A possible solution to the [α/Fe]-σ problem in early-type galaxies within a hierarchical galaxy formation model

Abstract

The most massive elliptical galaxies apparently formed the fastest, because the ratio of α elements (such as oxygen) to iron is the smallest. In fact, iron is mainly produced from type Ia supernovae on a time‐scale of ∼0.1–1 Byr, while the α elements come from massive stars on time‐scales of a few tens of million years. Reproducing such a α/Fe correlation has long been a severe problem for cosmological theories of galaxy formation, which envisage massive galaxies to assemble gradually from smaller progenitors, and to be characterized by a star formation history too much extended towards late cosmic times. While it has recently become clear that feedback from Active Galactic Nuclei (AGNs) activity plays a role in the late quenching of star formation, and that early star formation history in the galaxy progenitors affect the α/Fe ratio, major mergers alone cannot enhance the star formation in the high‐redshift progenitors to the levels required to match the steepness of the observed α/Fe correlation. Here we report that the inclusion of the effects of fly‐by ‘harassments’, that trigger lower level starbursts, combined with the AGN quenching of the starburst activity, considerably enhances the capability to account for the observed α/Fe ratio in ellipticals within cosmological galaxy formation models. The critical difference between the earlier work and the present result is the effect of starbursts driven by fly‐by encounters that would have been very common amongst the high‐redshift progenitors of massive galaxies and which would have boosted star formation in the first 2 Byr after the big bang, combined with quenching of the burst activity within the first 3–4 Gyr.

Chemoarchaeological downsizing in a hierarchical universe: impact of a top-heavy IGIMF

“…The effect of starbursts triggered by companions is present in our model SAG through our particular implementation of disk instabilities (see Section 2.2.2). Forcing the free parameters of SAG to reproduce the [α/Fe]-stellar mass relation with a Salpeter IMF gives a similar slope than the best model of Calura & Menci (2011), although in both cases the slope is too shallow compared with observations. Even constraining the parameters with the relation itself, it is not possible to achieve a satisfactory slope.…”

“…Earlier attempts to model the abundance of [α/Fe] have failed to reproduce the observed relation with stellar mass, unless when invoking ad hoc hypotheses, like a change in the slope of the global IMF (Arrigoni et al 2010), an arbitrary SFR-dependent IMF (Calura & Menci 2009), starbursts in progenitor galaxies triggered by close encounters (Calura & Menci 2011) or empirically adopted SNIa delay times (Yates et al 2013). Motivated by our finding that a universal IMF does not naturally allow to reproduce the right slope in the observed trend between the [α/Fe] abundance ratio and the stellar mass of elliptical galaxies, we explore the impact of a variable IMF on their [α/Fe]-stellar mass relation.…”

“…We report here that when we change the slope of the IMF in SAG , we also have as a result a variation in the zero point, with no apreciable change in the slope, supporting the last conclusion. Calura & Menci (2011) include the effects of starbursts due to fly-by interactions at high redshift following the model developed by Cavaliere & Vittorini (2000) and studied by Menci et al (2004) in their SAM. They find that the slope of the [α/Fe]-stellar mass relation can be reproduced naturally when adding this effect, although the slope they obtain is shallower than expected from observations (see their Figure 1).…”

“…A similar assumption about the IMF has been considered by to explain the number counts of submillimiter galaxies at high redshift. Later on, Calura & Menci (2011) mention the importance of the harassment of satellite galaxies at high redshift as a trigger of SF in the progenitors of massive galaxies that form stars enhanced in α-elements, contributing to build up the relation. Arrigoni et al (2010) also analyse the [α/Fe]-mass relation by using the semi-analytic model of Somerville et al (2008) that includes AGN feedback, in which they implement a chemical enrichment model that involves SNe Ia and SNe II yields and the corresponding delay times for the SNe Ia explosions.…”

The metal enrichment of passive galaxies in cosmological simulations of galaxy formation

“…This assumption, together with quasar-mode AGN feedback and interaction-triggered starbursts, yields a better agreement with the [α/Fe]-σ and mass-metallicity relations than the case where the Salpeter IMF is assumed for all star formation activities. Calura & Menci (2011) employed the same model as Calura & Menci (2009), but computed chemical evolution using individual star formation histories of all progenitors of a galaxy considered. They claimed that inclusion of quasar-mode AGN feedback which quenches late star formation in large haloes and starbursts triggered by galaxy harassment, which boosts star formation activity at high redshift, is crucial and the variable IMF used in Calura & Menci (2009) is not necessary to explain the observed [α/Fe]-σ relation, although they did not show the mass-metallicity relation predicted by their model.…”