Hunting for the First Explosions at the High-Redshift Frontier Junehyoung Jeon, Volker Bromm, Alessandra Venditti, Steven L. Finkelstein, Tiger Yu-Yang Hsiao Abstract The James Webb Space Telescope (JWST) has spectroscopically confirmed galaxies up to z ∼ 14 z\sim 14 , 300 Myr after the Big Bang, and several candidates have been discovered at z ∼ 15 − 25 z\sim 15-25 , with one candidate as high as z ∼ 30 z\sim 30 , only 100 Myr after the Big Bang. Such objects are unexpected, since theoretical studies have not predicted the existence of detectable galaxies at z ∼ 30 z\sim 30 . While any z ∼ 30 z\sim 30 candidates may be contaminants at lower redshifts, we explore whether such extreme redshift sources could be consistent with hyper-energetic transient events linked to the formation of the first, metal-free, stars. Specifically, we consider pair-instability supernovae (PISNe), a predicted class of extreme thermonuclear explosions that leave no remnant behind. Using cosmological simulations, we investigate an overdense cosmic region, where star formation and subsequent PISNe occur at z ∼ 30 − 40 z\sim 30-40 , even within standard cosmology. Assessing the likelihood of such a region, the corresponding number of PISNe at z ≳ 20 z\gtrsim 20 , and their observed flux, we find that JWST has a non-negligible chance to detect a PISN event at extremely high redshifts. If a transient event were confirmed at z ∼ 30 z\sim 30 , this would provide a direct glimpse into the epoch of first star formation, dramatically extending the empirical reach of astronomy. I Introduction The frontier of observational cosmology has been pushed to increasingly high redshifts by the James Webb Space Telescope (JWST), with spectroscopically confirmed galaxies out to z\gtrsim 14 ( R. P. Naidu et al. 2025 ; S. Carniani et al. 2024 ) . Recently, photometrically selected candidates have been identified at even earlier times, to z\sim 25 , via the photometric dropout technique ( P. G. Pérez-González et al. 2025 ; M. Castellano et al. 2025 ) , and even z\sim 32 by spectral energy distribution (SED) fitting ( G. Gandolfi et al. 2025 ) , corresponding to only \sim 100 Myr after the Big Bang. If confirmed, these indications of stellar activity in the extremely early Universe would challenge our understanding of first star and galaxy formation ( V. Bromm & N. Yoshida 2011 ; P. Dayal & A. Ferrara 2018 ; R. S. Somerville et al. 2025 ; L. Y. A. Yung et al. 2025 , e.g.,) . The suggested period for these ultra-high-redshift candidates coincides with the formation of the first stars in the Universe. Primordial gas from the Big Bang, composed of hydrogen and helium, will collapse to form metal-free, Population III (Pop III) stars at z\sim 20-30 ( V. Bromm 2013 ; R. S. Klessen & S. C. O. Glover 2023 ) . Pop III stars are predicted to be more massive than the metal-enriched stars at subsequent epochs, reaching masses up to a few 100 M ⊙ ( T. Hosokawa et al. 2011 ; S. Hirano et al. 2014 ; A. Stacy et al. 2016 ; T. Hosokawa et al. 2016 ; K. Sugimura et al. 2020 ; M. A. Latif et al. 2022 ) . However, the inferred total stellar masses of Pop III systems formed within minihalos are low ( \lesssim 10^{3} M ⊙ ), and they are thus not expected to be directly observable at z\gtrsim 20 ( A. T. P. Schauer et al. 2023 ) . Alternatively, Pop III stars may be detectable when they end their lives as hyper-energetic pair-instability supernovae (PISNe). Such an event occurs when a progenitor star with a mass between 140 and 260 M ⊙ undergoes an extreme thermonuclear explosion 1 1 1 For rapidly rotating progenitors, the PISN mass range may extend to lower masses, impacting event rates and observational characteristics ( E. Chatzopoulos & J. C. Wheeler 2012 ; J. Smidt et al. 2015 , e.g.,) . , triggered by electron-positron production in its core ( A. Heger et al. 2003 ) . These reactions cause a rapid loss of radiation pressure, resulting in runaway collapse and the ignition of explosive oxygen and silicon burning that completely disrupts the star, with no remnant being left behind ( Z. Barkat et al. 1967 ; G. Rakavy & G. Shaviv 1967 ; G. S. Fraley 1968 ; J. R. Bond et al. 1984 ; C. L. Fryer et al. 2001 ; K.