⚛️ phenomenology

Testing Dark Matter Coscattering with the High-Energy LZ Recoil

This paper demonstrates that the high-energy nuclear recoil observed by the LZ experiment can directly probe dark matter scenarios where the relic abundance is determined by coscattering or conversion-driven freeze-out, specifically within a singlet-doublet fermion model coupled to a light dark photon, thereby testing the same interaction mechanisms that governed dark matter production in the early universe.

Original authors: Sreemanti Chakraborti, Soumyajit Datta

Published 2026-10-07

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Sreemanti Chakraborti, Soumyajit Datta

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

For decades, the most compelling evidence for dark matter has come from gravity alone. We see stars orbiting the centers of galaxies too fast to be held by visible matter, and we watch light bend around massive clusters in ways that suggest invisible mass is doing the work. Yet, despite this gravitational certainty, we have never directly caught a particle of dark matter. The leading theory suggests these particles are "weakly interacting massive particles," or WIMPs, which should occasionally bump into ordinary atoms in detectors buried deep underground. For years, these detectors have listened for the faint tap of such a collision, but the silence has been deafening. Recently, however, the LUX-ZEPLIN experiment, a massive tank of liquid xenon deep in a South Dakota mine, reported a single, unusual event. It saw a nucleus recoil with an energy far higher than typical background noise, a signal that, while not yet a discovery, has sparked intense interest in whether it could be the first glimpse of a dark matter particle that behaves differently than expected.

This new study explores a specific possibility for that signal: a dark matter particle that does not bounce off an atom like a billiard ball, but instead uses the energy of the collision to transform into a slightly heavier version of itself. The researchers, working within a framework where dark matter exists in two nearly identical states—a lighter, mostly hidden one and a heavier, more interactive one—asked if this transformation could explain the high-energy event seen by the LUX-ZEPLIN detector. They focused on a scenario where the universe's current amount of dark matter was not set by the usual process of particles annihilating each other, but by a conversion process where the lighter particles constantly turned into the heavier ones, which then disappeared. This mechanism, known as "coscattering," requires a very specific set of conditions that usually make dark matter invisible to detectors. However, the team proposed that a light, invisible force carrier, acting like a messenger between the dark and visible worlds, could boost the chances of this transformation happening today, making the high-energy signal visible.

The researchers built a detailed model to test this idea, involving a dark matter particle that is a mix of two types: one that barely interacts with anything and another that interacts more strongly. In their scenario, the lighter version is the dark matter we see today, while the heavier version is just slightly more massive, separated by a tiny gap of energy. In the early universe, these particles were hot and moving fast. As the universe cooled, the lighter particles struggled to disappear on their own because they interact so weakly. Instead, they relied on converting into the heavier particles, which could then easily annihilate away. This conversion became the bottleneck that determined how much dark matter survived to the present day. The challenge for this theory is that the same small mixing between the two states that makes this conversion possible in the early universe should also make it nearly impossible for the light particles to interact with detectors on Earth today.

To bridge this gap, the team introduced a light dark photon, a hypothetical particle that acts as a bridge between the dark sector and the visible world. This particle allows the light dark matter to absorb energy from a collision with a xenon nucleus and jump up to the heavier state. Because this process requires extra energy, it only happens if the dark matter particle is moving very fast, which naturally pushes the resulting signal to higher energies. This matches the specific high-energy recoil event reported by the LUX-ZEPLIN experiment, which occurred at an energy of 248 keV. The researchers ran extensive simulations to see if the same parameters that create the correct amount of dark matter in the universe could also produce exactly one such event in the detector, as observed.

