⚛️ phenomenology

Exothermic Dark Matter with a Dark Higgs: The LZ 248 keV Event and the Galactic Center Excess

This paper proposes an exothermic dark matter model within a dark $U(1)$ framework to simultaneously explain the LZ 248 keV nuclear-recoil event and the Galactic Center $\gamma$ -ray excess, while demonstrating that the lighter mass branch is excluded by cosmic microwave background constraints and the heavier branch remains viable under INTEGRAL/SPI limits.

Original authors: Jinheung Kim, Jongkuk Kim, Pyungwon Ko

Published 2026-10-20

📖 4 min read🧠 Deep dive

CC0 1.0

Original authors: Jinheung Kim, Jongkuk Kim, Pyungwon Ko

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

For decades, the hunt for dark matter has been a game of listening for a whisper in a noisy room. Scientists assume that the invisible substance making up most of the universe's mass occasionally bumps into ordinary atoms, creating a tiny, measurable kick. Standard theories predict these kicks should happen at very low energies, like a gentle tap. However, the LUX-ZEPLIN experiment, a massive tank of liquid xenon buried deep underground, recently reported something unusual: a single, much harder hit. This event registered an energy of 248 keV, far higher than the typical background noise and significantly stronger than the gentle taps most models expect. While this single event could simply be a random fluke or a hidden background signal, it sparked a new line of thinking. If it is real, it suggests that dark matter might not be a static, boring particle, but something more dynamic that can change its state when it collides with ordinary matter.

A team of researchers in Korea has proposed a specific scenario to explain this high-energy hit without breaking the rules of physics elsewhere in the universe. They suggest that dark matter consists of two versions of the same particle: a heavier, excited version and a lighter, ground-state version. In their model, when the heavier version smashes into an atomic nucleus, it drops down to the lighter version, releasing extra energy in the process. This "exothermic" transition acts like a spring-loaded mechanism, allowing even slow-moving dark matter particles to deliver the hard kick observed at 248 keV. To make this work, the researchers introduced a new force carrier, a dark photon, and a dark Higgs boson, which give the particles their masses and allow them to interact with the visible world in a very specific way.

The researchers did not just build a theory to fit one number; they ran a complex simulation to see if this idea could survive a gauntlet of other cosmic tests. They tracked how the population of these dark matter particles has evolved since the beginning of the universe. A critical part of their calculation was determining how many of the heavier particles are still around today to cause the collision. Their model showed that the interaction between the particles and the heat of the early universe naturally leaves just the right amount of the heavier version remaining to explain the single event seen by the detector. However, this same interaction creates a tightrope walk. If the force carrier is too light, the heavier particles would have decayed away too quickly, leaving nothing to cause the hit. If they are too heavy, the signal would be too weak.

The study found that the model only works in two specific ranges for the mass of the dark photon. One range involves a very light particle, and the other involves a much heavier one. The researchers then checked these two possibilities against other data from space. The light option was immediately ruled out. Observations of the cosmic microwave background, the afterglow of the Big Bang, show that if these light particles existed, they would have injected too much energy into the early universe, altering the pattern of radiation we see today. The heavy option, however, survived this test. Furthermore, this heavier version offers a bonus explanation for another long-standing mystery: an excess of gamma rays coming from the center of our galaxy. The model predicts that when dark matter particles collide and annihilate, they produce a spectrum of light that matches the glow seen by the Fermi-LAT telescope, specifically in the southern part of the sky where the view is clearest.

This solution is not a final proof, but it is a remarkably consistent picture. The researchers calculated that the rate of collisions in their model matches the single event seen by the underground detector, while simultaneously explaining the gamma-ray glow from the galactic center. They also showed that the model respects strict limits on how long the heavier dark matter particle can live before decaying into three photons, a process monitored by space-based observatories. The paper concludes that if the 248 keV event is indeed a signal from dark matter, it points toward a universe where dark matter has a hidden structure, a dark force, and a partner particle that helps it decay, all working together to produce a signal that is just strong enough to be seen, but not so strong that it contradicts the history of the cosmos. The next step for scientists is to see if other detectors can find more of these events or if the single hit was just a statistical fluke, but for now, this model provides a coherent, testable story for what might be happening in the dark.

Technical Summary: Exothermic Dark Matter with a Dark Higgs

Problem Statement
The LUX-ZEPLIN (LZ) collaboration recently reported a single nuclear-recoil candidate event at approximately 248 keV, with a global significance of 2.6 $\sigma$ . Standard direct-detection analyses typically assume spin-independent elastic scattering with a velocity-independent cross section, predicting a recoil spectrum that falls steeply and focuses on energies of order 10 keV. The LZ event lies well beyond this range. While background origins cannot be excluded, the event motivates mechanisms that can explain a high-energy recoil without overpredicting rates at lower or higher energies. Inelastic dark matter (DM) models, where DM scatters into a different mass eigenstate, offer a solution by restricting nuclear recoils to a finite energy window. Specifically, exothermic scattering (transitioning to a lighter state) allows even slow DM to produce high-energy recoils. However, such models require a sufficient abundance of the heavier excited state ( $\chi_2$ ) today, which is constrained by conversion processes and decays that deplete this state over cosmic time.

