GALAXY CLUSTER BARYON FRACTIONS REVISITED

We measure the baryons contained in both the stellar and hot-gas components for 12 galaxy clusters and groups at z similar to 0.1 with M = 1-5 x 10(14) M-circle dot. This paper improves upon our previous work through the addition of XMM-Newton X-ray data, enabling measurements of the total mass and masses of each major baryonic component-intracluster medium, intracluster stars, and stars in galaxies-for each system. We recover a mean relation for the stellar mass versus halo mass, M-star proportional to M-500(-0.52 +/- 0.04), that is 1 sigma shallower than in our previous result. We confirm that the partitioning of baryons between the stellar and hot-gas components is a strong function of M-500; the fractions of total mass in stars and X-ray gas within a sphere of radius r(500) scale as f(star) proportional to M-500(-0.45 +/- 0.04) and f(gas) proportional to M-500(0.26 +/- 0.03), respectively. We also confirm that the combination of the brightest cluster galaxy and intracluster stars is an increasingly important contributor to the stellar baryon budget in lower halo masses. Studies that fail to fully account for intracluster stars typically underestimate the normalization of the stellar baryon fraction versus M500 relation by similar to 25%. Our derived stellar baryon fractions are also higher, and the trend with halo mass weaker, than those derived from recent halo occupation distribution and abundance matching analyses. One difference from our previous work is the weak, but statistically significant, dependence here of the total baryon fraction upon halo mass: f(bary) proportional to M-500(0.16 +/- 0.04). For M-500 greater than or similar to 2 x 10(14), the total baryon fractions within r(500) are on average 18% below the universal value from the seven year Wilkinson Microwave Anisotropy Probe (WMAP) analysis, or 7% below for the cosmological parameters from the Planck analysis. In the latter case, the difference between the universal value and cluster baryon fractions is less than the systematic uncertainties associated with the M-500 determinations. The total baryon fractions exhibit significant scatter, particularly at M-500 < 2 x 10(14)M(circle dot) where they range from 60%-90%, or 65%-100%, of the universal value for WMAP7 and Planck, respectively. The ratio of the stellar-to-gas mass within r(500) (M-star/M-gas), a measure of integrated star-formation efficiency, strongly decreases with increasing M-500. This relation is tight, with an implied intrinsic scatter of 12%. The fact that this relation remains tight at low mass implies that the larger scatter in the total baryon fractions at these masses arises from either true scatter in the total baryon content or observational scatter in M-500 rather than late-time physical processes such as redistribution of gas to beyond r(500). If the scatter in the baryon content at low mass is physical, then our results imply that in this mass range, the integrated star-formation efficiency rather than the baryon fraction that is constant at fixed halo mass.