Quenched today, but star forming above z > 3
At z = 0, 93 per cent of the 352 TNG-Cluster BCGs are quenched, and only 23 remain star forming. The quenched galaxies have low central gas densities, while the star-forming ones retain high gas densities and disc-like morphologies in their inner regions.
That present-day split is not how the population began. At redshift z > 3, all main progenitors of these BCGs were still star forming. The transition to quiescence therefore happened later, and the paper's central question is what set its timing.
Rejuvenation episodes break the single-quenching picture
Individual BCG star-formation histories are complex rather than monotonic. Since z = 3, BCGs rejuvenate on average 2.1 ± 1.8 times, and typically undergo between one and three rejuvenation episodes. These episodes last about 500 Myr on average and become less frequent with cosmic time, while their duration is roughly independent of redshift.
The fraction of permanently quenched BCGs grows monotonically below z < 3. So the population-level trend toward quiescence coexists with repeated, object-to-object reversals, which is why the authors describe BCG histories as diverse rather than as a single rapid transition.
First and last quenching happen at different times
First quenching occurs on average 9.9 +0.9 −1.4 Gyr ago, corresponding to z = 1.7 +0.5 −0.5. The last and final quenching event is much later and far more broadly distributed: on average 7.0 +2.8 −5.0 Gyr ago, at z = 0.8 +0.8 −0.6, with an almost flat distribution from z ∼ 2 to today.
These two epochs answer different questions. First quenching marks when a BCG first leaves the star-forming population; last quenching marks when it stops rejuvenating. Treating them as one event would hide the long, varied interval in between.
No characteristic mass threshold for quenching
At first quenching, the halo, stellar, and SMBH masses show large scatter: log(M200c/M⊙) = 13.8 +0.4 −0.5, log(M*/M⊙) = 11.9 +0.3 −0.4, and log(MSMBH/M⊙) = 8.9 +0.4 −0.2. At last quenching the values are log(M200c/M⊙) = 14.2 +0.5 −0.7, log(M*/M⊙) = 12.0 +0.3 −0.3, and log(MSMBH/M⊙) = 9.3 +0.5 −0.5, again with large scatter.
The paper reads this spread as evidence that there is no characteristic mass at which quenching occurs. The SMBH mass at first quenching is almost an order of magnitude higher than the mass at which the SMBH transitioned into kinetic feedback mode, so the feedback-mode transition itself is not the trigger either.
Quenching waits for cumulative energy to exceed binding energy
The paper's proposed mechanism is energetic rather than mass-based. First quenching happens when the ratio between the cumulative kinetic energy injected by the SMBH and the BCG binding energy reaches β = 0.9 +2.2 −0.7, the point at which the injected energy becomes larger than the binding energy and gas is expelled from the BCG.
Timing matters for how that threshold is reached. First quenching occurs on average 2.3 +1.2 −1.2 Gyr after the first kinetic energy injection, so the first injection does not quench the galaxy. BCGs with gentler energy injection over a longer period reach β ≈ 1 on average, whereas BCGs with a rapid energy buildup can quench at β < 1, indicating that a more violent short injection can shut down star formation before enough energy is injected to fully unbind the gas.
Gas removal links feedback to star formation shutdown
In the IllustrisTNG model used by TNG-Cluster, SMBH feedback operates in two modes set by the accretion rate relative to the Eddington limit. At low accretion rates, feedback is kinetic, injecting energy as discrete momentum pulses in random directions; at higher accretion rates, it switches to a less efficient thermal mode that continuously injects thermal energy locally.
Kinetic feedback both ejects gas and suppresses later cooling and inflows. The paper's gas maps support this chain: quenched BCGs at z = 0 have lower central gas densities than star-forming BCGs, consistent with gas expulsion removing the fuel for star formation. The wide scatter in β at first quenching shows, however, that gas removal and quenching need not align perfectly in time.
Mergers modulate star formation without explaining first quenching
Rejuvenation requires a renewed cold-gas supply, and gas-rich mergers are one candidate. In the two haloes shown in the paper's gas-density sequences, gas density rises before rejuvenation and disc features appear once rejuvenation begins. Some rejuvenation events are preceded by one or several halo merger events, but others have no recent merger.
The paper therefore does not find a clear correlation between halo mergers and star-formation spikes. The merger catalogues used here define events on a halo scale, while much smaller radii are relevant for BCG rejuvenation, and not every halo merger immediately affects the gas at the cluster centre. The authors note that additional parameters, such as the gas-richness of the merger, would be needed to test the connection further.
What the TNG-Cluster result does and does not settle
The paper extends earlier IllustrisTNG evidence that kinetic-mode SMBH feedback drives quenching to the BCG regime, but it reframes the timing: quenching is not a rapid transition tied to a single event or mass threshold, but a process that develops over several billion years with significant object-to-object variation.
The scope is set by the simulation itself. TNG-Cluster provides 352 BCGs at z = 0 with stellar masses M*,<30 pkpc = 10^11.6–12.9 M⊙ and halo masses M200c = 10^14.3–15.4 M⊙, analysed between z = 6.5 and z = 0, with BCGs identified as the first subhalo in each of the 352 zoom-in haloes. The quenching criterion, the SMBH selection within 30 pkpc, and the halo-scale merger definition all shape the measured timings.
The most decision-relevant result is the β threshold rather than any mass cut: in this simulation, a BCG's first quenching is best predicted by whether its cumulative SMBH kinetic energy has caught up with its own binding energy, which is why the paper finds no characteristic quenching mass and why the delay after the first energy injection is a central quantity to carry forward.
Sources used in this answer
The star-forming past and quenching of brightest cluster galaxies in TNG-Cluster
In TNG-Cluster, 93 per cent of 352 brightest cluster galaxies are quenched at z = 0, but their star-formation histories include multiple rejuvenation episodes; first quenching occurs when cumulative SMBH kinetic energy reaches the BCG binding energy, about 2.3 Gyr after the first kinetic energy injection, with no characteristic quenching mass.