Near-IR studies

People involved at OAB: Comastri, Origlia.

Metal enrichment in starburst galaxies

The near-IR stellar luminosity of starburst galaxies is dominated by massive red supergiants. Such a stellar continuum generally largely dominates over the gas and dust emission (Oliva & Origlia 1998; Origlia & Oliva 2000), while in the visual range the nebular emission strongly dilutes the stellar absorption lines and dust can heavily obscure the central regions where most of the burst activity is concentrated. Their absorption spectra show many atomic and molecular lines which can be used to infer reliable abundances of key metals (e.g. C, O, Fe and other $\alpha$-elements). Metals locked in the stellar atmosphere of red supergiants trace the abundances just prior to the last burst of star formation. On the other hand, the hot gas in the nuclear region, probed by X-ray observations, is heated by type II SN explosions and therefore is related to the gas just enriched by the new generation of stars. The X-ray spectra obtained by the new generation of X-ray telescopes (Chandra and XMM-Newton) have a quality high enough to set good constraints on the metallicity of the hot gas in starburst galaxies and possible spatial gradients. We started with a successful observational campaign at the TNG with NICS, when we secured medium-resolution IR spectra of 4 starburst galaxies observed with Chandra and/or XMM-Newton, to infer reliable abundances of Fe, C, O, Si, Mg, Ca and Al and to obtain a detailed screening of the most important abundance patterns, namely [C/Fe] and $\alpha$/Fe], of the pre-burst medium, locked into the stellar photospheres. For the first time detailed stellar abundances in the nuclear region of the starburst galaxy M82 have been obtained. They are compared with those of the hot gas as derived from an accurate re-analysis of the XMM and Chandra nuclear X-ray spectra. The cool stars and the hot gas suggest [Fe/H] $=-0.35\pm0.2$ dex, and an overall [Si,Mg/Fe] enhancement by $\simeq 0.4$ and 0.5 dex, respectively. This is consistent with a major chemical enrichment by SNe II explosions in recursive bursts on short timescales. Oxygen is more puzzling to interpret since it is enhanced by $\simeq 0.3$ dex in stars and depleted by $\simeq 0.2$ dex in the hot gas. None of the standard enrichment scenarios can fully explain such a behavior when compared with the other $\alpha$-elements. The analysis of the IR and X-ray spectra of other 3 starburst galaxies is in progress. This work is carried out in collaboration with P. Ranalli (Astronomy Dept., University of Bologna), R. Maiolino and A. Marconi (INAF-Arcetri Obs.).

X-ray number counts and evolution of star-forming galaxies

The fluctuation analysis performed on the Chandra deep fields (Miyaji & Griffiths 2001) shows an excess of sources at faint fluxes ( $\lower.5ex\hbox{$\; \buildrel < \over \sim \;$}10^{-17}$ erg s$^{-1}$ cm$^{-2}$) with respect to the predictions of AGN synthesis models for the X-ray background (Comastri et al. 1995). The excess can be explained as the emergence of a population of ``normal galaxies''. The analysis of well defined samples of star-forming galaxies in the nearby and distant Universe indicates that a linear relation between X-ray and radio luminosity holds up to $z \simeq 1.3$ (Ranalli et al. 2003; Bauer et al. 2002) Since the deepest radio surveys (Fomalont et al. 1991; Richards 2000) show that at faint (sub-mJy) fluxes star-forming galaxies dominate the radio counts, it is possible to use the radio/X-ray relation to transform the observed radio Log$N$-Log$S$ in predicted X-ray number counts.

The predicted counts are in good agreement with the observed X-ray number counts (for fluxes larger than $5\times 10^{-17}$ erg s$^{-1}$ cm$^{-2}$) and, at fainter fluxes, with the observational limits from the fluctuation analysis in the deepest Chandra fields. Similar results are obtained if we consider the IR counts from the ISO ELAIS survey (Gruppioni et al. 2002), or if we integrate the X-ray luminosity function as derived from the radio and far infrared luminosity functions (Machalski & Godlowski 2000; Takeuchi et al. 2003; Sejeant et al. 2004).

This work is carried out in collaboration with P. Ranalli and G. Setti (Astronomy Dept., University of Bologna).