From FFB Starbursts at Cosmic Dawn to Quenching at Cosmic Morning: Hi-z Galaxy Bimodality

TL;DR

The paper proposes a mass-dependent bimodality in the early evolution of galaxies, distinguishing feedback-free starburst (FFB) halos from non-FFB halos and tracing how these tracks connect the bright galaxies observed at cosmic dawn ($z>8$) to massive quiescent galaxies and supermassive BHs at cosmic morning ($z\sim4-7$) while predicting a peak star-forming phase at cosmic noon. It argues that FFB halos, residing in high-$\\sigma$ peaks with $M_v\gtrsim10^{10.5}M_\\odot$ at $z\sim10$, undergo intense, bursty star formation followed by abrupt post-FFB quenching driven by compaction, hot CGM, and accelerated BH growth; non-FFB halos grow more gradually and fuel peak star formation around $z\sim2-3$ as cold streams remain able to feed via CGM entrainment. The paper explores multiple quenching pathways, notably compaction-driven gas depletion, AGN feedback, and disruption of cold streams through CGM turbulence or AGN photo-heating, acknowledging that fully shutting off gas supply by streams remains a challenging open problem requiring higher-resolution simulations. Observable predictions include a morphology bimodality at cosmic dawn and the existence of massive, compact post-FFB quiescent galaxies at cosmic morning with $n\gtrsim10^{-5}\,\mathrm{Mpc}^{-3}$ and UVLF peaks near $M_{uv}\sim-22.7$ (z~4) and $-21$ (z~6), offering concrete targets for JWST surveys.

Abstract

We propose a mass-dependent bimodality in the early evolution of galaxies. The massive track connects the super-bright galaxies at cosmic dawn ($z > 8$) to the super-massive quiescent galaxies and black holes (BHs) at cosmic morning ($z \sim 4 - 7$). The dark-matter halos $> 10^{10.5} {\rm M}_\odot$ at $z = 10$ are expected to undergo feedback-free starbursts (FFB) with high star-formation efficiency in dense star clusters within compact galaxies. The less massive halos avoid FFB and form stars gradually under stellar feedback, possibly leading to the peak star-forming galaxies at cosmic noon ($z \sim 1-3$). The FFB and non-FFB halos originate from $>4σ$ and $2-3σ$ density peaks, respectively. The post-FFB galaxies quench their star formation soon after the FFB phase and remain quiescent due to (a) gas depletion by the FFB starbursts and outflows, (b) compaction events driven by angular-momentum loss in colliding streams within the high-sigma-peak FFB halos, (c) turbulent circum-galactic medium (CGM) that suppresses feeding by cold streams, and (d) BH feedback, being a key for complete quenching. BH feedback is enhanced by FFB-driven BH seeding and growth. It seems capable of disrupting the streams by generating CGM turbulence or photo-heating, but this remains an open challenge. The cosmic-morning quiescent galaxies are expected to be massive, compact, showing signatures of compaction, outflows and AGN, with a comoving number density $\sim 10^{-5} {\rm Mpc}^{-3}$, comparable to the super-bright galaxies at cosmic dawn and the AGN at cosmic morning. Their UV luminosity function is predicted to peak about $M_ {\rm uv} \sim -22$ and contribute $\sim 10\%$ of the galaxies there.

Figures (13)

