Clouds of hot X-ray emitting gas associated with clusters of galaxies are the biggest aggregates of baryons that we know, except for the cosmic web. A typical cloud contains the nuclear-fusion products of billions of supernovae. Therefore, they are representative objects to study the enrichment of the bulk of the matter in the universe. The elemental abundances in this hot Intra-Cluster Medium (ICM) are able to reveal the clusters enrichment history and they can also be used to constrain supernova models. Using the current state-of-the-art in high-resolution X-ray spectroscopy, XMM-Newton, Jelle de Plaa has analysed data of 23 clusters of galaxies in total. He finds that detailed knowledge about the X-ray background, spectral lines, and the response of the instruments is absolutely necessary to obtain accurate abundances with realistic uncertainties. To improve the accuracy on the abundances and their uncertainties, he has developed a method to correct for the soft X-ray background. Moreover, systematic differences between the instruments also need to be taken into account. Therefore, analysis of extended sources, like clusters, cannot be done using a black-box approach, but a large effort is needed to extract robust values. Only when robust values are obtained, one can attempt to derive the enrichment history. De Plaa shows that EPIC and RGS can be used to obtain information about the spatial abundance distribution in clusters. From the RGS spectrum of Sérsic 159-03, he derives that the iron abundance profile (a typical supernova type Ia product) is peaked around the central cluster galaxy, while the oxygen abundance distribution (a typical core-collapse supernova product) is more extended. This confirms the idea that iron is still being added to the ICM specifically in the core by the supernovae type Ia, and that oxygen is an old well-mixed product of the core-collapse supernovae that exploded soon after the first star bursts. The spatial oxygen distribution is not required to be entirely flat, but may exhibit a more shallowly peaked distribution due to ram-pressure stripping of gas from in-falling galaxies. Perhaps this is the case in 2A 0335+096, were he finds a similar spatial profile for oxygen and iron with RGS. The abundances that he measures also provide constraints on the supernovae themselves. He concludes that the currently favoured supernova models are not able to match our data. An empirically modified supernova type Ia model provides a better fit to the Ar/Ca ratio that he finds. The fitted abundances can provide information to constrain physical parameters of type Ia supernovae, like the density where the sound wave that is ahead of the explosion turns into a shock. In his final chapter, De Plaa argues that new deep observations of clusters with XMM-Newton or future observatories will provide much more insight into supernova explosions and the enrichment of the largest bound objects we know.