Their results show that this scenario is indeed possible, but only within a narrow window of conditions. For the lightest dark photon mass they considered, around 250 MeV, the model works remarkably well. In this case, the region of parameters that produces the correct amount of dark matter through the conversion process overlaps almost perfectly with the region that would generate the high-energy event seen by LUX-ZEPLIN. This means the experiment could be directly testing the very same interaction that governed the creation of dark matter billions of years ago. The model predicts that the dark matter particle would have a mass between 750 GeV and 1.6 TeV, a range that is heavy but accessible to current physics. The mass splitting between the two dark states must be between 150 and 300 keV to allow the high-energy recoil to occur without exceeding the speed limits of the particles in our galaxy.

However, the story becomes more complicated as the mass of the invisible messenger particle increases. When the researchers increased the mass of this dark photon to 500 MeV or 1 GeV, the connection between the early universe and the present-day detector began to fray. A heavier messenger makes it harder for the dark matter to interact with the detector, requiring a stronger mixing between the two dark states to produce a signal. But if the mixing is too strong, the conversion process in the early universe becomes too efficient, and the resulting amount of dark matter would be far too low to match what we observe. Consequently, for the heaviest messenger mass, the region where the detector would see a signal no longer overlaps with the region where the conversion-driven history of the universe works. The high-energy event would still be possible, but it would no longer be linked to the mechanism that created the dark matter we see today.

The study concludes that the high-energy recoil event reported by LUX-ZEPLIN offers a unique opportunity to probe a specific type of dark matter history. If the signal is indeed from dark matter, it points to a scenario where the abundance of these particles was set by a conversion process rather than simple annihilation. This would mean that the same force that allowed dark matter to transform in the hot, dense early universe is the same force allowing it to transform in our detectors today. While the model requires a light dark photon to make this interaction visible, it remains consistent with all other known constraints from particle accelerators. The researchers emphasize that this is a simulation-based finding, suggesting a viable path forward rather than a confirmed discovery. If future data from LUX-ZEPLIN or other experiments confirms this high-energy signal, it could provide the first direct evidence that dark matter's history is written in the language of conversion, linking the cosmos of the Big Bang to the quiet depths of a mine in South Dakota.

Technical Summary: Testing Dark Matter Coscattering with the High-Energy LZ Recoil

Problem Statement
The LUX-ZEPLIN (LZ) experiment recently reported a single high-energy nuclear recoil event with an energy of $248 \pm 23 \text{ (stat)} \pm 23 \text{ (sys)}$ keV. While statistically below the discovery threshold, this event has motivated various dark matter (DM) interpretations, particularly those involving inelastic scattering where a light DM particle upscatters to a slightly heavier state ( $\chi_1 N \to \chi_2 N$ ). A specific theoretical challenge arises in models where the DM relic abundance is determined by "coscattering" (or conversion-driven freeze-out). In such scenarios, the relic density is set by the efficiency of converting a weakly interacting singlet-like DM state into a more strongly interacting doublet-like state. However, the small mixing angle required to suppress direct annihilation and realize this freeze-out mechanism typically suppresses the direct detection signal to unobservable levels. The central question addressed by this work is whether the high-energy LZ recoil can probe DM models where the relic abundance is set by this conversion-driven mechanism, specifically within a singlet-doublet fermion framework coupled to a light dark photon.

Methodology
The authors investigate a minimal Singlet-Doublet Dark Matter (SDDM) model containing a vector-like Dirac fermion singlet ( $\xi$ ) and a vector-like Dirac fermion doublet ( $\Psi$ ), coupled via a complex scalar singlet ( $\Phi$ ) and a light dark photon ( $A'$ ).

