Is feedback-free star formation possible?

A. Ferrara, D. Manzoni, E. Ntormousi · 2025

Read Is feedback-free star formation possible? by A. Ferrara, D. Manzoni, E. Ntormousi (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-02566.txt.

Introduction

A. Ferrara, D. Manzoni, and E. Ntormousi present a 2025 astrophysics paper challenging the idea of feedback-free star formation, where a dense cloud supposedly turns much of its gas into stars before massive stars explode. The proposed window lasts only a few million years.

The authors show that radiation pressure from Lyman-alpha photons, produced before any supernova, can break those conditions. For steady star formation, their calculations cap the star formation efficiency at low values across a wide range of cloud densities. Higher values need the driven shells to fragment and form stars themselves. Even under generous assumptions, the quoted efficiencies are upper limits. Dust, ionised regions, and turbulence were also considered, and the conclusion held.

Catalog record

CREATORA. Ferrara, D. Manzoni, E. Ntormousi
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE45 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-02566

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Is feedback-free star formation possible? A. Ferrara, D. Manzoni, E. Ntormousi Abstract It has been suggested that, if the free-fall time of star-forming clouds is shorter than the lifetime ( ≈ 3 \approx 3 Myr) of massive stars exploding as supernovae (SN), a large fraction of the cloud gas can be converted into stars during an allegedly ‘feedback-free’ phase. Here, we show that radiation pressure from Ly α \alpha photons produced in the pre-SN phase can instead erase feedback-free conditions, and severely limit the star formation efficiency (SFE). We find that, for a constant star formation rate, all clouds with gas surface density ( 37 − 1.7 × 10 5 ) ​ M ⊙ ​ pc − 2 (37-1.7\times 10^{5})\ M_{\odot}\ \rm pc^{-2} have ϵ ∗ < 0.08 \epsilon_{*}<0.08 . Higher SFE values can only be reached if Ly α \alpha -driven shells fragment and form stars. While advanced RHD simulations are required to establish the importance of this effect, adopting an optimistic guess, we find that the SFE increases with cloud surface density, rising from ϵ ∗ = 0.023 \epsilon_{*}=0.023 at Σ g = 37 ​ M ⊙ ​ pc − 2 \Sigma_{g}=37\ M_{\odot}\ \mathrm{pc^{-2}} to ϵ ∗ = 0.27 \epsilon_{*}=0.27 at Σ g = 1.7 × 10 5 ​ M ⊙ ​ pc − 2 \Sigma_{g}=1.7\times 10^{5}\ M_{\odot}\ \mathrm{pc^{-2}} . Given the optimistic assumptions adopted, these numbers should be regarded as upper limits. We conclude that Ly α \alpha radiation pressure strongly limits the SFE, even at solar metallicities, erasing the possibility that a feedback-free star formation mode with ϵ ∗ ∼ > 0.4 \epsilon_{*}\lower 2.15277pt\hbox{$\;\mathrel{\mathop{\kern 0.0pt\sim}\limits^{>}}\;$}0.4 exists in the pre-SN phase. This conclusion remains valid even when other effects such as dust destruction of Ly α \alpha photons, presence of H II \scriptstyle\rm II\ regions, velocity gradients, atomic recoil, and turbulence are considered. I Introduction The hydrogen Lyman- \alpha (Ly \alpha ) line is often the most prominent emission feature ( Osterbrock & Ferland 2006 ) in the spectra of galaxies and has historically been used as a key tracer of star-forming systems ( Partridge & Peebles 1967 ; Djorgovski & Thompson 1992 ; Rhoads et al. 2000 ; Taniguchi et al. 2005 ; Ouchi et al. 2009 ; Hu et al. 2010 ; Pentericci et al. 2011 ; Kashikawa et al. 2011 ; Ouchi et al. 2018 ; Shibuya et al. 2018 ) . Surveys such as HETDEX ( Hill & HETDEX Consortium 2016 ) , VLT/MUSE ( Herenz et al. 2019 ) , and SILVERRUSH ( Ouchi et al. 2018 ; Shibuya et al. 2018 ; Shibuya et al. 2019 ) have also revealed a significant population of Ly \alpha emitters across a wide redshift range. In comparison, much less attention has been paid to the dynamical role of Ly \alpha photons. As Ly \alpha often carries a significant fraction of the bolometric luminosity in young, metal-poor galaxies, it has long been suspected that its radiation pressure could significantly impact gas dynamics ( Cox 1985 ; Bithell 1990 ) . This physical effect might play a pivotal role in our understanding of the earliest