arXiv subjects

S. Borgani

Publications and source records attributed to S. Borgani.

At least 145 records · Page 8Linked to original sources

arXiv2006

Angular diameter distance estimates from the Sunyaev-Zel'dovich effect in hydrodynamical cluster simulations

The angular-diameter distance D_A of a galaxy cluster can be measuread by combining its X-ray emission with the cosmic microwave background fluctution due to the Sunyaev-Zeldovich effect. The application of this distance indicator usually assumes that the cluster is spherically symmetric, the gas is distributed according to the isothermal beta-model, and the X-ray temperature is an unbiased measure of the electron temperature. We test these assumptions with galaxy clusters extracted from an extended set of cosmological N-body/hydrodynamical simulations of a LCDM concordance cosmology, which include the effect of radiative cooling, star formation and energy feedback from supernovae. We find that, due to the steep temperature gradients which are present in the central regions of simulated clusters, the assumption of isothermal gas leads to a significant underestimate of D_A. This bias is efficiently corrected by using the polytropic version of the beta-model to account for the presence of temperature gradients. In this case, once irregular clusters are removed, the correct value of D_A is recovered with a ~ 5 per cent accuracy on average, with a ~ 20 per cent intrinsic scatter due to cluster asphericity. This result is valid when using either the electron temperature or a spectroscopic-like temperature. When using instead the emission-weighted definition for the temperature of the simulated clusters, D_A is biased low by \~ 20 per cent. We discuss the implications of our results for an accurate determination of the Hubble constant H_0 and of the density parameter Omega_m. We find that H_0 can be potentially recovered with exquisite precision, while the resulting estimate of Omega_m, which is unbiased, has typical errors Delta(Omega_m) ~ 0.05.

astro-ph↗

arXiv2006

Systematics in the X-ray Cluster Mass Estimators

We examine the systematics affecting the X-ray mass estimators applied to a set of five simulated galaxy clusters. They have been processed through the X-ray Map Simulator, X-MAS, to provide Chandra-like long exposures that are analyzed to reconstruct the gas temperature, density, and mass profiles used as input. We find that at R_2500 the mass profile obtained via a direct application of the hydrostatic equilibrium equation is consistent within 1 sigma with the actual mass; although we notice this estimator shows high statistical errors due to high level of Chandra background. Instead, the poorness of the beta-model in describing the gas density profile makes the evaluated masses to be underestimated by \sim 40 per cent with respect to the true mass, both with an isothermal and a polytropic temperature profile. We also test ways to recover the mass by adopting an analytic mass model, such as those proposed by Navarro et al. (1997) and Rasia et al. (2004), and fitting the temperature profile expected from the hydrostatic equilibrium equation to the observed one. These methods and the one of the hydrostatic equilibrium equation provide a more robust mass estimation than the ones based on the beta-model. In the present work the main limitation for a precise mass reconstruction is to ascribe to the relatively high level of the background chosen to reproduce the Chandra one. After artificially reducing it by a factor of 100, we find that the estimated mass significantly underestimates the true mass profiles. This is manly due (i) to the neglected contribution of the gas bulk motions to the total energy budget and (ii) to the bias towards lower values of the X-ray temperature measurements because of the complex thermal structure of the emitting plasma.

astro-ph↗

arXiv2006

X-ray spectral properties of AGN in the Chandra Deep Field South

We present a detailed X-ray spectral analysis of the sources in the 1Ms catalog of the Chandra Deep Field South (CDFS) taking advantage of optical spectroscopy and photometric redshifts for 321 sources. As a default spectral model, we adopt a power law with slope Gamma with an intrinsic redshifted absorption N_H, a fixed Galactic absorption and an unresolved Fe emission line. For 82 X-ray bright sources, we perform the X-ray spectral analysis leaving both Gamma and N_H free. The weighted mean value is ~ 1.75+-0.02, with an intrinsic dispersion of sigma~0.30. We do not find hints of a correlation between Gamma and the N_H. We detect the presence of a scattered component at soft energies in 8 sources, and a pure reflection spectrum, typical of Compton-thick AGN, in 14 sources (Compton-thick AGN candidates). The intrinsic N_H distribution shows a lognormal shape, peaking around log(N_H)~23.1 and with sigma~1.1. We find that the fraction of absorbed sources (with N_H>10^{22} cm^{-2}) in the sample is constant (at the level of about 75%) or moderately increasing with redshift. Finally, we compare the optical classification to the X-ray spectral properties, confirming that the correspondence of unabsorbed (absorbed) X-ray sources to optical Type I (Type II) AGN is accurate for at least 80% of the sources with spectral identification (1/3 of the total X-ray sample).

