🔭 astrophysics

Signatures of Very Massive Stars in the Epoch of Reionization

JWST observations of two high-redshift galaxies reveal spectral signatures of very massive stars, providing evidence that an extended initial mass function beyond 100 solar masses significantly enhanced ionizing output and shaped the early Universe's chemical and spectral evolution.

Original authors: Rui Marques-Chaves, Fabrice Martins, Daniel Schaerer, Miroslava Dessauges-Zavadsky, Ana Palacios

Published 2026-04-24

📖 4 min read☕ Coffee break read

CC0 1.0

Original authors: Rui Marques-Chaves, Fabrice Martins, Daniel Schaerer, Miroslava Dessauges-Zavadsky, Ana Palacios

Original paper dedicated to the public domain under CC0 1.0 (http://creativecommons.org/publicdomain/zero/1.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the early Universe as a vast, dark construction site just after the Big Bang. For a long time, it was shrouded in a thick fog of neutral gas. Then, the first lights turned on, burning away the fog in a process called the "Epoch of Reionization."

For a long time, astronomers thought the "bulldozers" clearing this fog were standard massive stars—big, bright, and hot, but within a known size limit. However, a new study using the James Webb Space Telescope (JWST) suggests that the construction crew included something much more extreme: Very Massive Stars (VMS). These aren't just big; they are cosmic giants, weighing in at over 100 times the mass of our Sun, some potentially reaching 300 times.

Here is the story of how we found them, explained simply.

The Detective Work: Looking for "Fingerprints"

The astronomers pointed the JWST at two incredibly bright, young galaxies from the early Universe (about 13 billion years ago), named CEERS-1019 and CEERS-1025. They didn't just take a picture; they took a "spectral fingerprint" of the light coming from these galaxies.

Think of a star's light like a song. Most stars sing a clear, steady tune. But these super-massive stars are like heavy metal bands playing at maximum volume with a massive wind machine blowing through the microphone.

The Wind: Because these stars are so heavy, they are unstable. They blow off their outer layers in powerful, screaming winds.
The Signature: These winds create a specific pattern in the light called a P-Cygni profile. Imagine a singer shouting a note while a giant fan blows the sound backward; you hear the note clearly, but with a weird "whoosh" trailing behind it.
The Smoking Gun: The team found two specific "notes" in the galaxies' light: a strong, broad emission of Helium and a very distinct "whoosh" pattern in Nitrogen and Carbon.

The Problem: The "Standard" Models Failed

The astronomers tried to match these fingerprints against a library of computer models.

The "Standard" Models: These models assumed the galaxies were made of normal massive stars (up to about 100 solar masses). It was like trying to explain a hurricane by saying it was just a strong breeze. The models failed miserably; they couldn't reproduce the strength of the "wind" or the weird Helium glow.
The "Super-Star" Models: Then, they tried a new model that included Very Massive Stars (up to 300 solar masses). Suddenly, the puzzle pieces clicked. The model predicted exactly the kind of screaming winds and bright Helium glow that the telescope saw.

The statistical evidence was overwhelming. It was like finding a fingerprint at a crime scene that matched a suspect with 99.9% certainty, while the "innocent" suspects had no match at all.

The Implications: Tiny, Bright, and Powerful

The study revealed three mind-blowing facts about these galaxies:

They are incredibly young: The stars are only about 1.5 to 2 million years old. In cosmic terms, this is a newborn. It's like catching a baby star in its first cry. Because these super-stars burn their fuel so fast, they die young. Finding them means we are catching these galaxies at the exact moment they are forming their most extreme stars.
They are ionizing machines: These stars are pumping out ionizing radiation (the energy needed to strip electrons from atoms) at a rate much higher than we thought possible. They are the heavy-duty power drills clearing the cosmic fog much faster than expected.
The "Top Heavy" Family Tree: This suggests that in the early Universe, the "Initial Mass Function" (the rulebook for how stars are born) was different. Instead of having mostly small stars and a few big ones, the early Universe seemed to be "top-heavy," creating a surplus of these 100-to-300-solar-mass monsters.

