⚛️ phenomenology

Boosted or Inelastic? Discriminating Interpretations of the LZ 248 keV Event

This paper proposes that the single 248 keV nuclear recoil event observed by the LUX-ZEPLIN experiment can be explained by either endothermic inelastic dark matter or boosted dark matter, demonstrating that these two scenarios are spectrally distinguishable across different effective operators and can be definitively resolved with additional data from the full LZ exposure.

Original authors: Satyabrata Mahapatra, Partha Kumar Paul

Published 2026-09-15

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Satyabrata Mahapatra, Partha Kumar Paul

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 the existence of invisible matter comes from how stars and galaxies move. Astronomers have long observed that visible stars orbit the centers of galaxies too quickly to be held in place by the gravity of the stars and gas we can see. Something else must be there, providing the extra gravitational pull to keep these cosmic structures from flying apart. This invisible substance, known as dark matter, is estimated to make up about five times more of the universe's energy than all the ordinary matter we are made of. Despite its overwhelming presence, no particle of dark matter has ever been directly detected. The leading theory suggests these particles are heavy and move relatively slowly, bumping into ordinary atomic nuclei in underground detectors and causing a tiny, measurable kick. However, for years, the most sensitive detectors have found nothing, pushing the limits of what these particles could be and forcing scientists to consider more exotic possibilities.

Recently, the LUX-ZEPLIN experiment, a massive detector buried deep underground in South Dakota, reported a single, intriguing event. The detector, filled with liquid xenon, recorded a nuclear recoil with an energy of 248 keV. While this energy level is high enough to be caused by a dark matter particle, the nature of the event poses a significant puzzle. In the standard picture of dark matter, particles rain down on Earth from the galactic halo, and their collisions with atomic nuclei produce a spectrum of energies that is highest at the lowest levels and falls off steadily as energy increases. If a heavy dark matter particle were responsible for this 248 keV kick, it should have produced many more, much smaller kicks at lower energies. The fact that the detector saw this single high-energy event without any accompanying swarm of low-energy events makes the standard explanation highly unlikely.

To solve this mystery, researchers Satyabrata Mahapatra and Partha Kumar Paul have proposed two distinct scenarios that could explain the event without violating the laws of physics. Both ideas rely on the dark matter particle behaving differently than the standard, slow-moving type. The first possibility is "inelastic" dark matter. In this scenario, the dark matter particle exists in two states with slightly different masses. When it hits a nucleus, it must jump to the heavier state, a process that requires a specific amount of energy to occur. This energy threshold acts like a gate, preventing the particle from creating low-energy recoils. It can only produce a kick if it has enough speed to pay the energy cost of the jump, which naturally filters out the small events and leaves only the high-energy ones. The second possibility is "boosted" dark matter. Here, the dark matter particles are not the slow, heavy residents of the galactic halo, but rather light particles that have been accelerated to near-light speeds by some energetic process within the dark sector. These fast-moving particles carry enough momentum to create a large recoil even if they are very light, bypassing the need for the heavy mass required in the standard model.

The researchers tested both of these ideas against the single observed event using a flexible mathematical framework that does not assume a specific type of interaction between dark matter and ordinary matter. They examined how different types of forces would shape the pattern of energy recoils. They found that the two scenarios leave very different fingerprints. The inelastic dark matter model naturally concentrates events near the observed energy, regardless of the specific force involved, because the energy threshold blocks the low-energy events entirely. In contrast, the boosted dark matter model produces a spectrum that depends heavily on the nature of the force. For some types of forces, the boosted model would predict that almost all events happen at low energies, which contradicts the observation. However, for a specific type of force involving a particular symmetry, the boosted model naturally places most of its events at high energies, matching the observation.

The study concludes that while both scenarios are physically possible, they can be distinguished by the pattern of future events. The inelastic model suggests that the mass difference between the two dark matter states is the key factor, and this value changes depending on the type of force. The boosted model suggests that the momentum dependence of the force is the critical element, with the specific force type determining whether the events cluster at high or low energies. The researchers emphasize that a single event is not enough to definitively prove either theory. However, if the LUX-ZEPLIN experiment continues to collect data, or if future detectors with even greater sensitivity come online, a handful of additional events would be sufficient to separate these two possibilities. The presence or absence of low-energy events in the coming data will tell scientists whether the dark matter is jumping between states or racing through the detector at relativistic speeds, finally revealing the true nature of this elusive cosmic component.

Technical Summary: Boosted or Inelastic? Discriminating Interpretations of the LZ 248 keV Event

Problem Statement
The LUX-ZEPLIN (LZ) experiment reported a single nuclear recoil candidate at $E_R = 248 \pm 23(\text{stat}) \pm 23(\text{sys})$ keV. While this event disfavors the background-only hypothesis at a global significance of $2.6\sigma$ , it presents a kinematic challenge for standard Weakly Interacting Massive Particle (WIMP) interpretations. In standard elastic scattering of Galactic halo dark matter (DM), the recoil spectrum is monotonically falling. To generate a 248 keV recoil, a momentum transfer of $q \simeq 246$ MeV is required, implying a DM mass $m_\chi \gtrsim 79$ GeV. However, such a heavy WIMP would inevitably produce a significant excess of low-energy recoils, which is not observed in the data. The absence of this low-energy tail necessitates an explanation that kinematically suppresses events below the observed energy.