astro-ph↗

arXiv2006

Properties of the diffuse X-ray background in a high-resolution hydrodynamical simulation

We study the properties of the diffuse X-ray background by using the results of a cosmological hydrodynamical simulation of the concordance LambdaCDM model. The simulation follows gravitational and gas dynamics and includes a treatment of physical processes like radiative cooling, star formation and supernova feedback. From the simulation outputs, we produce a set of two-dimensional maps of the intergalactic medium X-ray emission integrated over redshift. We find that the signal in the soft (0.5-2 keV) band is lognormally distributed with a mean intensity of about 4 10^-12 erg s^-1 cm^-2 deg^-2; approximately 40 per cent of the emission originates from warm-hot gas (defined as baryons with 10^5<T<10^7 K), and 90 per cent comes from structures at z<0.9. Since the spectrum is soft, being mostly provided by the intergalactic medium at low temperature, the total mean intensity in the hard (2-10 keV) X-ray band is smaller by a factor of about 4. In order to constrain the physical processes included in our simulation, we compare our results with the observed upper limit (1.2 +/- 0.3) 10^-12 erg s^-1 cm^-2 deg^-2 of the soft X-ray emission due to diffuse gas. To this purpose, we remove the contributions of observable extended objects (groups and clusters of galaxies) from the simulated maps by adopting different detectability criteria which are calibrated on the properties of systems at intermediate redshifts observed by Chandra. We show that the simulated diffuse soft X-ray emission is consistent with the present observed upper limit. However, if future measurements will decrease the level of the unresolved X-ray background by a factor of two, a more efficient feedback mechanism should be required to suppress the soft emission of the gas residing within filaments and group-size haloes.

astro-ph↗

arXiv2005

Hot and Cooled baryons in SPH simulations of galaxy clusters: physics and numerics

We discuss an extended set of Tree+SPH simulations of galaxy clusters, with the goal of investigating the interplay between numerical resolution effects and star-formation/feedback processes. The simulated clusters span the mass range (0.1-2.3) 10^{15}Msun/h, with mass resolution varying by several decades. At the highest achieved resolution, we resolve the virial region of a Virgo-like cluster with more than 2 million gas particles and with at least as many dark-matter (DM) particles. Our resolution study confirms that, in the absence of an efficient feedback mechanism, runaway cooling leads to about 35 per cent of baryons in clusters to be locked up in long lived stars at our highest resolution, with no evidence of convergence. However, including feedback causes the fraction of cooled baryons to converge at about 15 per cent already at modest resolution. Feedback also stabilizes other gas-related quantities, such as radial profiles of entropy, gas density and temperature, against variations due to changes in resolution. We also investigate the influence of the gravitational force softening length, and that of numerical heating of the gas induced by two-body encounters between DM and lighter gas particles. We show that simulations where more DM than gas particles are used, show a significantly enhanced efficiency of star formation at z>3. Our results are important for establishing and delineating the regime of numerical reliability of the present generation of hydrodynamical simulations of galaxy clusters.

astro-ph↗

arXiv2005

The baryon fraction in hydrodynamical simulations of galaxy clusters

We study the baryon mass fraction in a set of hydrodynamical simulations of galaxy clusters performed using the Tree+SPH code GADGET-2. We investigate the dependence of the baryon fraction upon the radiative cooling, star formation, feedback through galactic winds, conduction and redshift. Both the cold stellar component and the hot X-ray emitting gas have narrow distributions that, at large cluster-centric distances r>R500, are nearly independent of the physics included in the simulations. Only the non-radiative runs reproduce the gas fraction inferred from observations of the inner regions (r ~ R2500) of massive clusters. When cooling is turned on, the excess star formation is mitigated by the action of galactic winds, but yet not by the amount required by observational data. The baryon fraction within a fixed overdensity increases slightly with redshift, independent of the physical processes involved in the accumulation of baryons in the cluster potential well. In runs with cooling and feedback, the increase in baryons is associated with a larger stellar mass fraction that arises at high redshift as a consequence of more efficient gas cooling. For the same reason, the gas fraction appears less concentrated at higher redshift. We discuss the possible cosmological implications of our results and find that two assumptions generally adopted, (1) mean value of Yb = fb / (Omega_b/Omega_m) not evolving with redshift, and (2) a fixed ratio between f_star and f_gas independent of radius and redshift, might not be valid. In the estimate of the cosmic matter density parameter, this implies some systematic effects of the order of Delta Omega_m/Omega_m < +0.15 for non-radiative runs and Delta Omega_m/Omega_m ~ +0.05 and < -0.05 for radiative simulations.

