🔬 physics

At the EDGES of MADness: phenomenology of relic millicharged atomic dark matter

This paper proposes that atomic dark matter naturally produces a relic millicharged component capable of explaining the EDGES 21-cm absorption signal while remaining consistent with cosmological and laboratory constraints, and identifies distinctive signatures for future 21-cm observations.

Original authors: Eleonora Vanzan, Kimberly K. Boddy, Ely D. Kovetz

Published 2026-10-30

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Eleonora Vanzan, Kimberly K. Boddy, Ely D. Kovetz

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 is filled with invisible matter that holds galaxies together, yet we have never seen a single particle of it. This "dark matter" makes up most of the mass in the cosmos, but it refuses to interact with light, making it impossible to photograph directly. For decades, scientists have tried to catch a glimpse of its nature by watching how it moves and how it influences the ordinary matter we can see. One of the most promising ways to do this is by listening to the universe's oldest radio signal: a faint hum emitted by hydrogen gas when the first stars began to ignite, roughly 180 million years after the Big Bang. This signal, known as the 21-centimeter line, acts as a thermometer for the early universe. If dark matter interacts with ordinary gas, even slightly, it would change the temperature of that gas, leaving a distinct fingerprint on the radio signal. Recently, a controversial observation suggested this signal was much deeper than expected, hinting that dark matter might be interacting with normal matter in a way that cools the gas. However, for this to work without contradicting other observations of the early universe, the interacting dark matter would have to be a tiny, almost invisible fraction of the total dark matter supply.

The challenge for physicists has been explaining why this tiny fraction exists. In simple models, it is difficult to generate a scenario where only a small sliver of dark matter carries an electric charge while the rest remains neutral; it usually requires an arbitrary, unexplained adjustment to the numbers. A new study by Eleonora Vanzan, Kimberly K. Boddy, and Ely D. Kovetz proposes a natural solution to this puzzle. They suggest that dark matter is not just a single type of particle, but a complex "dark sector" that mirrors our own visible world. In their model, dark matter consists of heavy dark protons and light dark electrons that bind together to form dark atoms, much like the hydrogen atoms that make up the gas in our own universe. Just as ordinary hydrogen can be ionized into free protons and electrons, this dark hydrogen undergoes a similar process called "dark recombination." As the dark universe cools, most of the dark particles bind into neutral atoms, but a small, leftover fraction remains as free, charged ions. Because of a subtle connection between the dark world and our own, these remaining ions acquire a tiny electric charge, making them "millicharged." This mechanism naturally produces the exact small fraction of interacting dark matter needed to explain the radio signal, without requiring any artificial tuning.

To test this idea, the researchers built a sophisticated computer simulation that tracks the history of the universe from its earliest moments to the present day. They modified a standard cosmological software tool to include this specific type of dark matter, allowing them to calculate how the dark atoms form, how the leftover ions interact with ordinary gas, and how this affects the cosmic microwave background—the afterglow of the Big Bang. They then compared their simulation results against a vast collection of real-world data, including measurements of the cosmic microwave background from the Planck satellite, the distribution of galaxies across the sky, and the light from distant supernovae. The goal was to see if their model could fit all these observations simultaneously while still allowing for the specific interactions needed to cool the gas during the Cosmic Dawn.

The results show that this "millicharged atomic dark matter" model is a viable candidate for explaining the universe's structure. The simulations revealed that the model naturally produces a small fraction of charged dark particles, typically less than one percent of the total dark matter, which is exactly the range required to match the controversial radio signal observations. This fraction is not an input the scientists forced into the model; it is the automatic output of the dark atoms forming and leaving behind a residue of charged particles. The study found that the model works best when the dark protons are relatively heavy and the dark electrons are light, a combination that leaves just the right amount of free ions to interact with ordinary matter. Crucially, the model passes all the strict tests set by the cosmic microwave background and the large-scale structure of the universe, meaning it does not break our current understanding of how the cosmos evolved.