galaxies now routinely investigated by the James Webb Space Telescope (JWST). One of the major JWST findings so far has been the discovery of a population of surprisingly bright, blue galaxies at z\gtrsim 10 ( Naidu et al. 2022 ; Harikane et al. 2023 ; McLeod et al. 2024 ; Robertson et al. 2024 ) . These sources, often termed “blue monsters,” exhibit ultraviolet luminosities and colors that challenge existing models of early galaxy formation ( Mason et al. 2023 ; Mirocha & Furlanetto 2023 ) . To reconcile observations with theory, several mechanisms have been proposed: extremely low dust attenuation ( Ferrara et al. 2023 ) , top-heavy (or even PopIII-dominated Maiolino et al. 2024 ) initial mass functions ( Trinca et al. 2024 ; Schaerer et al. 2024 ) , or modified cosmological initial conditions ( Liu & Bromm 2022 ; Padmanabhan & Loeb 2023 ) . A particularly straightforward explanation suggests that such galaxies formed stars with extremely high efficiency, potentially converting nearly all their gas mass into stars during short-lived, ‘feedback-free’ bursts ( Dekel et al. 2023 ; Li et al. 2023a , FFB,) . In this scenario, if the free-fall time of a star-forming molecular cloud is shorter than the delay before the first supernovae (typically \sim 3 – 5 Myr), star formation proceeds essentially unregulated, yielding star formation efficiencies (SFEs) approaching unity. Analytical arguments and radiation-hydrodynamic (RHD) simulations have claimed that early feedback processes—such as photoionization heating and UV radiation pressure on dust—are insufficient to halt star formation under these conditions ( Li et al. 2023a ; Menon et al. 2023 ) . However, the FFB scenario neglects the momentum imparted by resonantly trapped Ly \alpha photons, which can become dynamically dominant well before the first SNe. At the extremely high neutral hydrogen column densities expected in compact, metal-poor systems, Ly \alpha photons undergo \gg 10^{6} scatterings, with the resulting radiation pressure capable of disrupting gas clouds and quenching further star formation ( Tomaselli & Ferrara 2021 ; Kimm et al. 2018a ; Nebrin et al. 2024 ) . Crucially, this occurs on timescales shorter than the SN delay time, contradicting the key assumption of the FFB model. As Ly \alpha often carries a significant fraction of the bolometric luminosity in young, metal-poor galaxies, it has long been suspected that its radiation pressure could impact gas dynamics ( Adams 1972 ; Harrington 1973 ; Neufeld 1990 ; Tan & McKee 2003 ; Oh & Haiman 2002 ) . Early analytic and numerical studies suggested that trapped Ly \alpha photons could slow accretion onto protostars and dark matter halos, while Monte Carlo radiative transfer simulations showed that the momentum transfer from multiple scatterings can drive supersonic outflows ( Dijkstra & Loeb 2008 ; Dijkstra & Loeb 2009a ) . The dynamical influence ot Ly \alpha photons is often characterized by the force multiplier M_{F}\equiv\dot{p}_{\alpha}/(L_{\alpha}/c) , which quantifies how photon trapping boosts the radiation force. In static, dust-free media, this multiplier scales as M_{F}\propto\tau_{0}^{1/3} , where \tau_{0} is the Ly \alpha line-center optical depth ( Adams 1972 ; Smith et al. 2017 ; Kimm et al. 2018a ) . For typical star-forming clouds at high redshift, the optical depths are enormous ( \log\tau_{0}\gtrsim 10 ), leading to M_{F}\sim 100 or higher ( Tomaselli & Ferrara 2021 ; Nebrin et al. 2024 ) , even in the presence of a moderate dust content ( Tomaselli & Ferrara 2021 ) . The first hydrodynamical simulations explicitly incorporating Ly \alpha momentum coupling found that it can significantly alter the dynamics of early star-forming clouds. Smith et al. 2016 ; Smith et al. 2017 ; Smith et al. 2018 ; Smith et al. 2019 showed that Ly \alpha pressure can drive gas shell expansion faster than ionizing radiation alone, while Kimm et al. 2018a implemented a subgrid model calibrated via MCRT simulations using the RASCAS code ( Michel-Dansac et al. 2020 ) , accounting for recoil, dust, and deuterium effects. Their results suggested that Ly \alpha feedback can potentially reduce star formation efficiency (SFE) and regulate cloud evolution prior to supernova

Read the complete text