Methodology
The authors propose a dark $U(1)_X$ gauge model where a Dirac fermion DM candidate ( $\chi$ ) acquires a mass splitting ( $\delta$ ) between two Majorana states ( $\chi_1, \chi_2$ ) via a dark Higgs field ( $\Phi$ ). The dark Higgs also generates the mass of the dark photon ( $Z'$ ). The model includes kinetic mixing ( $\epsilon$ ) between the dark photon and the Standard Model hypercharge.

Key methodological steps include:

Relic Density and Excited Fraction ( $f_2$ ): Unlike previous works that often fix the excited-state fraction $f_2$ or treat it as a free parameter, the authors evolve $f_2$ from the early universe to the present day. They compute the non-perturbative conversion rate ( $\chi_2\chi_2 \leftrightarrow \chi_1\chi_1$ ), scattering on the thermal bath ( $\chi_2\ell \leftrightarrow \chi_1\ell$ ), and decays ( $\chi_2 \to \chi_1 + \text{SM}$ ). The gauge coupling $g_X$ is fixed at each parameter point to reproduce the observed relic density ( $\Omega h^2 = 0.120$ ).
Direct Detection Likelihood: The authors construct a likelihood function for the LZ event, incorporating the expected signal in the Region of Interest (ROI) and the High-Energy Sideband (SB). They assume the event is a signal and use the generalized Navarro–Frenk–White (NFW) halo profile.
Multi-Messenger Constraints: The model is tested against:
- INTEGRAL/SPI: Limits on the three-photon decay $\chi_2 \to \chi_1 + 3\gamma$ .
- Cosmic Microwave Background (CMB): Limits on energy injection from $\chi_2$ decays during the dark ages.
- Galactic Center Excess (GCE): The prompt $\gamma$ -ray spectrum from DM annihilation in the Galactic Center, specifically the southern-sky spectrum endorsed by Cholis et al. (2022).
Coannihilation Channel: The model includes the coannihilation channel $\chi_1\chi_2 \to Z'\phi$ (dark Higgsstrahlung), mediated by the dark Higgs boson ( $\phi$ ), which is crucial for reproducing the GCE spectrum.

Key Contributions and Results
The analysis reveals that the interplay between the direct-detection cross section and the present-day excited fraction $f_2$ creates two distinct solution branches for the dark photon mass ( $m_{Z'}$ ) at a fixed kinetic mixing $\epsilon = 10^{-4}$ :

The Low- $m_{Z'}$ Branch:
- Characterized by a large scattering cross section but a very small $f_2$ (due to efficient conversion and fast decays).
- Exclusion: This branch is excluded near its likelihood minimum by CMB limits. The energy injected by $\chi_2$ decays during the dark ages exceeds observational bounds by a factor of ~49 at the benchmark point. Additionally, INTEGRAL/SPI limits on the 3 $\gamma$ flux constrain the rate.
The High- $m_{Z'}$ Branch:
- Characterized by a smaller scattering cross section but a larger $f_2$ (approaching 0.5).
- LZ Fit: This branch provides a good fit to the LZ 248 keV event. The best-fit parameters are $m_\chi \approx 43.9$ GeV, $\delta \approx 1$ MeV, and $m_{Z'} \approx 25.1$ GeV.
- GCE Agreement: Crucially, this branch agrees with the southern-sky Galactic Center $\gamma$ -ray excess. The coannihilation channel $\chi_1\chi_2 \to Z'\phi$ , followed by the decay of $\phi$ (and $Z'$ ) into hadrons (dominated by $b\bar{b}$ ), reproduces the observed $\gamma$ -ray spectrum. The normalization of this flux is consistent with the relic cross section when a common halo model is used.
- Role of the Dark Higgs: The presence of the dark Higgs boson with mass $m_\phi \approx m_{Z'}$ is essential. Without the $\chi_1\chi_2 \to Z'\phi$ channel, the agreement with the GCE worsens significantly.

Significance and Claims
The paper claims that if the LZ 248 keV event is interpreted as a dark matter signal, it selects a specific region of parameter space within an exothermic dark matter model with a dark Higgs.

Model Viability: The model successfully explains the LZ event while satisfying INTEGRAL/SPI limits and CMB constraints, provided the solution lies on the high- $m_{Z'}$ branch.
Unified Explanation: The high- $m_{Z'}$ branch offers a unified explanation for both the LZ nuclear-recoil candidate and the Galactic Center $\gamma$ -ray excess, linking them through the same particle physics parameters ( $m_\chi, \delta, m_{Z'}, m_\phi, g_X, \epsilon$ ).
Predictive Power: The model predicts that the excited-state fraction $f_2$ is dynamically determined rather than fixed, leading to a specific correlation between the scattering rate and the decay width. This ties the LZ rate to the 3 $\gamma$ flux, making the scenario testable by future MeV $\gamma$ -ray observations.
Future Tests: The authors note that existing data from XENONnT and PandaX-4T in extended analysis windows can already test this scenario. Furthermore, argon-based detectors could provide a clean test at higher recoil energies where the signal lies below the first zero of the form factor.

The paper concludes modestly that while the background origin of the LZ event cannot be ruled out, the proposed exothermic dark matter scenario with a dark Higgs provides a consistent and testable framework that accommodates the event and the Galactic Center excess simultaneously.

Technical Summary: Exothermic Dark Matter with a Dark Higgs

More like this