Annu. Rev. Astron. Astrophys. 2002. 40: 539-577 Copyright © 2002 by Annual Reviews. All rights reserved

2.2. Cooling in the Intra Cluster Medium

In order to characterize the role of cooling in the ICM, it is useful to define the cooling time-scale, which for an emission process characterized by a cooling function _c(T), is defined as t_cool = k_BT / (n (T)), n being the number density of gas particles. For a pure bremsstrahlung emission: t_cool 8.5 × 10¹⁰ yr (n / 10^-3 cm^-3)^-1 (T / 10⁸ K)^1/2 (e.g. Sarazin 1988). Therefore, the cooling time in central cluster regions can be shorter than the age of the Universe. A substantial fraction of gas undergoes cooling in these regions, and consequently drops out of the hot diffuse, X-ray emitting phase. Studies with the ROSAT and ASCA satellites indicate that the decrease of the ICM temperature in central regions has been recognized as a widespread feature among fairly relaxed clusters (see Fabian 1994, and references therein). The canonical picture of cooling flows predicted that, as the high-density gas in the cluster core cools down, the lack of pressure support causes external gas to flow in, thus creating a superpositions of many gas phases, each one characterized by a different temperature. Our understanding of the ICM cooling structure is now undergoing a revolution thanks to the much improved spatial and spectral resolution provided by Newton-XMM. Recent observations have shown the absence of metal lines associated with gas at temperature 3 keV (e.g. Peterson et al. 2001, Tamura et al. 2001), in stark contrast with the standard cooling flow prediction for the presence of low-temperature gas (e.g. Böhringer et al. 2002a, Fabian et al. 2001a).

Radiative cooling has been also suggested as an alternative to extra heating to explain the lack of ICM self-similarity (e.g. Bryan 2000, Voit & Bryan 2002). When the recently shocked gas residing in external cluster regions leaves the hot phase and flows in, it increases the central entropy level of the remaining gas. The decreased amount of hot gas in the central regions causes a suppression of the X-ray emission (Pearce et al. 2000, Muanwong et al. 2001). This solution has a number of problems. Cooling in itself is a runaway process, leading to a quite large fraction of gas leaving the hot diffuse phase inside clusters. Analytical arguments and numerical simulations have shown that this fraction can be as large as ~ 50%, whereas observational data indicates that only 10% of the cluster baryons are locked into stars (e.g. Bower et al. 2001, Balogh et al. 2001). This calls for the presence of a feedback mechanisms, such as supernova explosions (e.g. Menci & Cavaliere 2000, Finoguenov et al. 2000, Pipino et al. 2002; Kravtsov & Yepes 2000) or Active Galactic Nuclei (e.g. Valageas & Silk 1999, Wu et al. 2000, Yamada & Fujita 2001), which, given reasonable efficiencies of coupling to the hot ICM, may be able to provide an adequate amount of extra energy to balance overcooling.