The Big Picture

Why does this matter?

Clearing the Fog: These super-stars were likely the main reason the early Universe became transparent. They were the "sunlight" that burned away the cosmic fog.
Chemical Alchemy: As these stars blow their winds, they scatter heavy elements (like nitrogen and carbon) into space. This is the "soil" that future generations of stars and planets (like our own) are made of.
A New Era: This study proves that the early Universe was a wilder, more extreme place than we imagined. It wasn't just a gentle nursery of stars; it was a factory churning out cosmic giants that shaped the destiny of the entire Universe.

In short: By listening to the "windy songs" of two ancient galaxies, astronomers have confirmed that the early Universe was home to a population of stellar giants that were far more massive and powerful than anything we see today, acting as the primary engines that lit up the cosmos.

1. Problem Statement

The existence and prevalence of Very Massive Stars (VMS; $M_\star \gtrsim 100\,M_\odot$ ) in the early Universe (Epoch of Reionization, $z \sim 8-10$ ) remain a critical open question in astrophysics.

Theoretical Context: VMS are predicted to form in dense, low-metallicity environments typical of high-redshift galaxies. They possess extreme luminosities, powerful stellar winds, and short lifetimes ( $\sim 2$ Myr).
Observational Challenge: Identifying VMS in distant galaxies is difficult because:
1. Their lifetimes are extremely short, requiring observation of very young stellar populations.
2. Their spectroscopic signatures (broad emission lines, P-Cygni profiles) are often subtle in integrated galaxy spectra and can be confused with nebular emission or Active Galactic Nuclei (AGN) activity.
3. High signal-to-noise (SNR) and moderate spectral resolution in the rest-frame UV are required to disentangle these features, which has historically been observationally prohibitive at $z > 6$ .
Specific Gap: Previous studies at intermediate redshifts ( $z \sim 2-4$ ) suggested VMS presence, but direct evidence at $z \sim 8.7$ (the Epoch of Reionization) was lacking.

2. Methodology

The authors utilized ultra-deep spectroscopy from the James Webb Space Telescope (JWST) to analyze two UV-bright galaxies: CEERS-1019 and CEERS-1025 at $z \approx 8.7$ .

Observations:
- Instrument: NIRSpec (G140M/F100LP configuration).
- Exposure Time: Ultra-deep integrations of $\sim 29.5$ hours for CEERS-1019 and $\sim 19.6$ hours for CEERS-1025.
- Resolution: $R \simeq 1000$ covering $\lambda \approx 0.9 - 1.84\,\mu\text{m}$ (rest-frame UV).
- Targets: Both galaxies reside in an overdense region with low metallicity ( $Z_{\text{neb}} \simeq 0.1\,Z_\odot$ ).
Spectral Analysis & Modeling:
- Code: Used FiCUS to model observed spectra as linear combinations of Simple Stellar Population (SSP) templates.
- Model Comparison: Compared standard stellar population models against models including VMS:
  - Standard Models: Starburst99 (SB99; IMF cutoff $120\,M_\odot$ ) and BPASS (IMF cutoff $100\,M_\odot$ ).
  - VMS Models: Models by Martins et al. (2025) with an extended IMF up to $300\,M_\odot$ .
- Mass-Loss Prescriptions: Tested two VMS mass-loss ( $\dot{M}$ $\dot{M}$ ) scenarios:
  1. VMS_GR: Mass loss independent of metallicity (LMC-like).
  2. VMS_GRZ: Mass loss scaling linearly with metallicity relative to LMC values.
- Statistical Validation: Used $\chi^2$ minimization, Akaike Information Criterion (AIC), and Bayesian Information Criterion (BIC) to evaluate model fits.