• Figure 1: FFB-like starbursts in the histories of massive quiescent galaxies. Shown are star formation histories as estimated from the spectra of six characteristic massive quiescent galaxies observed by JWST at cosmic morning, $z \!=\! 4\!-\! 7$ (marked by circles). The galaxies are ZF-UDS-6496 at $z \!=\! 3.99$carnall24, PRIMER-EXCELS-117560 at $z \!=\! 4.619$carnall24, PRIMER-EXCELS-109760 at $z \!=\! 4.623$carnall24, GS-9209 at $z \!=\! 4.66$carnall23, RUBIES-EGS-QG-1 at $z \!=\! 4.9$degraaff24, and RUBIES-UDS-QG-z7 at $z \!=\! 7.3$weibel25. Small vertical offsets are introduced to enable distinction between the curves below $10 M_\odot\,{\rm yr}^{-1}$, where they are below the detection limit. Each recovered SFH seems to consist of a starburst of a global duration $100\!-\! 200\,{\rm Myr}$, at a level of a few hundred $M_\odot \,{\rm yr}^{-1}$, followed by a robust quiescent period of several hundred Megayears till the observed time. The crudely estimated stellar masses are between $10^{10.3}M_\odot$ and $10^{11}M_\odot$ for RUBIES-UDS-QG-z7 and RUBIES-UDS-QG-1 respectively. The surface densities are well above $10^{4}M_\odot\,{\rm pc}^{-2}$weibel25. The estimated cosmological number density of these galaxies is between $10^{-6}$ and $10^{-4} \,{\rm Mpc}^{-3}$, well above the expectations from current cosmological simulations weibel25.
• Figure 2: Bimodal evolutionary tracks of galaxies: FFB and non-FFB. Shown in a diagram of halo mass versus redshift are average growth tracks for halos that at $z_0 \!=\! 10$ have masses $M_{\rm v} \!=\! 10^{9.84}, 10^{10.54}, 10^{11.24}M_\odot$ (three blue thick diagonal lines). In comparison, the threshold for FFB in a disk (red curve, marked "FFB") consists of thresholds for gas density $n$ and surface density $\Sigma$ following Fig. 6 of dekel23. The $N \sigma$ peaks in the density fluctuation field (dashed grey curves) indicate that the middle halo is roughly a $4\sigma$ peak. Galaxy comoving number densities of $10^{-4}$ and $10^{-6} \,{\rm Mpc}^{-3}$ are marked (thin dashed orange curves), confining the middle halo at $z \!\leq\! 11$. We learn that the halos of mass $\geq\!10^{10.5}M_\odot$ at $z \!=\! 10$ ($\geq\!4\sigma$ peaks) undergo an FFB phase, while halos of lower masses (lower-$\sigma$ peaks) never become FFB. The "fiducial", most abundant FFB galaxies, those near the middle blue curve, enter the FFB phase near $z \!\sim\! 11$ (rightmost open black circle) and exit at $z \!\sim\! 8$ (leftmost open black circle), after $\sim\! 170\,{\rm Myr}$ and when $M_{\rm v} \!\hbox{$\; \buildrel > \over \sim \;$}\! 10^{11}M_\odot$. Soon after, the fiducial FFB halos cross the "golden mass" dekel19_gold$M_{\rm v} \!\sim\! 10^{11.5}M_\odot$, in the shaded area between the horizontal dashed lines that marks the onset of quenching by several different mechanisms. These include a hot CGM db06, wet compaction events zolotov15lapiner23, the rapid growth of super-massive BHs and AGN lapiner21, as well as long-lived extended disks and rings dekel20_flipdekel20_ring. The quenching of the FFB galaxies is efficient because of the intense compaction activity in high-sigma halos dubois12 and the compaction-driven AGN feedback lapiner21 that is boosted by FFB BH seeding and growth dekel25. By cosmic morning, at $z \!\sim\! 5$, these quenching mechanisms lead to quiescent galaxies of stellar mass $M_{*} \!\sim\! 10^{10.5}M_\odot$ in halos of $M_{\rm v} \!\hbox{$\; \buildrel > \over \sim \;$}\! 10^{12}M_\odot$ (black square), as observed. These halos eventually become quiescent groups and clusters of galaxies. The lower-sigma halos with masses below the FFB threshold accrete gas and form stars more gradually as they grow. At cosmic noon, $z \!\sim\! 2\!-\! 3$, when these halos are above $10^{12}M_\odot$, cold gas streams penetrate efficiently through the hot CGM dekel09, boosted by entrainment of hot CGM gas onto the cold streams aung24, giving rise to star forming galaxies at their peak (blue diamond). At $z \!\leq\! 1$, cold streams do not penetrate the hot halos db06, such that these galaxies end up with lower SFRs.