Model Structure: The model features a small mass splitting ( $\delta \sim \mathcal{O}(100)$ keV) between the lightest neutral state ( $\chi_1$ , predominantly singlet) and the next-to-lightest state ( $\chi_2$ , predominantly doublet). The interaction is mediated by a dark photon with kinetic mixing $\epsilon$ and a dark gauge coupling $g_D$ .
Cosmological Evolution: The authors utilize the sector decomposition method (following Ref. [47]) to solve coupled Boltzmann equations. They track the singlet sector ( $\chi_1$ ) and the doublet sector ( $\chi_2, \psi^\pm$ ) independently. The relic abundance is calculated using micrOMEGAs, accounting for annihilation, co-annihilation, and crucially, bath-assisted conversion processes ( $\chi_1 + \text{bath} \leftrightarrow \chi_2 + \text{bath}$ ). The regime of interest is "coscattering," where the conversion rate becomes the bottleneck for freeze-out rather than annihilation.
Direct Detection: The authors calculate the differential recoil rate for Xenon isotopes in the LZ detector. They focus on the endothermic scattering process $\chi_1 N \to \chi_2 N$ , mediated primarily by the dark photon. The kinematics of this process are sensitive to the mass splitting $\delta$ , which shifts the recoil spectrum to higher energies, potentially matching the observed 248 keV event.
Constraints: The analysis incorporates constraints from accelerator searches for visible dark photon decays and dark Higgstrahlung, fixing the kinetic mixing at $\epsilon = 10^{-4}$ .

Key Contributions

Unified Probe of Cosmology and Detection: The paper demonstrates that in the coscattering regime, the relic abundance is intrinsically sensitive to the same off-diagonal interaction structure ( $\chi_1 \leftrightarrow \chi_2$ ) that governs the inelastic direct detection signal. This contrasts with models where relic density is set by secluded annihilation, decoupling the two observables.
Role of the Light Dark Photon: The authors show that a light dark photon ( $m_{A'} \sim \mathcal{O}(100)$ MeV) is essential. It enhances the endothermic transition rate $\chi_1 N \to \chi_2 N$ sufficiently to generate an observable signal at LZ, despite the small singlet-doublet mixing ( $\sin \theta$ ) required to maintain the conversion-driven freeze-out history.
Parameter Space Mapping: The study maps the viable parameter space in the $(m_{\chi_1}, \sin \theta)$ plane for different mediator masses ( $m_{A'} = 250, 500, 1000$ MeV) and mass splittings ( $\delta = 150, 200, 300$ keV).

Results

Mass Range: For the mediator masses considered, the combined requirements of the correct relic density and the LZ event rate select DM masses in the range of approximately 750 GeV to 1.6 TeV.
Mediator Mass Dependence:
- For $m_{A'} = 250$ MeV, the LZ-sensitive region (defined by an expected event rate $N_{LZ} \approx 1$ ) covers essentially the full relic-compatible parameter space, including the dominantly conversion-driven regime (characterized by small $\sin \theta$ ).
- As the mediator mass increases to 500 MeV and 1 GeV, the inelastic scattering cross-section is suppressed by the propagator. Consequently, the LZ sensitivity shifts toward larger mixing angles ( $\sin \theta$ ).
- For $m_{A'} = 1$ GeV, the overlap between the LZ-sensitive region and the deep conversion-driven regime (small mixing) is significantly reduced. Specifically, for large splittings ( $\delta = 300$ keV), the LZ contour no longer overlaps with the region where conversion drives the relic abundance.
Kinematic Constraints: The high-energy recoil favors mass splittings in the range $\delta \sim 150\text{--}300$ keV. Splittings smaller than this place the event deep in the suppressed high-energy tail of the spectrum, while splittings approaching $\sim 350$ keV hit the kinematic endpoint of the Galactic DM velocity distribution.

Significance
The paper claims that the high-energy LZ recoil offers a unique opportunity to directly test the interaction structure responsible for the non-standard thermal history of DM. Unlike scenarios where direct detection and relic density probe different sectors, this work establishes that in the coscattering regime, the same physical parameters (dark coupling, mixing, and mediator mass) control both the early Universe freeze-out and the present-day scattering event. The light dark photon acts as a crucial bridge, making a regime that would otherwise be invisible to direct detection experimentally accessible. Thus, the LZ event, if interpreted within this framework, would provide a probe of the same conversion mechanism across vastly different cosmological epochs.

More like this