3. Key Results

A. Spectral Features

The high-SNR spectra revealed distinct signatures of massive stellar winds that standard models could not reproduce:

P-Cygni Profiles: Exceptionally strong P-Cygni profiles were detected in wind lines N v $\lambda$ 1240 and C iv $\lambda$ 1550.
He ii $\lambda$ 1640 Emission: Significant broad emission was detected in both galaxies:
- CEERS-1019: $EW \approx 2.6 \pm 0.6$ Å, FWHM $\approx 1380$ km s $^{-1}$ .
- CEERS-1025: $EW \approx 4.0 \pm 0.8$ Å, FWHM $\approx 1150$ km s $^{-1}$ .
- Note: AGN activity was ruled out as the primary source due to the lack of dominant AGN signatures and the presence of similar broad components in [O III] lines (indicating outflows).

B. Model Fitting & Statistical Evidence

Superiority of VMS Models: Models including VMS provided a markedly improved fit to the data compared to non-VMS models.
- Statistical Significance: $\Delta$ AIC and $\Delta$ BIC values were $> 70$ (ranging up to $\sim 280$ ) in favor of VMS models. This indicates overwhelming statistical support.
- Failure of Standard Models: Non-VMS models (SB99, BPASS) significantly underestimated the strength of the N v P-Cygni emission and the broad He ii component. At $Z \approx 0.1\,Z_\odot$ , classical Wolf-Rayet (WR) stars are predicted to have weak winds, insufficient to explain the observed line strengths.

C. Derived Physical Parameters

Based on the best-fit VMS models (specifically VMS_GRZ):

Stellar Ages: Extremely young populations, $1.5 - 2.0$ Myr, consistent with the short lifetimes of VMS.
Ionizing Photon Efficiency: High production efficiency of ionizing photons, $\log(\xi_{\text{ion}}) \gtrsim 25.8$ Hz erg $^{-1}$ . This exceeds values from non-VMS models by $\sim 0.1 - 0.2$ dex.
Metallicity: Consistent with gas-phase measurements ( $Z \approx 0.1\,Z_\odot$ ).

D. Comparative Analysis

The spectra of CEERS-1019 and CEERS-1025 were compared to local VMS-dominated systems (e.g., R136 in the LMC, MrK71-A, Sunburst cluster at $z=2.37$ ).

The spectral shapes (C iv P-Cygni and broad He ii wings) closely resemble those of local VMS-dominated clusters, despite the large difference in redshift and metallicity.

4. Significance and Implications

Direct Evidence at High-z: This study provides the first direct spectroscopic evidence for an overabundance of VMS in galaxies during the Epoch of Reionization ( $z \sim 8.7$ ).
IMF Constraints: The results suggest that the Initial Mass Function (IMF) in early galaxies extends well beyond $100\,M_\odot$ , potentially up to $300\,M_\odot$ or higher.
Cosmic Reionization: VMS significantly enhance the ionizing photon output ( $\xi_{\text{ion}}$ ). If ubiquitous, VMS populations could play a dominant role in driving cosmic reionization, a process previously attributed mainly to standard massive stars or AGN.
Chemical Enrichment: The presence of VMS helps explain the nitrogen-enhanced nature of CEERS-1019 (high N/O ratios), as VMS winds are efficient at ejecting processed material early in a galaxy's life.
Future Outlook: The study underscores the necessity of deep rest-frame UV spectroscopy (JWST and future ELT) to statistically characterize VMS populations and refine models of early galaxy evolution.

Conclusion

Marques-Chaves et al. demonstrate that standard stellar population models fail to reproduce the extreme wind features in $z \sim 8.7$ galaxies. The inclusion of Very Massive Stars ( $M > 100\,M_\odot$ ) is statistically required to explain the observed P-Cygni profiles and broad He ii emission. This confirms that VMS are a significant component of the first generations of stars, profoundly influencing the ionizing budget and chemical evolution of the early Universe.