Methodology
The authors investigate two distinct kinematic mechanisms capable of evading the low-energy excess problem: Endothermic Inelastic Dark Matter (iDM) and Boosted Dark Matter (bDM).

Framework: The analysis employs a model-independent effective field theory (EFT) approach, utilizing dimension-six contact operators suppressed by a cutoff scale $\Lambda$ $Λ$ . Three specific operators are analyzed on equal footing:
1. Scalar–Scalar ( $O_{ss}$ ): $\frac{1}{\Lambda^2}(\bar{\chi}_1\chi_2)(\bar{N}N)$
2. Pseudoscalar–Scalar ( $O_{ps}$ ): $\frac{1}{\Lambda^2}(\bar{\chi}_1 i\gamma_5 \chi_2)(\bar{N}N)$
3. Pseudoscalar–Pseudoscalar ( $O_{pp}$ ): $\frac{1}{\Lambda^2}(\bar{\chi}_1 i\gamma_5 \chi_2)(\bar{N} i\gamma_5 N)$
iDM Scenario: Assumes a two-state dark sector with a mass splitting $\delta = m_{\chi_2} - m_{\chi_1} > 0$ . Scattering proceeds via $\chi_1 N \to \chi_2 N$ . The mass splitting imposes a minimum velocity threshold $v_{\min}$ that diverges as $E_R \to 0$ , kinematically forbidding low-energy recoils.
bDM Scenario: Assumes a sub-dominant, light DM species $\chi_1$ is produced relativistically via the annihilation of a heavier component $\chi_2$ in the Galactic halo ( $\chi_2\chi_2 \to \chi_1\chi_1$ ). The incident flux is monoenergetic with a Lorentz factor $\gamma = m_{\chi_2}/m_{\chi_1}$ , providing the necessary momentum transfer even for light DM masses.
Analysis: The authors construct full predicted recoil spectra, incorporating nuclear form factors (Helm), LZ detection efficiency, and energy resolution. They perform a profile likelihood analysis on the single observed event to map confidence regions in the relevant parameter spaces: $\{m_{\chi_1}, \delta, \Lambda\}$ for iDM and $\{m_{\chi_1}, v, \Lambda\}$ (or equivalently $\langle \sigma v \rangle$ ) for bDM.

Key Contributions and Results

Spectral Distinguishability: The two scenarios produce qualitatively different recoil spectra.
- iDM: For all three operators, the spectra are concentrated near the observed 248 keV energy. The mass splitting $\delta$ naturally suppresses low-energy events, with only 6–17% of events falling below 150 keV, depending on the operator.
- bDM: The spectral shape is critically dependent on the operator's momentum dependence.
  - $O_{ss}$ and $O_{ps}$ interactions place 99% and 92% of events, respectively, below 150 keV, failing to match the high-energy concentration of the LZ event without significant fine-tuning.
  - $O_{pp}$ , due to its $q^4$ momentum dependence, shifts the spectrum significantly, placing 75% of events above 150 keV. This makes $O_{pp}$ the preferred operator for the bDM interpretation of the LZ event.
Operator-Dependent Kinematics (iDM): In the iDM scenario, the preferred mass splitting $\delta$ decreases monotonically as the momentum dependence of the operator increases ( $O_{ss} \to O_{ps} \to O_{pp}$ ). Operators with higher powers of momentum transfer ( $q$ ) naturally weight the spectrum toward higher energies, partially compensating for a smaller mass splitting. This creates a degeneracy between the operator structure and $\delta$ , meaning a single event cannot uniquely determine the splitting without additional data.
Kinematic Constraints:
- iDM: The mass splitting is bounded by $\delta \lesssim 385$ keV for $m_{\chi_1} \gg m_N$ , determined by the maximum Galactic escape velocity.
- bDM: A kinematic boundary exists at $m_{\chi_1} \simeq 78.7$ GeV. For $m_{\chi_1} \gtrsim 78.7$ GeV, the velocity required to produce a 248 keV recoil is below the maximum halo velocity ( $v_{\max}$ ), meaning the observed recoil can be produced without requiring any additional boost. Conversely, for $m_{\chi_1} < 78.7$ GeV, the required velocity exceeds $v_{\max}$ , necessitating a boost to explain the event.
- EFT Validity: The authors identify a region where the effective cutoff scale $\Lambda$ drops below $\sim 69$ MeV (specifically for $O_{pp}$ with $v \approx 0.05c$ and $m_{\chi_1} \lesssim 22$ GeV). In this region, the EFT description breaks down as the mediator mass would be comparable to the momentum transfer.
Discrimination Potential: The paper demonstrates that the scenarios are distinguishable across the entire high-energy window. While the single event allows for multiple interpretations, the distinct low-energy tails predicted by different operators (especially in the bDM case) and the different spectral shapes of iDM vs. bDM mean that a handful of additional events would be sufficient to discriminate between these hypotheses.

Significance
The paper argues that the 248 keV LZ event cannot be explained by standard elastic WIMP scattering due to the lack of a low-energy excess. It establishes that both inelastic and boosted dark matter scenarios can kinematically accommodate the event while suppressing low-energy recoils. Crucially, the authors show that these two mechanisms are spectrally distinct and that the interpretation of a boosted signal is highly sensitive to the underlying effective operator. The work concludes that the full LZ exposure and next-generation xenon observatories will be capable of resolving the ambiguity between these scenarios or ruling them out entirely, placing the interpretation of this event within immediate experimental reach.

More like this