astro-ph↗

arXiv2005

An X-ray Galaxy Cluster Survey for Investigations of Dark Energy

The amount and nature of dark energy (DE) can be tightly constrained by measuring the spatial correlation features and evolution of a sample of ~ 100,000 galaxy clusters over the redshift range 0<z < 1.5. Such an X-ray survey will discover all collapsed structures with mass above 3.5e14 solar masss at redshifts z<2 (i.e. the full range where such objects are expected) in the high Galactic latitude sky. Above this mass threshold the tight correlations between X-ray observables and mass allow direct interpretation of the data. We describe the constraints on Dark Energy that can be inferred from such a survey, using powerful self-calibration techniques to relate the X-ray observables (luminosity and temperature) to the underlying mass.

astro-ph↗

arXiv2005

Internal dynamics of the z sim 0.8 cluster RX J0152.7-1357

We present the results from the dynamical analysis of the cluster of galaxies RX J0152-1357, which shows a complex structure in its X-ray emission, with two major clumps in the central region, and a third clump in the Eastern region. Our analysis is based on redshift data for 187 galaxies. We compute the line-of-sight velocity dispersion of galaxies, sigma_V=1322^+74_-68 kms^-1, which is significantly larger than what is expected in the case of a relaxed cluster with an observed X-ray temperature of 5-6 keV. We find evidence that this cluster is far from dynamical equilibrium, as shown by the non Gaussianity of the velocity distribution, the presence of a velocity gradient and significant substructure. Our analysis shows that the high value of sigma_V is due to the complex structure of RX J0152-1357, i.e. to the presence of three galaxy clumps of different mean velocity. Using optical data we detect a low-velocity clump (with sigma_V=300-500 kms^-1) in the central SW region and a high--velocity clump (with sigma_V sim 700 kms^-1) in the E region, well corresponding to the SW and E peaks detected in the X-ray emission. The central NE X-ray peak is associated to the main galaxy structure with a velocity which is intermediate between those of the other two clumps and sigma_V sim 900 kms^-1. The mass of the whole system within 2 Mpc is estimated to lie in the range (1.2-2.2)X10^{15} M_sun. Analytic calculations based on the two-body model indicate that the system is most likely bound, currently undergoing merging. In particular, we suggest that the SW clump is not a small group, but rather the dense cluster-core of a massive cluster.

astro-ph↗

arXiv2005

Entropy amplification from energy feedback in simulated galaxy groups and clusters

We use hydrodynamical simulations of galaxy clusters and groups to study the effect of pre-heating on the entropy structure of the ICM. Our simulations account for non-gravitational heating of the gas either by imposing a minimum entropy floor at redshift z=3, or by considering feedback by galactic winds powered by supernova (SN) energy. In the adiabatic simulations we find that the entropy is increased out to the external regions as a consequence of the transition from clumpy to smooth accretion induced by extra heating. This result is in line with the predictions of the semi-analytical model by Voit et al. However, the introduction of radiative cooling substantially reduces this entropy amplification effect. While galactic winds of increasing strength are effective in regulating star formation, they have a negligible effect on the entropy profile of cluster-sized halos. Only in models where the action of the winds is complemented with diffuse heating corresponding to a pre-collapse entropy do we find a sizable entropy amplification out to the virial radius of the groups. Observational evidence for entropy amplification in the outskirts of galaxy clusters and groups therefore favours a scenario for feedback that distributes heating energy in a more diffuse way than predicted by the model for galactic winds from SN explosions explored here.

astro-ph↗

arXiv2004

Spectroscopic-Like Temperature of Clusters of Galaxies and Cosmological Implications

