Annu. Rev. Astron. Astrophys. 2000. 38: 289-335 Copyright © 2000 by Annual Reviews. All rights reserved

5.9. The Origin and Evolution of the Intragroup Medium

The presence of heavy elements in the intragroup medium indicates that a substantial fraction of the diffuse gas must have passed through stars. The presence of iron is particularly important because it suggests that supernovae played an important role in the enrichment of the gas. In principle, X-ray spectroscopy can provide detailed constraints on the stars responsible for the enrichment. For example, the relative abundance of the -burning elements to iron is a measure of the relative importance of Type II to Type 1a supernovae (Renzini et al 1993, Renzini 1997, Gibson et al 1997). For the gas temperatures characteristic of groups (~ 1 keV), strong emission lines are expected for many of the elements including oxygen, neon, magnesium, silicon and sulfur. Although most ASCA studies suggest that the / Fe ratio is approximately solar in groups, this result is somewhat inconclusive at present because of uncertainties in the spectral modeling.

Renzini and collaborators have used the concept of iron mass-to-light ratio to study the history of the hot gas in groups and clusters (Renzini et al 1993, Renzini 1997). They find that the X-ray emitting gas in rich clusters contains ~ 0.01 h^-1/2 M of iron for each L of blue light. The iron mass-to-light ratio is effectively constant for clusters with temperature between ~ 2 and 10 keV. However, this ratio is typically a factor of ~ 50 lower in X-ray groups (Renzini et al 1993, Renzini 1997, Davis et al 1999). The iron mass-to-light ratios of groups are lower than those of clusters because both the overall iron abundance and the gas-to-stellar mass ratio are lower in groups than in clusters (Renzini 1997). The low iron mass-to-light ratios may be evidence that a significant amount of mass has been lost in groups. The escape velocities of groups are comparable to the escape velocities of individual galaxies. Thus, material that is ejected from galaxies may also escape the group. Several mechanisms have been proposed to eject material from groups, including galactic winds and outflows powered by supernovae or nuclear activity (Renzini 1997). The material lost from groups may have contributed significantly to the enrichment of the intergalactic medium (Davis et al 1999).

The iron mass-to-light ratios of groups could be somewhat underestimated if the true iron abundances are higher than the sub-solar values usually derived from isothermal model fits. However, the gas-mass estimates are less sensitive to the iron abundance assumed and uncertainties in the iron abundances likely lead to inaccuracies in the gas-mass estimates of at most ~ 50% (Pildis et al 1995). A potentially bigger problem is that many groups are detected to a much smaller fraction of the virial radius than their rich clusters counterparts. Thus, the true gas masses in some groups may be significantly underestimated from the existing X-ray data. In fact, it is possible that the differences in the iron mass-to-light ratios of groups and clusters may largely be a result of this effect and not necessarily evidence for mass loss.

The mechanisms responsible for producing metals may also inject energy into the gas. Numerical simulations indicate that in the absence of such non-gravitational heating, the density profiles of groups and clusters are nearly identical (Navarro et al 1997). There is now considerable evidence for departures from such uniformity. In the standard hierarchical clustering models, the X-ray luminosity is expected to scale with temperature as L_X T² (e.g. Kaiser 1991). The observed relationship is considerably steeper, especially for small groups (see Figure 6). Furthermore, the ratio of specific energy of the galaxies to specific energy of the gas (i.e. the parameter) is less than one for low-mass systems. (However, see Section 3.3.3 for a discussion of why the observed values for groups may be biased low.) Both of these observations suggest that the gas temperature may not be a good indicator of the virial temperature in poor groups. Entropy profiles for groups and clusters indicate that the entropy of the group gas is also higher than can be achieved through gravitational collapse alone (David et al 1996, Ponman et al 1999, Lloyd-Davies et al 2000). All of these observations are consistent with preheating models for the hot gas (Kaiser 1991; Evrard & Henry 1991; Metzler & Evrard 1994; Knight & Ponman 1997; Cavaliere et al 1997, 1998, 1999; Arnaud & Evrard 1999; Balogh et al 1999; Tozzi et al 2000; Loewenstein 2000;; Tozzi & Norman 2000). Such preheating leads to a more extended gas component in groups than in rich clusters (i.e. lower central gas densities and shallower density slopes). Moreover, without preheating, groups appear to over-produce the X-ray background (Wu et al 2000).

Ponman and collaborators have estimated the excess entropy associated with the preheating in groups and find that it corresponds to a temperature of ~ 0.3 keV (Ponman et al 1999, Lloyd-Davies et al 2000). The preheating temperature can be combined with the excess entropy to estimate the electron density of the gas into which the energy was injected. The resulting value (n ~ 4 × 10^-4 h₁₀₀^0.5 cm^-3) implies that the heating occurred prior to the cluster collapse but after a redshift of z ~ 10 (Lloyd-Davies et al 2000). The current estimates for the entropy associated with the preheating have been based on rather small samples of groups and clusters, and these techniques will undoubtably improve with the next generation of X-ray telescopes. Already it is clear that such research can provide considerable insight into the history of the gas and group formation.