⚛️ phenomenology

Late reheating in the era of precision cosmology

This paper demonstrates that a cosmic reheating temperature as low as 0.4 MeV, driven by light particle decays into electrons and neutrinos, is consistent with current primordial abundance and CMB data, offering a scenario that reduces active neutrino mass density and suppresses sterile neutrino production while leaving distinct signatures for future cosmological surveys.

Original authors: Kensuke Akita, Oleksii Ihnatenko, Maksym Ovchynnikov

Published 2026-10-28

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Kensuke Akita, Oleksii Ihnatenko, Maksym Ovchynnikov

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ 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

The universe began as an incredibly hot, dense point that expanded and cooled, a story we call the Big Bang. As it cooled, the energy that filled space transformed into the particles that make up everything we see today: protons, neutrons, electrons, and neutrinos. Neutrinos are ghostly particles that rarely interact with anything else, yet they were once so abundant and energetic that they helped shape the formation of the first atomic nuclei. This process, known as Big Bang nucleosynthesis, occurred when the universe was just a few minutes old and the temperature had dropped to a specific range. If the universe had cooled too quickly or too slowly, the balance of elements would have been different, and the stars and galaxies we see today might never have formed.

For decades, scientists have assumed that the universe cooled smoothly and continuously from its fiery birth. However, there is a possibility that the cooling was interrupted. Imagine a long-lived particle, heavier than the atoms we know, that dominated the universe's energy for a while before finally decaying. When it decayed, it would have dumped a fresh burst of energy into the cosmic soup, effectively reheating the universe. This event, called reheating, would have set the temperature at which the first elements were forged. The question is: how cold could the universe have been before this reheating event, and still produce the universe we observe?

A team of researchers has investigated this question with new precision, using the most recent measurements of the cosmic microwave background—the afterglow of the Big Bang—and the observed amounts of light elements like helium and deuterium. They found that the universe could have been much colder than previously thought before this reheating event occurred. Specifically, they discovered that if the decaying particle produced both electrons and neutrinos, the reheating temperature could have been as low as 0.4 million degrees Kelvin, or 0.4 MeV. This is significantly colder than the previous limit of about 5 to 6 MeV, which was based on scenarios where the particle decayed only into electromagnetic radiation.

The key to this discovery lies in how the decaying particle shared its energy. In older models, scientists assumed the particle decayed into electrons and photons, which would heat up the plasma of the early universe. However, this process does not efficiently heat the neutrinos, which are crucial for converting protons into neutrons. Without enough neutrinos, the balance of protons and neutrons shifts, leading to too much helium and too little deuterium, which contradicts what we observe in the cosmos. The researchers realized that if the particle decayed directly into neutrinos as well as electrons, it could replenish the missing neutrinos. This direct injection allowed the universe to remain in a state compatible with our observations even at much lower temperatures.

To test this idea, the team ran detailed computer simulations that tracked the behavior of these particles as the universe expanded and cooled. They followed the energy of the decaying particles, the temperature of the plasma, and the complex interactions between different types of neutrinos. They found that when the decaying particle was relatively light, with a mass of just a few to ten times the mass of a proton, the neutrinos it produced were "softer," meaning they carried less energy. These softer neutrinos were less likely to convert protons into neutrons, which prevented the overproduction of helium that had ruled out lower temperatures in the past. By carefully adjusting the ratio of electrons to different types of neutrinos in their models, they found a sweet spot where the universe could be reheated to 0.4 MeV without violating any cosmological constraints.

This finding has profound implications for our understanding of the universe's history. It suggests that the universe could have spent a longer time in a cold, dormant state before the final burst of heating that allowed life to eventually emerge. Furthermore, this scenario changes the properties of the neutrinos that fill the universe today. Because the reheating process produced fewer neutrinos but with higher average energy, the total mass of these relic neutrinos could be significantly lower than standard models predict. This reduction in mass density would subtly change how matter clumps together in the universe, potentially altering the distribution of galaxies in a way that future telescopes could detect.