The thermal properties of hydrodynamical simulations of galaxy clusters are usually compared to observations by relying on the emission-weighted temperature T_ew, instead of on the spectroscopic X-ray temperature T_spec, which is obtained by actual observational data. Here we show that, if the intra-cluster medium is thermally complex, T_ew fails at reproducing T_spec. We propose a new formula, the spectroscopic-like temperature, T_sl, which approximates T_spec better than a few per cent. By analyzing a set of hydrodynamical simulations of galaxy clusters, we also find that T_sl is lower than T_ew by 20-30 per cent. As a consequence, the normalization of the M-T relation from the simulations is larger than the observed one by about 50 per cent. If masses in simulated clusters are estimated by following the same assumptions of hydrostatic equilibrium and beta-model gas density profile, as often done for observed clusters, then the M-T relation decreases by about 40 per cent, and significantly reduces its scatter. Based on this result, we conclude that using the observed M-T relation to infer the amplitude of the power spectrum from the X--ray temperature function could bias low sigma_8 by 10-20 per cent. This may alleviate the tension between the value of sigma_8 inferred from the cluster number density and those from cosmic microwave background and large scale structure.

astro-ph↗

arXiv2004

A VLT spectroscopic survey of RX J0152.7-1357, a forming cluster of galaxies at z=0.837

We present the results of an extensive spectroscopic survey of RX J0152.7-1357, one of the most massive distant clusters of galaxies known. Multi-object spectroscopy, carried out with FORS1 and FORS2 on the ESO Very Large Telescope (VLT), has allowed us to measure more than 200 redshifts in the cluster field and to confirm 102 galaxies as cluster members. The mean redshift of the cluster is $z=0.837 \pm 0.001$ and we estimate the velocity dispersion of the overall cluster galaxy distribution to be $\sim 1600 \mathrm{km \ s^{-1}}$. The distribution of cluster members is clearly irregular, with two main clumps that follow the X-ray cluster emission mapped by Chandra. A third clump of galaxies to the east of the central structure and at the cluster redshift has also been identified. The two main clumps have velocity dispersions of $\sim919$ and $\sim737 \mathrm{km s^{-1}}$ respectively, and the peculiar velocity of the two clumps suggests that they will merge into a single more massive cluster. A segregation in the star formation activity of the member galaxies is observed. All star forming galaxies are located outside the high-density peaks, which are populated only by passive galaxies. A population of red galaxies (belonging to the cluster red sequence) with clear post-starburst spectral features and [OII] ($λ$3727) emission lines is observed in the outskirts of the cluster. Two AGNs, which were previously confused with the diffuse X-ray emission from the intracluster medium in ROSAT and BeppoSAX observations, are found to be cluster members.

astro-ph↗

arXiv2004

Measuring cluster peculiar velocities with the Sunyaev-Zeldovich effects: scaling relations and systematics

The fluctuations in the Cosmic Microwave Background (CMB) intensity due to the Sunyaev-Zeldovich (SZ) effect are the sum of a thermal and a kinetic contribution. Separating the two components to measure the peculiar velocity of galaxy clusters requires radio and microwave observations at three or more frequencies, and knowledge of the temperature T_e of the intracluster medium weighted by the electron number density. To quantify the systematics of this procedure, we extract a sample of 117 massive clusters at redshift z=0 from an N-body hydrodynamical simulation, with 2x480^3 particles, of a cosmological volume 192 Mpc/h on a side of a flat Cold Dark Matter model with Omega_0=0.3 and Lambda=0.7. Our simulation includes radiative cooling, star formation and the effect of feedback and galactic winds from supernovae. We find that (1) our simulated clusters reproduce the observed scaling relations between X-ray and SZ properties; (2) bulk flows internal to the intracluster medium affect the velocity estimate by less than 200 km/s in 93 per cent of the cases; (3) using the X-ray emission weighted temperature, as an estimate of T_e, can overestimate the peculiar velocity by 20-50 per cent, if the microwave observations do not spatially resolve the cluster. For spatially resolved clusters, the assumptions on the spatial distribution of the ICM, required to separate the two SZ components, still produce a velocity overestimate of 10-20 per cent, even with an unbiased measure of T_e. Thanks to the large size of our cluster samples, these results set a robust lower limit of 200 km/s to the systematic errors that will affect upcoming measures of cluster peculiar velocities with the SZ effect.

astro-ph↗

arXiv2004

The spatial clustering of X-ray selected AGN and galaxies in the Chandra Deep Field South and North

