7 Abundance

Figure 9: The radial abundance profile of A1413, derived in the radial range where the spectra have clear detections of the Fe K line. Errors are $1\sigma$.

The combined MOS+pn spectral fits also allow us to calculate a radial abundance profile, which can be compared with the profiles derived for cluster samples observed with BeppoSAX (Irwin & Bregman 2001 (IB01); De Grandi & Molendi 2001 (DM01)). Figure 9 shows the profile derived in the conservative radial range where we have information from the Fe K$\alpha$ line.

The average abundance is $0.27\pm0.03$, more typical of the value found by DM01 for non-CF clusters ( $0.25\pm0.01$) than for CF clusters ( $0.34\pm0.01$). There is a clear decrease of abundance with radius. However, the profile exhibits a two-step behaviour, rather than a steadily declining profile, as found by DM01 and IB01. It remains more or less constant (a mean of 0.34, excluding the central bin) up to $\sim$400 kpc, dropping to another plateau (with a mean of 0.22) beyond. IB01 found that inside  0.075 r₂₀₀ their CF subsample had an average abundance of  $0.46 \pm 0.06$, while the non-CF subsample had an average of  $0.33 \pm 0.04$. The outer region abundances were  $0.30 \pm 0.02$ and  $0.24 \pm 0.03$ for the CF and non-CF subsamples, respectively. The inner two bins of our observation correspond roughly to  0.075 r₂₀₀. The mean error-weighted value for this region is  $0.34 \pm 0.08$ (3$\sigma$ errors, for direct comparison with IB01). Outside this region, the abundance value is  $0.28 \pm 0.09$. Thus when the errors are taken into account, we cannot distinguish between the CF and non-CF subsamples of IB01. However, taken at face value, these figures appear to suggest that the abundance profile shape of A1413 displays characteristics intermediate between the CF and non-CF subsamples. This may be because A1413 appears to host only a modest CF. It must also be noted that the abundance profiles of A1795 and A2142, both clusters with strong CF signatures (Peres et al. 1998) appear relatively flat, as shown in IB01 and DM01. It has been suggested that both of these clusters are undergoing (or have undergone) mergers (Oegerle & Hill 1994; Oegerle et al. 1995), which have presumably not been sufficiently strong to disrupt the CF, but which have effectively mixed the metals and thus flattened the radial profile. A study of cluster abundance profiles as a function of the strength of the CF signature and dynamical state would help to better understand the origin of cluster abundance gradients.