Quantifying AGN-Driven Metal-Enhanced Outflows in Chemodynamical Simulations

Philip Taylor, Chiaki Kobayashi

Research output: Contribution to journalArticlepeer-review

22 Citations (Scopus)
42 Downloads (Pure)


We show the effects of AGN-driven outflows on the ejection of heavy elements using our cosmological simulations, where super-massive black holes originate from the first stars. In the most massive galaxy, we have identified two strong outflows unambiguously driven by AGN feedback. These outflows have a speed greater than $\sim 8000$ km\,s$^{-1}$ near the AGN, and travel out to a half Mpc with $\sim 3000$ km\,s$^{-1}$. These outflows remove the remaining gas ($\sim 3$ per cent of baryons) and significant amounts of metals ($\sim 2$ per cent of total produced metals) from the host galaxy, chemically enriching the circumgalactic medium (CGM) and the intergalactic medium (IGM). 17.6 per cent of metals from this galaxy, and 18.4 per cent of total produced metals in the simulation, end up in the CGM and IGM, respectively. The metallicities of the CGM and IGM are higher with AGN feedback, while the mass--metallicity relation of galaxies is not affected very much. We also find `selective' mass-loss where iron is more effectively ejected than oxygen because of the time-delay of Type Ia Supernovae. AGN-driven outflows play an essential role not only in quenching of star formation in massive galaxies to match with observed down-sizing phenomena, but also in a large-scale chemical enrichment in the Universe. Observational constraints of metallicities and elemental abundance ratios in outflows are important to test the modelling of AGN feedback in galaxy formation.
Original languageEnglish
Pages (from-to)L59-L63
Number of pages4
JournalMonthly Notices of the Royal Astronomical Society
Issue number1
Early online date22 Jul 2015
Publication statusPublished - 30 Sept 2015


  • astro-ph.GA


Dive into the research topics of 'Quantifying AGN-Driven Metal-Enhanced Outflows in Chemodynamical Simulations'. Together they form a unique fingerprint.

Cite this