Oxygen-rich droplets and the enrichment of the ISM

TL;DR

This paper proposes that metal-rich droplets formed from supernova ejecta explain the observed abundance discrepancies in HII regions, linking ISM enrichment processes with emission line observations.

Contribution

It introduces a scenario where supernova-enriched droplets cause ORL-CEL abundance discrepancies, connecting stellar feedback with nebular emission line diagnostics.

Findings

Metal-rich droplets form after supernova explosions.

These droplets cause enhanced recombination lines of metals.

The scenario explains the ORL-CEL abundance discrepancy.

Abstract

We argue that the discrepancies observed in HII regions between abundances derived from optical recombination lines (ORLs) and collisionally excited lines (CELs) might well be the signature of a scenario of the enrichment of the interstellar medium (ISM) proposed by Tenorio-Tagle (1996). In this scenario, the fresh oxygen released during massive supernova explosions is confined within the hot superbubbles as long as supernovae continue to explode. Only after the last massive supernova explosion, the metal-rich gas starts cooling down and falls on the galaxy within metal-rich droplets. Full mixing of these metal-rich droplets and the ISM occurs during photoionization by the next generations of massive stars. During this process, the metal-rich droplets give rise to strong recombination lines of the metals, leading to the observed ORL-CEL discrepancy. (The full version of this work is submitted to Astronomy and Astrophysics.)

Figures

Full text

1 Introduction

There is no doubt that galaxies suffer chemical enrichment during their lives (see e.g. Cid Fernandes et al. 2006 for a recent systematic approach using a large data base of galaxies from the Sloan Digital Survey Data Release 5 – Adelman-Mac Carthy J.K. et al., 2007). The main source of oxygen production has since long been identified as due to supernovae from massive stars (type II supernovae). Yet, the exact process by which chemical enrichment proceeds is poorly known (see a review by Scalo & Elmegreen 2004).

Ten years ago, Tenorio-Tagle (1996, hereafter T-T96) proposed a scenario in which the metal-enhanced ejecta from supernovae follow a long excursion in galactic haloes before falling down on the galaxies in the form of oxygen-rich droplets.

In the present work (the full version of which has been submitted to Astronomy & Astrophysics, Stasińska et al. 2007), we suggest that the discrepancy between the oxygen abundances derived from optical recombination lines (ORLs) and from collisionally excited lines (CELs) in HII regions (see e.g. García-Rojas et al. 2006 and references therein) might well be the signature of those oxygen-rich droplets. In fact, Tsamis et al. (2003) and Péquignot & Tsamis (2005) already suggested that the ORL/CEL discrepancy in HII regions is the result of inhomegeneities in the chemical composition in these objects. Our aim is to explicit the link between the ORL/CEL discrepancy and the T-T96 scenario, and to check whether what is known of the oxygen yields allows one to explain the ORL/CEL discrepancy in a quantitative way.

2 The Tenorio-Tagle (1996) scenario

Figures 1–5 present the T-T96 scenario in cartoon format.

3 The ORL-CEL discrepancy in the context of the T-T96 scenario

The details of the physical arguments concerning the amount of oxygen available in the droplets, the mixing processes, as well as the simulation of the ORL-CEL discrepancy with a multizone photoionization model are described in Stasińska et al. (2007). Here, we simply give the most important conclusions.

Photoionization of the oxygen-rich droplets predicted by the T-T96 scenario can reproduce the observed abundance discrepancy factors (ADFs, i.e. the ratios of abundances obtained from ORLs and from CELs) derived for Galactic and extragalactic HII regions. The recombination lines arising from the highly metallic droplets thus show mixing at work.

We find that, if our scenario holds, the recombination lines strongly overestimate the metallicities of the fully mixed HII regions. The collisionally excited lines may also overestimate them, although in much smaller proportion. In absence of any recipe to correct for these biases, we recommend to discard objects showing large ADFs to probe the chemical evolution of galaxies.

To proceed further with this question of inhomogeneities, one needs as many observational constraints as possible.

On the theoretical side, one needs more robust estimates of the integrated stellar yields as well as a better knowledge of the impact of massive stars on the ISM and of the role of turbulence. All these issues are relevant to our understanding of the metal enrichment of the Universe.