The LX--T relation and intracluster gas fractions of X-ray clusters

We re-examine the X-ray luminosity-temperature relation using a nearly homogeneous data set of 24 clusters selected for statistically accurate temperature measurements and absence of strong cooling flows. The data exhibit a remarkably tight power-law relation between bolometric luminosity and temperature with a slope 2.88 ± 0.15. With reasonable assumptions regarding cluster structure, we infer an upper limit on fractional variations in the intracluster gas fraction 〈(δ fgas fgas)2〉1/2≤ 15 per cent. A strictly homogeneous Ginga subset of 18 clusters places a more stringent limit of 9 per cent. Imaging data from the literature are employed to determine absolute values of fgas within spheres encompassing density contrasts δc=500 and 200 with respect to the critical density. Comparing binding mass estimates based on the virial theorem (VT) and the hydrostatic β-model (BM), we find a temperature-dependent discrepancy in fgas between the two methods caused by systematic variation of the outer slope parameter β with temperature. Mean values (for H0=50 km s−1 Mpc−1) range from f¯gas=0.10 for cool (T4 keV) clusters using the BM at δc=200. There is evidence that cool clusters have a lower mean gas fraction than hot clusters, but it is not possible to assess the statistical significance of this effect in the present data set. The T dependence of the intracluster medium (ICM) density structure, coupled with the increase of the gas fraction with T in the VT approach, explains the steepening of the LX-T relation. The small variation about the mean gas fraction within this majority subpopulation of clusters presents an important constraint for theories of galaxy formation and supports arguments against an Einstein-de Sitter universe based on the population mean gas fraction and conventional, primordial nucleosynthesis. The apparent trend of lower gas fractions and more extended atmospheres in low-temperature systems is consistent with expectations of models incorporating the effects of galactic winds on the ICM.