The ICM and the IGM show metal lines in the X-ray spectra. These metals cannot have been produced in the gas, but they must have been produced in the galaxies and subsequently transported from the galaxies into the ICM/IGM by certain processes like e.g. ram-pressure stripping, galactic winds, galaxy-galaxy interaction or jets from active galaxies. The metallicity is the best indicator for finding out which of these processes are most important. Of special interest is the distribution of metals. So far there are only few examples of measured metallicity variations in real 2D maps and not only profiles. In CL0939+4713 we find different metallicity in the different subclusters (De Filippis et al. 2002). In the Perseus cluster also clear metallicity variations were found (Schmidt et al. 2002). 1D profiles are not very useful in this context because photons from regions in the cluster which are very far apart are accumulated in the same spectrum.
Apart from the metallicity distribution also the evolution of the metallicity is interesting. As soon as enough XMM and CHANDRA observations of distant clusters are available we can compare the metallicities in these clusters with those of nearby clusters. This is another way of distinguishing between the enrichment processes as different processes have different time dependence. In addition element ratios can be derived, e.g of Fe and -elements to get information on the different types of supernovae that have contributed to the metal enrichment.
Various processes have been suggested for the transport of gas from the galaxies to the ICM/IGM. 30 years ago Gunn & Gott (1972) suggested ram-pressure stripping: as the galaxy moves through the cluster and approaches the cluster centre it feels the increasing pressure of the intra-cluster gas. At some point the galaxy is not able anymore to retain its ISM. The ISM is stripped off and lost to the ICM and with it all its metals. Many numerical simulations have been performed to investigate this process, first 2D models (Takeda et al. 1984; Gaetz et al. 1987; Portnoy et al. 1993; Balsara et al. 1994). With increasing computing power also more detailed 3D models could be calculated (Abadi et al. 1999; Quilis et al. 2000; Vollmer et al. 2001; Schulz & Struck 2001; Toniazzo & Schindler 2001). In Fig. 6 such a simulated stripping process is shown for an elliptical galaxy.
Figure 6. Gas density (grey scale) and pressure (contours) of a galaxy moving downwards towards the cluster centre. The arrows show the Mach vectors (white when M > 1, black otherwise). The gas of the galaxy is stripped due to ram pressure (from Toniazzo & Schindler 2001).
Another possible process is galactic winds e.g. driven by supernovae (De Young 1978). Also for this process simulations have been performed on order to see whether only winds can account for the observed metallicities. The results were quite discordant as the following two examples show. Metzler & Evrard (1994, 1997) found that winds can account for the metals, while Murakami & Babul (1999) concluded that winds are not very efficient for the metal enrichment. In the simulations of Metzler & Evrard quite steep metallicity gradients showed up which are not in agreement with observations.
A third possible process is galaxy-galaxy interactions, like tidal stripping or galaxy harassment. Also during these interactions a lot of ISM can be lost to the ICM and IGM. This process is very likely more efficient in groups of galaxies, because in these systems the relative velocities are smaller and therefore the interaction timescales are longer. The ram-pressure stripping on the other hand is probably less efficient in groups because not only the pressure of the IGM is lower than that of the ICM, but also the velocities are smaller. This is also very important as the stripping is about proportional to gas v2.
A forth possible mechanism is jets emitted by active galaxies. These jets can also carry metals. Fig. 7 shows the interaction of jets with the ICM as it was discovered by X-ray observations. In the cluster RBS797 minima in the X-ray emission have been detected in a CHANDRA observation (Schindler et al. 2001). The X-ray depressions are arranged opposite with respect to the cluster centre. It is very likely that the pressure of the relativistic particles in the jets push away the X-ray gas. Preliminary radio observations with the VLA confirm this hypothesis.
Figure 7. CHANDRA image of the central part of the cluster RBS797 (from Schindler et al. 2001). There are depressions in the X-ray emission which are located opposite to each other with respect to the cluster centre (see arrows). These depressions can be explained by an active galaxy in the centre of the cluster, which has two jets. The pressure of the relativistic particles in the jets push away the X-ray gas resulting in minima in the X-ray emission.
Simulations with different enrichment processes were also performed on cosmological scales. Also here quite discordant results have been found as the two following examples show. Gnedin (1998) found that galactic winds play only a minor role, while galaxy mergers eject most of the gas. In contrary to these results Aguirre et al. (2001) concluded that winds are most important and ram-pressure stripping is not very efficient. The reason for these differences are probably the large ranges in scale that are covered by these simulations, from cosmological scale down to galaxy scales. Therefore only a small number of particles are left for each single galaxy and hence galaxies are not well resolved. This can be the reason for the discordant results.
In order to clarify this we are currently performing comprehensive simulations, which include the different enrichment processes.