-J. Chen et al. 2014 ) , reaching absolute UV magnitudes up to -22 , or even brighter values close to the initial peak ( D. Kasen et al. 2011 ; L. Dessart et al. 2013 ; M. S. Gilmer et al. 2017 , e.g.,) . Motivated by the newly discovered ultra-high-redshift candidates, we explore the possibility of JWST detecting such transient phenomena triggered by the first stars at z\sim 30 , following up on earlier studies ( J. A. Hummel et al. 2012 ; R. S. de Souza et al. 2013 ; D. J. Whalen et al. 2013 , e.g.,) . Because of their extreme explosion energies (\sim 10^{53} erg), PISNe may be observable even up to z\sim 30 (see Section IV ). Previous works have argued that PISNe, should they exist, could be observed even up to z\sim 25 ( S. M. Weinmann & S. J. Lilly 2005 ) . The main challenge in observing these events is their small number density: As the predicted visibility time of individual PISN events is short, of order \sim 10 yr in the observed frame ( J. A. Hummel et al. 2012 ) , and given the limited JWST survey area, previous studies found that the probability of detecting a PISN event at z\gtrsim 8 is low, the expected number of events being \lesssim 0.1 ( J. A. Hummel et al. 2012 ; F. Gabrielli et al. 2024 ; A. Venditti et al. 2024 ) . Lower redshifts have been proposed to be more promising for PISN detection with JWST, or with upcoming wide-field Roman Space Telescope surveys ( T. J. Moriya et al. 2022a ; T. J. Moriya et al. 2022b ; E. Regős et al. 2020 ) . The discovery of z\sim 30 candidates prompts us to revisit this question, using cosmological simulations of highly-biased, overdense regions to trace the formation of the first stars in such extreme environments, and to examine the probability of detecting the resulting PISNe at the (high- z ) tail-end of early star formation. We specifically consider these related questions: Does JWST observe a sufficiently large volume, including archival searches, to capture possible host systems for a PISN at extremely high redshifts (Section III.1 )? Does JWST cover sufficiently long periods to detect a PISN event (Section III.2 )? Are PISN events luminous enough for JWST to observe them at such early times (Section III.3 )? II Methodology Table 1: Simulation parameters for the overdense region. Lengths are given in comoving units. \sigma_{8} Box size Particle number (Gas and DM) DM mass Gas mass Stellar mass Minimum gravitational softening length h^{-1} Mpc M ⊙ M ⊙ M ⊙ h^{-1} kpc 1.5 6 2\times 512^{3} 1.76\times 10^{5} 3.16\times 10^{4} 1.98\times 10^{3} 0.5 We run cosmological simulations of biased high-density regions that are capable of producing multiple Pop III stars and the resulting PISNe, employing the gizmo code ( P. F. Hopkins 2015 ) that inherits the gravity solver from the gadget-2 framework ( V. Springel 2005 ) , and includes accurate Lagrangian hydrodynamics, here in the meshless finite-mass (MFM) implementation. We use a modified version of gizmo ( B. Liu & V. Bromm 2020 ; J. Jeon et al. 2023 ; J. Jeon et al. 2025a ) with updated sub-grid models for star formation and feedback (Section II.2 ), primordial chemistry, cooling, and metal enrichment, which have been tested against high-resolution simulations of first galaxy formation and high-redshift observations ( J. Jaacks et al. 2018 ; J. Jaacks et al. 2019 ) . II.1 Initializing a Biased Region To be able to form a sufficient number of Pop III stars and their resulting PISNe as early as z\sim 30 , we need an extreme overdensity from primordial fluctuations. Zoom-in simulations of biased regions from larger parent simulations can be run in order to capture such an overdensity ( Y. M. Bahé et al. 2017 ; J. Jeon et al. 2025a , e.g.,) . However, to sample the rarest and most extreme regions, very large parent simulations are required, which can be computationally costly. An alternative method to simulate an overdense region is to artificially boost the amplitude of primordial density fluctuations ( T. H. Greif et al. 2011 , e.g.,) . Then, when sampling randomly from the boosted fluctuations, the probability of choosing a biased, overdense region is much higher. Following this approach, we use standard Planck cosmological parameters ( Planck Collaboration et al. 2016 ) : \Omega_{m}=0.315 , \Omega_{b}=0.048 , n_{s}=0.966 , and h=0.6774 , but we change the value of \sigma_{8} here. Instead of the \sigma_{8}=0.829 observed by Planck , we increase it to \sigma_{8}=1.5 This overdense region, while rare, is predicted to be observable in the survey volumes JWST has achieved so far (Section III.1 ). The initial conditions are generated with the MUSIC code ( O. Hahn & T. Abel 2011 ) at z=99 , including both gas (baryonic) and dark matter (DM) components. We choose a box with side length 6 comoving h^{-1} Mpc, which translates to \sim 2-3 arcmin at z\sim 30 , corresponding