The study also offers a new perspective on the search for sterile neutrinos, a hypothetical type of particle that has been a target for laboratory experiments. In standard scenarios, the high temperatures of the early universe would have produced too many of these particles, creating a density of dark matter that contradicts observations. However, the late reheating scenario described in this paper suppresses the production of sterile neutrinos. This means that sterile neutrinos with masses and interaction strengths that are currently being searched for in laboratories could exist without violating cosmological limits. The researchers showed that in this specific reheating history, the abundance of these particles could be reduced by a factor of forty compared to standard models, bringing them into alignment with what we see in the sky.

While these results are based on sophisticated simulations and theoretical models, they provide a concrete pathway for testing these ideas. The researchers noted that the specific conditions required for this scenario—such as the particle decaying into both electrons and neutrinos with specific ratios—are consistent with certain theoretical extensions of the Standard Model of particle physics. They also highlighted that future experiments, such as those measuring the distribution of galaxies or the polarization of the cosmic microwave background, could potentially detect the subtle signatures of this late reheating. The work does not claim to have solved the mystery of the early universe, but it has significantly narrowed the range of possibilities, showing that the universe could have been far colder and more dynamic in its infancy than we previously imagined.

Technical Summary: Late Reheating in the Era of Precision Cosmology

Problem Statement
The paper investigates the minimum allowable cosmic reheating temperature ( $T_{RH}$ ) permitted by current precision measurements of primordial nuclear abundances (Big Bang Nucleosynthesis, BBN) and the Cosmic Microwave Background (CMB). Standard cosmological models assume reheating occurs via the decay of a long-lived massive particle ( $X$ ) into relativistic species. Previous studies have established lower bounds on $T_{RH}$ (typically 5–6 MeV) for decays into electromagnetic (EM) or hadronic channels. These bounds arise because EM decays heat the photon-electron plasma but fail to fully populate neutrinos via weak interactions, leading to a deficit in the effective number of neutrinos ( $N_{eff}$ ) and altering the neutron-to-proton ratio, which in turn affects light element yields (specifically Helium-4 and Deuterium). While a previous study (Ref. [18]) suggested that "mixed decays" (into both EM particles and neutrinos) could lower this bound to $\sim 1$ MeV, that analysis relied on approximations (omitting neutrino oscillations, collisions, and Pauli blocking) and used outdated data. This work re-evaluates these constraints using modern, high-precision data and a rigorous treatment of non-equilibrium neutrino kinetics.

Methodology
The authors model a scalar "reheater" particle $X$ that dominates the energy density before decaying. They employ the Neutrino Direct Simulation Monte Carlo (νDsmc) method to track the coupled evolution of:

Neutrino momentum distributions (resolving flavor, particle/antiparticle, and momentum).
The cold and moving populations of the reheater $X$ .
The electromagnetic plasma temperature.
The cosmological scale factor.

The simulation includes scattering on electrons and neutrinos, pair annihilation, decays, inverse decays, quantum statistical factors, and flavor oscillations. The nuclear reaction network utilizes Primat rates with finite nucleon mass corrections.

To constrain the models, the authors combine:

CMB Data: A joint likelihood (CMB-SPA) from Planck, ACT, and SPT observations, constraining the baryon density ( $\omega_b$ ), $N_{eff}$ , and the helium mass fraction used in recombination ( $Y_{CMB}$ ).
BBN Data: Updated measurements of primordial Helium-4 ( $Y_P$ ) and Deuterium-to-Hydrogen ( $D/H$ ) ratios from recent literature.
Statistical Analysis: They profile over $\omega_b$ and branching fractions to minimize the joint $\chi^2$ statistic, defining the allowed regions at 95% confidence level (CL).