• Figure 3: Two evolutionary tracks in the plane of halo mass function, comoving number density versus halo mass, complementing Fig. \ref{['fig:post']}. The halo mass functions at different redshifts are shown in the background, based on watson13. The red curve (the bottom curve at high masses) refers to the evolving fiducial FFB halos that reach $\sim\! 10^{10.5}M_\odot$ at $z\!=\! 10$ (roughly $4\sigma$ peaks). The cyan curve(the bottom curve at low masses) is the threshold for FFB, $M_{{\rm v},10.8} (1+z)_{10}^{6.2} \!=\! 0.8$. The curves stretching on the left of $z \!=\! 8$ (dashed, $M_{\rm v} \!<\! 10^{11.2}M_\odot$) refer to the FFB phase, and the curves to the right of $z \!=\! 8$ (solid, $M_{\rm v} \!>\! 10^{11.2}M_\odot$) correspond to the post-FFB quiescent phase. The typical halo number density in the post-FFB zone is $n \!\hbox{$\; \buildrel > \over \sim \;$}\! 10^{-5}\,{\rm Mpc}^{-3}$. The uppermost, blue curve near $n \!\sim\! 10^{-3}\,{\rm Mpc}^{-3}$ corresponds to the non-FFB evolution track reaching the peak of star-forming galaxies at $z \!\sim\! 2$. Colored symbols mark the fiducial starbursting FFB halos at cosmic dawn ($M_{\rm v} \!\sim\! 10^{10.5}M_\odot$ at $z \!\sim\! 10$ where $n \!\sim\! 10^{-4.3}\,{\rm Mpc}^{-3}$), the fiducial quiescent halos at cosmic morning ($M_{\rm v} \!\sim\! 10^{12.2}M_\odot$ at $z \!\sim\! 5$ where $n \!\sim\! 10^{-5.3}\,{\rm Mpc}^{-3}$), and the fiducial non-FFB star-forming galaxies at cosmic noon ($M_{\rm v} \!\sim\! 10^{12.6}M_\odot$ at $z \!\sim\! 2$ where $n \!\sim\! 10^{-3.5}\,{\rm Mpc}^{-3}$). Shown for comparison (black symbols with error bars) are number densities of AGN observed by JWST harikane23_agn at $z \!=\! 5 \!-\! 7$ in quiescent galaxies brighter than the corresponding predicted UV magnitudes of post-FFB galaxies with quenching onset at $z_{\rm quench} \!=\! 8$ or $6.5$ (see Fig. \ref{['fig:Muv']}). The abundance of BHs is in the ball park of that predicted for the quiescent galaxies when the onset of quenching is at $z \!\sim\! 6.5$, and is somewhat higher for higher $z_{\rm quench}$. (See also the number densities in Fig. \ref{['fig:n_post']} below.)
• Figure 4: Compaction-driven quenching without (top) and with (bottom) AGN feedback. Shown are the evolution of quantities of interest in the VELA simulations without AGN feedback lapiner23 (top) and in the NewHorizon simulations that incorporate AGN feedback lapiner21 (bottom). The compaction event is characterized by the peak in gas mass within the central $1\,{\rm kpc}$ (blue dashed line, peak marked by a vertical line) and the subsequent decay of the BH orbit to the center lapiner21 (green, thin line, right axis for BH orbital radius). Following the compaction, in the presence of AGN the SFR drops (magenta, gradually declining line, right axis for sSFR) and the stellar mass (solid red line at the top) tends to remain rather constant over the whole galaxy (as well as in the inner $1\,{\rm kpc}$, dashed red). Galaxy NH6, with two successive compaction events, shows two successive quenching events. On the contrary, in the absence of AGN feedback, the quenching is only partial, limited to the inner $1\,{\rm kpc}$, while the stellar mass over the whole galaxy keeps growing. We learn that AGN feedback is crucial for the deep quenching at cosmic morning, consistent with the findings of other simulations (see text). We thus suspect that the boosted BH growth due to the FFB phase may be a key for post-FFB quenching.
• Figure 5: Stream disruption by CGM turbulence. Shown is an estimate of the critical stream radius for disruption by CGM turbulence, ${R}_{\rm turb}$, with respect to the actual stream radius, as a function of the turbulent Mach number. The transonic point is marked by a vertical dashed line. This estimate is based on a crude extrapolation of the results for clouds gronke22 to streams. The cases shown refer to stream properties in $M_{\rm v} \!\sim\! 10^{12}M_\odot$ halos during the central phase of the post-FFB quiescent regime at cosmic morning, $z \!\sim\! 6$. The relevant parameters of the four cases are listed in Table \ref{['tab:models']}. We learn that, with respect to ${R}_{\rm turb}/R_{\rm s} \!=\! 1$ (the horizontal line separating the shaded and un-shaded areas), the slow cooling stream is expected to be capable of being disrupted even by subsonic turbulence. The low gas stream may disrupt even by transonic or slightly supersonic turbulence On the other hand, the fiducial stream and the fast cooling stream seem to require supersonic turbulence of $\mathcal{M}_{\rm t} \!\sim\! 3$ for disruption. With respect to the more relaxed threshold ${R}_{\rm turb}/R_{\rm s} \!=\! 0.05$ (which is motivated by observations - the horizontal doted line), the fiducial streams are expected to be disrupted even by transonic turbulence.