We investigate the spatial clustering of X-ray selected sources in the two deepest X-ray fields to date, namely the 2Msec Chandra Deep Field North (CDFN) and the 1Msec Chandra Deep Field South (CDFS). The projected correlation function w(r_p), measured on scales ~0.2-10 h^-1 Mpc for a sample of 240 sources with spectroscopic redshift in the CDFN and 124 sources in the CDFS at a median redshift of z~0.8, is used to constrain the amplitude and slope of the real space correlation function xi(r)=(r/r0)^-gamma. The clustering signal is detected at high confidence (>~ 7 sigma) in both fields. The amplitude of the correlation is found to be significantly different in the two fields, the correlation length r0 being 8.6 +- 1.2 h^-1 Mpc in the CDFS and 4.2 +- 0.4 h^-1 Mpc in the CDFN, while the correlation slope gamma is found to be flat in both fields: gamma=1.33 +- 0.11 in the CDFS and gamma=1.42 +- 0.07 in the CDFN (a flat Universe with Omega_m=0.3 and Omega_L=0.7 is assumed; 1 sigma Poisson error estimates are considered). The correlation function has been also measured separately for sources classified as AGN or galaxies. In both fields AGN have a median redshift of z~0.9 and a median 0.5-10 keV luminosity of L_x~10^43 erg s^-1, i.e. they are generally in the Seyfert luminosity regime. As in the case of the total samples, we found a significant difference in the AGN clustering amplitude between the two fields, the best fit correlation parameters being r0=10.3 +- 1.7 h^-1 Mpc, gamma=1.33 +- 0.14 in the CDFS, and r0=5.5 +- 0.6 h^-1 Mpc, gamma=1.50 +- 0.12 in the CDFN. Within each field no statistically significant difference is found between soft and hard X-ray selected sources or between type 1 and type 2 AGN. (abridged)

astro-ph↗

arXiv2004

Simulating the Soft X-ray excess in clusters of galaxies

The detection of excess of soft X-ray or Extreme Ultraviolet (EUV) radiation, above the thermal contribution from the hot intracluster medium (ICM), has been a controversial subject ever since the initial discovery of this phenomenon. We use a large--scale hydrodynamical simulation of a concordance $Λ$CDM model, to investigate the possible thermal origin for such an excess in a set of 20 simulated clusters having temperatures in the range 1--7 keV. Simulated clusters are analysed by mimicking the observational procedure applied to ROSAT--PSPC data, which for the first time showed evidences for the soft X-ray excess. For cluster--centric distances $0.4< R/R_{\rm vir}< 0.7$ we detect a significant excess in most of the simulated clusters, whose relative amount changes from cluster to cluster and, for the same cluster, by changing the projection direction. In about 30 per cent of the cases, the soft X-ray flux is measured to be at least 50 per cent larger than predicted by the one--temperature plasma model. We find that this excess is generated in most cases within the cluster virialized regions. It is mainly contributed by low--entropy and high--density gas associated with merging sub--halos, rather than to diffuse warm gas. Only in a few cases the excess arises from fore/background groups observed in projection, while no evidence is found for a significant contribution from gas lying within large--scale filaments. We compute the distribution of the relative soft excess, as a function of the cluster--centric distance, and compare it with the observational result by Bonamente et al. (2003) for the Coma cluster. Similar to observations, we find that the relative excess increases with the distance from the cluster center, with no significant excess detected for $R<0.4R_{\rm vir}$. (abridged)

astro-ph↗

arXiv2004

Mismatch between X-ray and emission-weighted temperatures in galaxy clusters: cosmological implications

The thermal properties of hydrodynamical simulations of galaxy clusters are usually compared to observations by relying on the emission-weighted temperature T_ew, instead of on the spectroscopic X-ray temperature T_spec, which is obtained by actual observational data. In a recent paper Mazzotta et al. show that, if the cluster is thermally complex, T_ew fails at reproducing T_spec, and propose a new formula, the spectroscopic-like temperature, T_sl, which approximates T_spec better than a few per cent. By analyzing a set of hydrodynamical simulations of galaxy clusters, we find that T_sl is lower than T_ew by 20-30 per cent. As a consequence, the normalization of the M-T_sl relation from the simulations is larger than the observed one by about 50 per cent. If masses in simulated clusters are estimated by following the same assumptions of hydrostatic equilibrium and β--model gas density profile, as often done for observed clusters, then the M-T relation decreases by about 40 per cent, and significantly reduces its scatter. Based on this result, we conclude that using the observed M-T relation to infer the amplitude of the power spectrum from the X-ray temperature function could bias low σ_8 by 10-20 per cent. This may alleviate the tension between the value of σ_8 inferred from the cluster number density and those from cosmic microwave background and large scale structure.

