-end slope of the cluster luminosity function. Slopes a < -1.35 would imply more light from undetected galaxies than is observed in the diffuse component.