MEGATRON: The environments of Population III stars at Cosmic Dawn and their connection to present day galaxies

TL;DR

The study investigates how Population III stars form and persist in a Milky Way–mass progenitor during Cosmic Dawn using the MEGATRON suite of high-resolution, radiation-hydrodynamic simulations that self-consistently model non-equilibrium chemistry and a spatially varying Lyman-Werner background. By resolving gas down to near-pc scales and following both Pop III and Pop II modes across four physics variants, the work shows an initial Pop III phase in minihalos driven by H$_2$ cooling, followed by a transition to atomic-cooling halos as LW feedback builds up, with a global SFR around $10^{-3}\,M_\odot\,\mathrm{yr}^{-1}$ by $z\sim20$. The results reveal rare Pop III starbursts in massive halos (up to $\sim3\times10^8\,M_\odot$) that can host $\sim$20–130 shining Pop III stars, and they trace the later enrichment that leads to Pop II star formation; they also connect the present-day distribution of Pop III remnants to either the main halo or subhalos, with most remnants ending up in the stellar halo. Observationally, Pop III spectra are nebular-dominated with strong H I and He II lines, but the brightest systems are typically too faint for JWST unless aided by gravitational lensing, and proximity to UV-bright galaxies is generally unlikely, underscoring the challenge of direct Pop III detections at high redshift. Overall, the work provides a self-consistent baseline linking the environments of the first stars to the assembly of a Milky Way–like galaxy and informs strategies for near-field cosmology and JWST-era searches.

Abstract

We present results of Pop. III formation in the MEGATRON suite of simulations, which self-consistently follows radiation and non-equilibrium chemistry, and resolves gas at near-pc resolution of a Milky Way-mass halo at Cosmic Dawn. While the very first Pop. III stars form in halos with masses well below the atomic cooling limit, whose cooling is dominated by molecular hydrogen, the majority of Pop. III stars form in more massive systems above the $10^4$~K atomic cooling threshold. The shift in cooling regime of halos hosting new Pop. III stars occurs within $100$ Myr of the first Pop. III star as the Lyman-Werner (LW) background rapidly increases to $10^{-21}\,\rm erg\,s^{-1}\,cm^{-2}\,Hz^{-1}\,sr^{-1}$. We find that the global Pop. III star formation rate stabilizes to a value of $10^{-3}\,\rm M_\odot\,yr^{-1}$ at $z=20$. Among the three processes that quench Pop. III star formation in mini-halos, the LW background, gas starvation, and external chemical enrichment, the LW background is most important. A small fraction of haloes undergo multiple episodes of Pop. III star formation when the earlier forming stars all directly collapse to black holes. If the halos become massive enough, they can form up to $\sim100$ Pop. III stars in a single burst, which may be observable by JWST with moderate gravitational lensing. Pop. III stars form at a wide range of distances from UV-bright galaxies, with only $0.06\%$ of Pop. III stars forming within the virial radius of galaxies with $M_{\rm UV} < -17$. Finally, by tracking Pop. III star remnants down to $z=0$, we find that $75-80\,$% reside in the stellar halo of our simulated Milky Way analogue, while the remainder are gravitationally bound to lower-mass systems, including satellite halos.

Figures (10)

• Figure 1: The Initial Mass Function (IMF) of Pop. III stars used in the megatron suite of simulations. Histograms of Pop. III masses for each run are included to show they all sample the log-normal IMF. Regions in the background are coloured according to the feedback channels a star of particular initial mass will follow, starting with hypernova (HN) and moving to black hole (BH) formation with a narrow mass regime where the star can undergo pair-instability supernova (PISN). The Pop. III model also includes CCSN at lower masses. They are however probabilistically unlikely to be sampled due to the chosen characteristic mass of $\sim 100\,\rm M_\odot$.
• Figure 2: (Top) Star formation rate of Pop. III stars as a function of time from the start of the simulations. The SFR quickly rises to $10^{-3}\,\rm M_\odot\,yr^{-1}$ and remains mostly constant for all simulations until the end of the simulation at $z=8.5$. The gray dashed line at $\sim 1.7\times10^{-4}\;\rm M_\odot \,yr^{-1}$ is the level for the contour in the lower panel. (Bottom) Contours in halo mass and time of halos hosting Pop. III stars weighed by the SFR of Pop. III stars, for each simulation. The mass-time tracks for halos with a virial temperature of $10^4\;$K Barkana_Loeb_2001 are shown as the gray dash-dotted lines, labelled by mean molecular weight ($\mu=1.22$: neutral gas; $\mu=0.61$: ionized gas). For all simulations, Pop. III formation migrates from mini-halos to atomic cooling halos on a short timescale. Pop. III starbursts in $10^8\,\rm M_\odot$ halos are visible for the Efficient SF and Variable IMF runs.
• Figure 3: (left) Projection of the gas density of the Lagrangian volume in the Efficient SF simulation at $z=8.5$. All Pop. III stars which have formed by that redshift are over-plotted on the projection, with symbol sizes corresponding to their age, and colours/markers indicating if the star is still alive or a remnant. Most of the stars older than $5\;\rm Myr$ are now remnants, either a black hole or pair-instability supernova, and are mostly concentrated in the high-density filamentary structure, compared to young Pop. III stars which live on the outskirts. In the insets of the figure, we highlight three different halos displaying the three stages of Pop. III formation: before the star forms, while the star is alive, and the resulting PISN if the star meets the initial mass requirement $(140 < M_\star < 300)$. (right) Each row is for a particular stage of Pop. III evolution. The first column is the gas density projection in the inner part of the halo, the second column is the halo's phase diagram weighed by mass and coloured by dominant cooling process (blue: H$_{2}$, red: atomic, green: metal cooling), and the third column is the halo's intrinsic spectrum built up from nebular emission and continuum (and for the middle row the Pop. III SED.)
• Figure 4: The fraction of molecular hydrogen $f_{\rm H_2}$ in the inner $0.1 r_{\rm vir}$ of halos undergoing their first Pop. III formation (within $\sim5$ Myr), as a function of LW intensity (parametrized as $J_{21}$) in a thin shell at the virial radius of the halo, for the Efficient SF simulation. Halos are coloured by the electron fraction $f_e$ in the inner $0.1 r_{\rm vir}$, and the sizes of points correspond to the age of Universe at that time. Quantities which are computed in the inner halo are density weighted, while quantities computed in the shell are volume weighted. Two distinct populations emerge: halos with low electron fractions ($f_e \eqsim 10^{-4}$, the Pop. III.1 case) forming at early times, and halos with elevated electron fractions ($f_e > 10^{-4}$, the Pop. III.2 case) forming at later times, across a wide range of LW strengths.
• Figure 5: (Top) The fraction of halos containing no stars within their virial radii in the Bursty SF simulation which have density-weighted metallicities $Z > 2\times10^{-8}$ (the threshold for Pop. II formation) as a function of time. Lines are coloured by halo mass. Starless and metal enriched halos become more common over time, and make up $\sim 20-30\%$ of all starless halos at 600 Myr. (Middle) The median gas mass of starless halos at a given time divided by the maximum value reached at any time. Red represents the total gas mass in halos out to the virial radius, and in blue indicates the inner gas mass out to $0.1\,r_{\rm vir}$. (Bottom) The median Lyman Werner intensity across starless halos, as measured at the virial radius, as a function of time. The LW intensity increases rapidly after the first stars form, reaching $J_{21}$ at 200 Myr, which suppresses Pop. III formation in mini-halos.