While the model fits the data well, the researchers are careful to note that the original radio signal observation remains controversial and has not been confirmed by other experiments. Therefore, this work does not prove that the signal was caused by dark matter. Instead, it demonstrates that if such a signal is real, there is a physically sound and natural explanation for it that does not rely on arbitrary assumptions. The study identifies specific regions of the model's parameters where the theory remains consistent with all known constraints, including laboratory limits on charged particles and the cooling of ancient supernovae. These viable regions overlap with the specific range of particle charges and masses that would explain the cooling of gas in the early universe.

The most exciting implication of this work is that it points to a new way to test these ideas. The model predicts that the interaction between the dark ions and ordinary gas leaves a unique signature on the temperature of the gas in the early universe, distinct from other theories. This signature is not visible in the cosmic microwave background, which is why the model has survived previous tests, but it should be detectable by the next generation of radio telescopes designed to map the 21-centimeter signal. If future observations confirm the specific temperature patterns predicted by this model, it would provide strong evidence that dark matter is not a single, simple particle, but a rich, complex sector with its own chemistry and history. Until then, the model stands as a compelling, self-consistent possibility that bridges the gap between the mysterious dark sector and the observable universe, offering a clear path forward for the next decade of cosmological discovery.

Technical Summary: At the EDGES of MADness: phenomenology of relic millicharged atomic dark matter

Problem Statement
Millicharged dark matter (DM), consisting of particles with a fractional electric charge $\epsilon e$ , offers a compelling explanation for the anomalously deep 21-cm absorption signal reported by the EDGES experiment at Cosmic Dawn ( $z \sim 17$ ). This signal suggests efficient cooling of baryons, potentially mediated by DM-baryon scattering. However, cosmological constraints, particularly from the Cosmic Microwave Background (CMB), require that any millicharged component constitute only a subpercent fraction ( $\lesssim 0.4\%$ ) of the total dark matter abundance to avoid excessive damping of CMB anisotropies and suppression of small-scale structure. In minimal single-particle millicharged DM models, invoking such a low interacting fraction is theoretically ad hoc. The central problem addressed by this work is whether a physically motivated mechanism exists within a dark sector model that naturally produces this small, relic millicharged fraction without fine-tuning.

Methodology
The authors propose and analyze a model of Millicharged Atomic Dark Matter (MAD). This scenario extends the Atomic Dark Matter (ADM) framework, where a secluded dark sector contains two fermions (a dark proton $p_D$ and a dark electron $e_D$ ) interacting via a massless dark photon ( $\gamma_D$ ). The model introduces kinetic mixing between the dark photon and the Standard Model (SM) photon, parameterized by $\epsilon$ .

Key methodological steps include:

Model Construction: The dark sector undergoes a thermal history mirroring the visible sector. As the universe cools, dark recombination occurs, forming neutral dark hydrogen and leaving a relic ionized fraction ( $x_{D}$ ). Due to kinetic mixing, this residual ionized component acquires a millicharge $\epsilon e$ , while the neutral component remains effectively dark. The fraction of millicharged DM is thus dynamically determined by the dark recombination history rather than being an arbitrary input.
Numerical Implementation: The authors modified the cosmological Boltzmann solver CLASS (v3.1.1) to incorporate the MAD scenario. This involved:
- Implementing the thermal evolution equations for both visible and dark baryon temperatures, accounting for intra-sector processes (dark Compton scattering, bremsstrahlung, recombination) and inter-sector heat/momentum transfer rates driven by the millicharge.
- Calculating momentum-transfer cross-sections ( $\sigma \propto v^{-4}$ ) regulated by Debye screening in the plasma.
- Defining an effective interaction parameter $R$ , the ratio of DM-SM heat-exchange rates to dark Compton rates at dark decoupling, to ensure the DM-dark radiation (DR) tight-coupling approximation remains valid.
Statistical Analysis: A Markov Chain Monte Carlo (MCMC) analysis was performed using MontePython against a comprehensive dataset: Planck 2018 CMB data (temperature, polarization, lensing), ACT DR6 lensing, BAO data (BOSS, SDSS, 6dFGS, DESI), and Pantheon supernovae. The analysis varied five MAD parameters: $\Delta N_{\text{eff}}$ (dark radiation contribution), $\alpha_D$ (dark fine-structure constant), $m_{p_D}$ , $m_{e_D}$ , and $\log_{10} R$ .