astro-ph↗

arXiv2004

Evolution at z>0.5 of the X-ray properties of simulated galaxy clusters: comparison with the observational constraints

(ABRIDGED) We analyze the X-ray properties of a sample of local and high redshift galaxy clusters extracted from a large cosmological hydrodynamical simulation. This simulation has been realized using the Tree+SPH code GADGET-2 for a LambdaCDM model. In our analysis, we consider only objects with T_ew >2 keV and adopt an approach that mimics observations, associating with each measurement an error comparable with recent observations and providing best-fit results via robust techniques. Within the clusters, baryons are distributed among (i) a cold neutral phase, with a relative contribution that increases from less than 1 to 3 per cent at higher redshift, (ii) stars which contribute with about 20 per cent and (iii) the X-ray emitting plasma that contributes by 80 (76) per cent at z=0 (1) to the total baryonic budget. A depletion of the cosmic baryon fraction of ~7 (at z=0) and 5 (at z=1) per cent is measured at the virial radius, R_vir, in good agreement with adiabatic hydrodynamical simulations. We confirm that, also at redshift >0.5, power-law relations hold between gas temperature, T, bolometric luminosity, L, central entropy, S, gas mass, M_gas, and total gravitating mass, M_tot and that these relations are steeper than predicted by simple gravitational collapse. A significant, negative evolution in the L-T and L-M_tot relations and positive evolution in the S-T relation are detected at 0.5 < z < 1 in this set of simulated galaxy clusters. This is partially consistent with recent analyses of the observed properties of z>0.5 X-ray galaxy clusters. By fixing the slope to the values predicted by simple gravitational collapse, we measure at high redshift normalizations lower by 10-40 per cent in the L-T, M_tot-T, M_gas-T, f_gas-T and L-M_tot relations than the observed estimates.

astro-ph↗

arXiv2004

The Diffuse Light in Simulations of Galaxy Clusters

We study the properties of the diffuse light in galaxy clusters forming in a large hydrodynamical cosmological simulation of the Lambda-CDM cosmology. The simulation includes a model for radiative cooling, star formation in dense cold gas, and feedback by SN-II explosions. We select clusters having mass M>10^(14) h^(-1) Msun and study the spatial distribution of their star particles. While most stellar light is concentrated in gravitationally bound galaxies orbiting in the cluster potential, we find evidence for a substantial diffuse component, which may account for the extended halos of light observed around central cD galaxies. We find that more massive simulated clusters have a larger fraction of stars in the diffuse light than the less massive ones. The intracluster light is more centrally concentrated than the galaxy light, and the stars in the diffuse component are on average older than the stars in cluster galaxies, supporting the view that the diffuse light is not a random sampling of the stellar population in the cluster galaxies. We thus expect that at least ~10% of the stars in a cluster may be distributed as intracluster light, largely hidden thus far due to its very low surface brightness.

astro-ph↗

arXiv2004

Simulating the metal enrichment of the ICM

We present results from Tree+SPH simulations of a galaxy cluster, aimed at studying the metal enrichment of the intra--cluster medium (ICM). The simulation code includes a fairly advanced treatment of star formation, as well as the release of energy feedback and detailed yields from both type-II and type-Ia supernovae, also accurately accounting for the lifetimes of different stellar populations. We perform simulations of a cluster with virial mass ~ 3.9x 10^14 Msun, to investigate the effect of varying the feedback strength and the stellar initial mass function (IMF). Although most of the models are able to produce acceptable amounts of Fe mass, we find that the profiles of the iron abundance are always steeper than observed. The [O/Fe] ratio is found to be sub--solar for a Salpeter IMF, with [O/Fe] -0.2 at R >~ 0.1R200, whereas increasing to super-solar values in central regions, as a result of recent star formation. Using a top--heavier IMF gives a larger [O/Fe] over the whole cluster, at variance with observations. On the other hand, the adoption of a variable IMF, which becomes top-heavier at z>2, provides a roughly solar [O/Fe] ratio. Our results indicate that our simulations still lack a feedback mechanism which should quench star formation at low redshift and transport metals away from the star forming regions.

astro-ph↗