[Paper Review] On the origin of star formation quenching in massive galaxies at $z \gtrsim 3$ in the cosmological simulations IllustrisTNG

Shalini Kurinchi-Vendhan, Marion Farcy+2 more|arXiv (Cornell University)|Oct 4, 2023

Galaxies: Formation, Evolution, Phenomena6 citations

TL;DR

Using the IllustrisTNG cosmological simulations, this study identifies active galactic nucleus (AGN) feedback—specifically the kinetic, jet-mode feedback—as the primary driver of early star-formation quenching in massive galaxies at $z \gtrsim 3$. Quenched galaxies experience earlier and stronger AGN feedback due to faster black hole growth, which suppresses cold gas inflows and maintains low gas fractions, leading to quiescence.

ABSTRACT

Using the cosmological simulations IllustrisTNG, we perform a comprehensive analysis of quiescent, massive galaxies at $z \gtrsim 3$. The goal is to understand what suppresses their star formation so early in cosmic time, and how other similar mass galaxies remain highly star-forming. As a first-order result, the simulations are able to produce massive, quiescent galaxies in this high-redshift regime. We find that active galactic nuclei (AGN) feedback is the primary cause of halting star formation in early, massive galaxies. Not only do the central, supermassive black holes (SMBHs) of the quenched galaxies have earlier seed times, but they also grow faster than in star-forming galaxies. As a result, the quenched galaxies are exposed to AGN feedback for longer, and experience the kinetic, jet-mode of the AGN feedback earlier than the star-forming galaxies. The release of kinetic energy reduces inflows of gas while likely maintaining outflows, which keeps a low cold gas fraction and decreases the star formation of the galaxies down to a state of quiescence. In addition to AGN feedback, we also investigate the influence of the large-scale environment. While mergers do not play a significant role in the quenching process, the quenched galaxies tend to reside in more massive halos and denser regions during their evolution. As this provides a greater initial amount of infalling gas to the galaxies, the large-scale environment can mildly affect the fate of the central SMBH growth and, via AGN feedback, contribute to star formation quenching.

Motivation & Objective

• To understand the physical mechanisms responsible for quenching star formation in massive galaxies at high redshift ($z \gtrsim 3$), where observations reveal unexpectedly early quiescence.
• To determine whether internal processes like AGN feedback or external factors such as large-scale environment or mergers are primarily responsible for early quenching.
• To investigate the role of black hole growth and feedback mode in triggering quiescence, particularly in comparison to stellar feedback and cosmological starvation.
• To assess the relative importance of merger events and environmental density in shaping the quenching pathway of massive galaxies at high redshift.
• To evaluate how the timing and efficiency of AGN feedback models in simulations influence the emergence of quiescent galaxies at early cosmic times.

Proposed method

• Utilized the IllustrisTNG cosmological simulation suite (TNG100-1 and TNG300-1) to extract samples of massive, quiescent and star-forming galaxies at $z = 3$ and $z = 3.7$.
• Tracked the accretion history and feedback output of supermassive black holes in quenched versus star-forming galaxies, focusing on the kinetic, jet-mode AGN feedback in the radiatively inefficient regime.
• Compared black hole seed times, accretion rates, and energy output between quenched and star-forming galaxies to assess the timing and intensity of AGN feedback.
• Analyzed host halo mass and large-scale environment density (via dark matter halo properties and proximity to cosmic filaments) to evaluate environmental contributions to quenching.
• Used statistical comparisons and time-series analysis of galaxy and black hole evolution to isolate causal links between AGN feedback and quenching.
• Evaluated the role of mergers by tracking interaction histories and dynamical interactions, finding minimal impact on quenching at $z \gtrsim 3$.

Figure 1: Fraction of quiescent galaxies (top row) and specific star-formation rate (bottom row) versus stellar mass for the TNG100-1 $z=3.7$ (left), TNG100-1 $z=3.0$ (middle), and TNG300-1 (right) samples. In each panel, we compare the quiescent (red shade, triangles) and star-forming (blue shade,

Experimental results

Research questions

• RQ1 What physical mechanism is primarily responsible for quenching star formation in massive galaxies at $z \gtrsim 3$ in the IllustrisTNG simulations?
• RQ2 How does the timing and strength of AGN feedback differ between quenched and star-forming massive galaxies at high redshift?
• RQ3 To what extent do large-scale environmental conditions, such as host halo mass and density, contribute to early black hole growth and subsequent quenching?
• RQ4 Are merger events a significant driver of quenching in massive galaxies at $z \gtrsim 3$, or do they play a minor role?
• RQ5 How does the implementation of kinetic, jet-mode AGN feedback in the simulations influence the emergence of quiescent galaxies at early cosmic times?

Key findings

• AGN feedback, particularly the kinetic, jet-mode feedback in the radiatively inefficient regime, is the dominant mechanism responsible for quenching star formation in massive galaxies at $z \gtrsim 3$.
• Quenched galaxies host black holes with earlier seed times and higher accretion rates than star-forming galaxies, leading to prolonged and stronger AGN feedback exposure.
• The onset of quenching coincides with the triggering of kinetic AGN feedback when black hole mass and accretion rate cross a critical threshold, suppressing cold gas inflows and maintaining low gas fractions.
• Galaxies in denser, more massive host halos accrete gas and dark matter at higher rates, enabling faster black hole growth and earlier activation of AGN feedback, thus contributing to quenching.
• Merger events do not play a significant role in distinguishing quenched from star-forming galaxies at $z \gtrsim 3$, as quenching occurs independently of interaction history.
• The simulated number density of massive quenched galaxies at $z > 3$ is underestimated in current IllustrisTNG models, suggesting that improvements in AGN feedback modeling—especially in the radiatively efficient regime—may be necessary to better match observations.

Figure 2: Relative frequency of galaxies with respect to the critical redshift of BH mass growth. We define the “critical time of BH mass growth” as the time when the galaxy’s central BH mass exceeds $M_{\mathrm{BH}}\geq 10^{6.5}\>\mathrm{M}_{\odot}$ . We show separate histograms for the TNG100-1 $z