Key Contributions

Natural Origin of Low Fractions: The paper demonstrates that the required subpercent fraction of millicharged DM arises naturally as the relic ionization left over after dark recombination in an ADM scenario. This removes the need for ad hoc tuning of the interacting fraction.
Code Release: The authors publicly released MAD CLASS, a modified version of the standard Boltzmann solver, enabling systematic cosmological constraints on atomic dark matter with kinetic mixing.
Parameter Space Mapping: The study provides the first comprehensive cosmological constraints on the MAD parameter space, specifically mapping the viable regions in the mass-millicharge plane ( $m_{p_D}, m_{e_D}, \epsilon$ ) against CMB, Large Scale Structure (LSS), and supernovae data.

Results

Viable Parameter Windows: The analysis identifies cosmologically viable regions where the millicharge $\epsilon$ falls within the windows $10^{-6} \lesssim \epsilon \lesssim 10^{-4}$ and $\epsilon \lesssim 10^{-8}$ . These regions are consistent with laboratory bounds (SLAC) and astrophysical bounds (SN 1987A cooling).
EDGES Compatibility: The viable regions overlap significantly with the parameter space invoked in EDGES-style explanations for baryon cooling at Cosmic Dawn. Crucially, the relic millicharged fraction in these viable regions is naturally $\lesssim 0.4\%$ , satisfying CMB constraints without manual intervention.
Thermal Histories: The study identifies three benchmark regimes based on the timing of dark recombination relative to SM recombination:
- SM-like: Ruled out by CMB peak shifts.
- Early Dark Recombination: Behaves like interacting DM, showing characteristic CMB peak enhancements/suppressions.
- Late Dark Recombination: The most promising regime for 21-cm signals. Here, dark recombination occurs late enough that the residual ionized fraction significantly alters the late-time thermal history of baryons and dark baryons (cooling/heating effects) while leaving CMB anisotropies and the matter power spectrum nearly indistinguishable from $\Lambda$ CDM or standard ADM.
Hubble and $S_8$ Tensions: The MAD model inherits the ability of ADM to alleviate the $H_0$ and $S_8$ tensions via its dark radiation component ( $\Delta N_{\text{eff}}$ ) and dark acoustic oscillations. The millicharge interaction itself contributes only marginally to resolving these tensions; the primary driver remains the dark radiation density. The analysis recovers the strong bound $\Delta N_{\text{eff}} \lesssim 0.3$ .
Mass Dependence: The relic ionized fraction is primarily controlled by the dark proton mass ( $m_{p_D}$ ) and the dark electron mass ( $m_{e_D}$ ). Heavier dark protons lead to lower number densities and larger residual ionization fractions. A broad compatibility band exists around $m_{e_D} \sim 10$ MeV.

Significance
The paper argues that MAD represents a well-motivated intersection of high-energy physics (solutions to the Hierarchy Problem like Twin Higgs models) and cosmology. It provides a physically robust mechanism to generate the specific low-fraction millicharged DM required to explain potential Cosmic Dawn anomalies without violating early-universe constraints. While the EDGES result remains controversial and likely not a definitive Cosmic Dawn signal, the MAD framework predicts distinctive cosmological signatures in the late-time thermal history of the intergalactic medium. These signatures are accessible to upcoming 21-cm line-intensity mapping surveys, making MAD a prime target for the next generation of cosmological probes. The work emphasizes that the "least constrained" regions of the MAD parameter space are precisely those that leave the most significant imprints on 21-cm observables.

More like this