Key Contributions and Results

Lowered Reheating Bound: The primary result is that mixed decays of a light particle into electrons and neutrinos are compatible with joint CMB and BBN constraints for reheating temperatures as low as $T_{RH} \simeq 0.4$ MeV. This is an order of magnitude lower than the bound for purely electromagnetic decays.
- Mechanism: Direct injection of neutrinos compensates for the neutrino deficit caused by EM heating. The specific branching fractions into neutrino flavors ( $B_{\nu_e}, B_{\nu_\mu}, B_{\nu_\tau}$ ) are critical. For $T_{RH} \sim 0.4$ MeV, the compatible histories favor a large branching fraction into electron neutrinos ( $B_{\nu_e}$ ) and a total neutrino branching fraction $B_\nu \approx B_{ee} \approx 0.5$ .
- Flavor Dynamics: The study demonstrates that treating neutrino flavors as a single fluid is insufficient. Fast oscillations do not necessarily equalize populations, and the specific injection of $\nu_e$ vs. $\nu_\mu/\nu_\tau$ significantly impacts the neutron-to-proton conversion rates and the resulting $Y_P$ .
Mass Dependence and Pion Production:
- For heavier reheaters ( $m_X \gtrsim 40$ MeV), energetic neutrinos transfer energy efficiently to the EM plasma and convert protons to neutrons, strengthening the lower bound on $T_{RH}$ .
- Lighter reheaters ( $m_X \sim \text{few to } 10$ MeV) inject softer neutrinos, which suppress proton-to-neutron conversion more than the reverse, allowing for lower $T_{RH}$ without overproducing Helium.
- The analysis includes subdominant decays into charged pions (e.g., $X \to \gamma \pi^+ \pi^-$ ), which can convert protons to neutrons. This effect raises the bound for photon decays at $m_X \gtrsim 2m_{\pi^\pm}$ but is negligible for electron decays in the relevant mass range.
Refutation of Previous "Mixed" Scenarios: The authors show that the previously allowed region near $T_{RH} = 1$ MeV with high neutrino branching fractions ( $B_\nu \approx 0.9$ ) is excluded. Such scenarios overproduce radiation ( $N_{eff} \sim 9.6$ ) and Helium ( $Y_P \sim 0.33$ ) due to the energetic nature of the injected neutrinos, which is inconsistent with observations.
Implications for Neutrino Physics:
- Active Neutrino Mass Density: Late reheating can result in a relic neutrino population with nearly standard radiation density ( $N_{eff} \approx 2.9$ ) but fewer particles carrying higher mean momenta. Consequently, for a fixed physical neutrino mass sum, the present-day neutrino mass density can be reduced by up to a factor of five (for normal ordering). This alters the suppression of matter clustering, potentially raising the linear matter power spectrum by $\sim 1\%$ .
- Sterile Neutrinos: The scenario suppresses the production of sterile neutrinos ( $\nu_s$ ) that mix with active flavors. For mixings within the projected reach of laboratory experiments (e.g., KATRIN, nuSCOPE, SHiP), late reheating can reduce the sterile abundance by a factor of $\sim 40$ compared to instantaneous reheating, rendering them compatible with hot dark matter limits.

Significance and Claims
The paper claims to establish that the "standard" lower bound on reheating temperature is not absolute but depends critically on the decay channels of the reheating particle. By demonstrating that $T_{RH} \simeq 0.4$ MeV is viable for mixed decays, the work opens a new window for cosmological models involving light, long-lived particles.

The authors emphasize that this scenario has distinct observational signatures:

It modifies the relationship between cosmological observables (lensing, galaxy clustering) and physical neutrino masses, potentially alleviating tensions between oscillation data and CMB/DESI mass constraints.
It provides a mechanism to suppress sterile neutrino production to levels consistent with non-observation in current and future experiments, even for mixing angles that would otherwise be excluded.
The results rely on the specific particle model (scalar $X$ with specific couplings) and the rigorous treatment of non-equilibrium kinetics; other spins or couplings may shift the bounds.

The work concludes that while the minimum temperature is lowered, the constraints from light element abundances and radiation density remain robust, provided the branching fractions and reheater mass fall within the specific "compatible" regions identified. Future galaxy surveys and lensing measurements are expected to probe the percent-level changes in matter power predicted by this scenario.

Technical Summary: Late Reheating in the Era of Precision Cosmology

More like this