to the field of view of one JWST NIRCam pointing. The simulation parameters are summarized in Table 1 . To assess the probability of realizing such a biased region within realistic cosmological density fluctuations, described by the standard \sigma_{8}=0.829 parameter, we characterize the resulting bias by the peak height (\nu) of the most massive halo in the simulation volume, defined as \nu(M_{h},z)=\frac{\delta_{c}}{\sigma(M_{h},z)}\mbox{\ ,} (1) where \delta_{c}=1.686 is the critical overdensity for spherical collapse, and \sigma(M_{h},z) the variance of the power spectrum on the scale of the halo with mass M_{h} at redshift z . If \nu<1 at a given redshift, a halo of mass M_{h} on average would have already collapsed at that redshift and if \nu>1 , the halo on average will collapse in the future. We identify the dark matter halos in post processing using Rockstar ( P. S. Behroozi et al. 2013 ) , and the \nu required for a given M_{h} at z=30.4 is derived with the Colossus package ( B. Diemer 2018 ) assuming the virial halo mass and radius definitions ( G. L. Bryan & M. L. Norman 1998 ) . Fig. 1 shows the most massive halo with mass 1.2\times 10^{8}~\mathrm{M_{\odot}} and the corresponding peak height \nu\sim 5 at z=30.4 . We quantify the number density of halos of this mass using the halo mass function from the Gadget at Ultrahigh Redshift with Extra-Fine Timesteps ( GUREFT ) simulations, designed to capture the halo merger histories at extremely high- z ( L. Y. A. Yung et al. 2024 ) . The derived halo number density is 8\times 10^{-7} Mpc -3 dex -1 , or, employing the R. K. Sheth & G. Tormen 1999 halo mass function instead, 1.2\times 10^{-6} Mpc -3 dex -1 . Thus, the number density of the halo considered here, within the biased region, is approximately \sim 10^{-6} Mpc -3 dex -1 , which renders our simulated volume extremely rare, but still achievable in the early Universe ( S. Naoz et al. 2006 , see also) . We examine the likelihood of JWST observing such a region in Section III.1 . Figure 1: Biased, overdense simulation volume. Left: The most massive dark matter halo in our simulation with M_{h}\simeq 1.2\times 10^{8} M ⊙ at z=30.4 and the corresponding peak height \nu\simeq 5 . The corresponding number densities are also shown, giving 8\times 10^{-7} Mpc -3 dex -1 for the most massive halo, based on the GUREFT halo mass function ( L. Y. A. Yung et al. 2024 ) . A similar number density of \sim 10^{-6} Mpc -3 dex -1 is derived when using the R. K. Sheth & G. Tormen 1999 halo mass function. The overdense region is unlikely but not impossible to encounter in the early Universe. Right: The projected gas density distribution of our simulation volume at z\sim 30.4 . The length is in physical units. Stellar particles are marked as black dots and the most massive halo as an orange circle, with sizes not to scale, showing that SF and the subsequent transient events can occur in such highly biased regions within the early Universe. II.2 Star formation and feedback We adopt the stochastic star formation (SF) models developed and validated in previous studies ( J. Jaacks et al. 2018 ; J. Jaacks et al. 2019 ; B. Liu & V. Bromm 2020 , see) . Specifically, the metal-free Pop III models are calibrated to be consistent with extremely high-resolution zoom-in simulations, and the metal-enriched, Population II (Pop II) stellar feedback models to reproduce the observed star formation histories at z\sim 5-10 . In the stochastic model, a gas particle becomes a SF candidate, when its hydrogen number density is n_{\rm H}>100 cm -3 and its temperature T\leq 10^{3} K. For a SF candidate, the probability of spawning a stellar particle is p_{\rm SF}=\frac{m_{\rm SF}}{m_{*}}[1-\exp(-\eta_{*}\Delta t/t_{{\rm ff},i})]\mbox{\ ,} (2) where m_{\rm SF} is the mass of the candidate gas particle, m_{*} the mass of the stellar particle to be formed, \eta_{*} the SF efficiency, \Delta t the simulation timestep, and t_{{\rm ff},i}=\sqrt{3\pi/(32G\rho_{i})} the free-fall timescale of the gas particle with density \rho_{i} . For Pop III stars, we set \eta_{*}=0.05 and for Pop II \eta_{*}=0.1 ( J. Jaacks et al. 2019 ) , reflecting the lower star-formation efficiencies of Pop III minihalos. For both populations, we set m_{*}\simeq 600 M ⊙ based on high-resolution Pop III star formation simulations ( V. Bromm 2013 ; A. Stacy et al. 2016 ; R. S. Klessen & S. C. O. Glover 2023 ) . A random number p is generated for the uniform distribution [0,1] and a stellar particle is formed when p