Dark Energy, Dark Matter, and Cosmology
From Cooling Flows to "Cool Cores" in Massive Galaxy Clusters
Clusters of galaxies have very deep potential wells with virial velocities equivalent to temperatures of 107-108 K. Gravitationally driven processes like accretion shocks and adiabatic compression should therefore heat gas accumulating within a cluster to X-ray-emitting temperatures. Spectroscopic X-ray observations show that most of a cluster's gas is indeed near the virial temperature Tvirial = 7.1 × 107 σ10002 K or 6.2 σ10002 keV, where σ1000 is the line-of-sight velocity dispersion in units of 1,000 km s-1 (Sarazin 1986).
In roughly 30% of all clusters we find a significant fraction of the baryons at a temperature significantly less than the virial temperature (see Donahue and Voit 2004 for a recent review). This gas would be considered quite hot in other astrophysical context, but in order to be cooler than the virial temperature today it must have either avoided the gravitational heating experienced by the rest of the cluster or cooled significantly after entering the cluster.
Because this gas is dense enough to radiate an energy equivalent to its thermal energy in less than a Hubble time, astronomers have long speculated that it cools and contracts, forming a "cooling flow" of condensing gas in the cluster core (e.g., Fabian and Nulsen 1977). All the stars in a cluster's galaxies are made of such gas, implying that at least some cooling and condensation must have occurred during the assembly of the cluster.
The primary question concerning cool gas in clusters is whether these pieces – cool X-ray gas, stars, nebulae, molecular clouds – all fit together into a single, coherent picture of condensation and star formation. If so, then studies of cluster cores may have much to teach us about the processes that govern galaxy formation. What we have learned in the Chandra and XMM-Newton era is that simple cluster cooling flows do not occur (Molendi and Pizzolato 2001; Peterson et al. 2001, 2003). High-resolution spectroscopic observations with XMM-Newton and Chandra are now revealing a deficit of emission from gas below Tvirial/3, which is not predicted by cooling flow models.
XMM-Newton RGS data for X-ray bright cluster 2A0335+096. The OVIII Lyα and Lyβ lines were detected, but no Fe XVII is apparent at the expected wavelengths of 15.014 Å or 16.78 Å. (Data courtesy J. Peterson; Peterson et al. 2003)
Observations from the present generation of X-ray telescopes have challenged the simple cooling flow hypothesis, but what will take its place? Star formation, radio jets, and conduction may all have important roles to play in the development of cluster cores. Looking for hallmarks of episodic feedback from both AGNs and supernovae may be more fruitful.
If feedback is episodic, then the state of the central intracluster medium should be closely related to other goings-on in the cluster core. Thus, it would be interesting to test whether the Tvirial/3 scaling of the minimum plasma temperature apparent in the early sample of XMM-Newton clusters from Peterson et al. (2003) holds for a large sample of cool core clusters with various levels of core activity. For Constellation-X this means taking the ambitious high-spectral resolution observations of Peterson et al. (2003) to a much larger sample (requiring much larger mirror-collecting area). We will answer the questions of: How do the X-ray emission-line spectra of clusters with radio-loud nuclei differ from those of clusters with radio-quiet nuclei? Are there any correlations between X-ray line emission and the presence of obvious star formation or emission-line nebulae?
Episodic heating also leads to a predictable pattern in the evolution of the core entropy distribution (Kaiser and Binney 2003). Thus, studying the core entropy distributions of a large sample of clusters may reveal a tell-tale pattern of entropy evolution with time.
References
Donahue, M., and Voit, G.M., 2004, Carnegie Observatories Astrophysics Series, Vol. 3, Edited by J.S. Mulchaey, A. Dressler, and A. Oemler, p. 144Fabian, A. C., and Nulsen, P. E. J., 1977, MNRAS, 180, 479
Kaiser, C. R., and Binney, J., 2003, MNRAS, 338, 837
Molendi, S., and Pizzolato, F., 2001, ApJ, 560, 194
Peterson, J. R., et al., 2003, ApJ, 590, 207
Peterson, J. R. et al., 2001, A&A, 365, L104
Sarazin, C. L. 1986, Rev. Mod. Phys., 58, 1
