Blazar-Boosted Dark Matter Overview

• Blazar-boosted dark matter is a non-thermal population formed when cold dark matter near active galactic nuclei is accelerated by relativistic jet particles.
• It connects astrophysical signals from neutrinos and gamma rays with Earth-side detection via elastic scattering and deep inelastic processes.
• Detection constraints depend critically on dark-matter spike profiles and jet modeling, influencing both mediator interactions and multi-messenger observations.

Blazar-boosted dark matter denotes a non-thermal dark-matter population generated when initially cold dark matter in the immediate environment of an active galactic nucleus is accelerated by relativistic particles in a blazar jet. In the elastic realization, the boosted dark matter free-streams to Earth and can induce electron or nuclear recoils in neutrino or direct-detection experiments; in the inelastic realization, the same dark-matter–proton interactions inside the source generate hadronic showers and therefore neutrinos and gamma rays observable from the blazar itself. The modern literature has treated both aspects in a unified astroparticle framework, with TXS 0506+056 and BL Lacertae as the main benchmark sources and with later extensions to AP Librae, Centaurus A, and stacked blazar samples (Wang et al., 2021, Granelli et al., 2022, Marchi et al., 2024, Marchi et al., 6 Jun 2025).

1. Origin and scope of the concept

Blazar-boosted dark matter emerged as a specific realization of the broader boosted-dark-matter program in which a cold dark-matter population is upscattered by energetic Standard Model particles. Its distinctive feature is the conjunction of two extreme environments: a relativistic jet with bulk Lorentz factor typically in the range $\Gamma_B \sim 10$–40, and a dense dark-matter spike around a supermassive black hole. Early treatments emphasized the Earth-bound flux of boosted dark matter generated by elastic proton–dark-matter collisions in blazar jets and constrained it with XENON1T, MiniBooNE, Borexino, and Super-Kamiokande (Wang et al., 2021, Granelli et al., 2022). Subsequent work extended the same setup to electron-jet boosting, explicit mediator models, xenon-TPC response modeling, and deep inelastic scattering, thereby connecting blazar dark-matter interactions to source neutrinos and gamma rays as well as to direct detection (Bhowmick et al., 2022, Barillier et al., 8 Sep 2025, Wang, 28 Mar 2025).

The field contains two observational branches. One branch treats the boosted dark matter itself as the signal and searches for its scattering on electrons or nuclei in terrestrial detectors. The other treats dark matter as a target in the source, so that proton–dark-matter deep inelastic scattering produces a neutrino and gamma-ray signal from the blazar. Later work made this distinction explicit, noting that earlier studies emphasized dark matter accelerated by blazar jets and detected on Earth, whereas the newer neutrino-oriented literature uses the same proton–dark-matter interactions to explain high-energy neutrinos from TXS 0506+056 and even the diffuse astrophysical neutrino flux at high energies (Marchi et al., 2024, Marchi et al., 6 Jun 2025, Marchi et al., 16 Jul 2025).

A common misconception is that the relevant dark matter is the local Milky Way halo. In the canonical blazar-boosted setup, the target population is instead the dark matter gravitationally bound to the blazar host halo and, especially, to the central spike around the black hole. Another common misconception is that blazar-boosted dark matter is synonymous with one detector class. In practice, the literature spans water Cherenkov detectors, scintillator experiments, germanium detectors, and xenon time projection chambers, as well as source-side neutrino and gamma-ray observations (Xu et al., 2024, Xia et al., 2024, Barillier et al., 8 Sep 2025).

2. Source environment and acceleration mechanism

The basic astrophysical picture begins with a Gondolo–Silk-type dark-matter spike around the supermassive black hole. Starting from an NFW-like inner halo $\rho_\chi^{\rm halo}(r)=\mathcal{N} r^{-1}$, adiabatic black-hole growth steepens the profile to $\rho_\chi^{\rm spike}(r)\propto r^{-7/3}$ for $\gamma=1$, often supplemented by a relativistic capture factor $g(r)=\left(1-2R_S/r\right)^{3/2}$ close to the horizon. Different papers normalize this spike either by requiring the dark-matter mass in the spike to be comparable to $M_{\rm BH}$ inside $4R_S$–$10^5R_S$, or by imposing a spike radius $R_{\rm sp}\simeq R_\star\sim 10^6R_S$ and a spike mass of order $10\%$–$\rho_\chi^{\rm halo}(r)=\mathcal{N} r^{-1}$0 of the black-hole mass (Xu et al., 2024, Marchi et al., 2024, Marchi et al., 6 Jun 2025).

The key line-of-sight quantity is the dark-matter column density along the jet,

$\rho_\chi^{\rm halo}(r)=\mathcal{N} r^{-1}$1

In benchmark TXS 0506+056 models with $\rho_\chi^{\rm halo}(r)=\mathcal{N} r^{-1}$2 or $\rho_\chi^{\rm halo}(r)=\mathcal{N} r^{-1}$3, this gives $\rho_\chi^{\rm halo}(r)=\mathcal{N} r^{-1}$4 and $\rho_\chi^{\rm halo}(r)=\mathcal{N} r^{-1}$5, respectively. These two cases, usually labeled BMCI and BMCII, quantify the strong dependence of the signal on the uncertain inner overlap between the accelerated jet and the dark-matter spike (Marchi et al., 2024, Marchi et al., 6 Jun 2025).

Jets are modeled as relativistically moving blobs whose electrons or protons are isotropic in the blob frame and obey power-law distributions. In the observer frame, the proton emission rate used in several analyses is

$\rho_\chi^{\rm halo}(r)=\mathcal{N} r^{-1}$6

with analogous expressions for electrons. TXS 0506+056 and BL Lacertae are the two standard benchmarks. For TXS 0506+056, representative parameters include $\rho_\chi^{\rm halo}(r)=\mathcal{N} r^{-1}$7, $\rho_\chi^{\rm halo}(r)=\mathcal{N} r^{-1}$8, $\rho_\chi^{\rm halo}(r)=\mathcal{N} r^{-1}$9, $\rho_\chi^{\rm spike}(r)\propto r^{-7/3}$0, $\rho_\chi^{\rm spike}(r)\propto r^{-7/3}$1, $\rho_\chi^{\rm spike}(r)\propto r^{-7/3}$2, and $\rho_\chi^{\rm spike}(r)\propto r^{-7/3}$3 in one CDEX-10 study, while neutrino-oriented analyses also use $\rho_\chi^{\rm spike}(r)\propto r^{-7/3}$4 and $\rho_\chi^{\rm spike}(r)\propto r^{-7/3}$5 from Keivani et al. fits. For BL Lacertae, representative values are $\rho_\chi^{\rm spike}(r)\propto r^{-7/3}$6, $\rho_\chi^{\rm spike}(r)\propto r^{-7/3}$7, $\rho_\chi^{\rm spike}(r)\propto r^{-7/3}$8, $\rho_\chi^{\rm spike}(r)\propto r^{-7/3}$9, $\gamma=1$0, $\gamma=1$1, and $\gamma=1$2 (Xu et al., 2024, Marchi et al., 2024).

The acceleration kinematics is controlled by elastic two-body scattering. For proton boosting, the maximal outgoing dark-matter kinetic energy is

$\gamma=1$3

and analogous formulas hold for electron boosting. Electron-jet models sometimes express the outgoing dark-matter energy directly as

$\gamma=1$4

with $\gamma=1$5. Because the jet is collimated and the scattering is forward-peaked in the observer frame, a large fraction of the boosted dark matter is emitted close to the jet axis, which for a blazar is close to the line of sight to Earth (Xu et al., 2024, Xia et al., 2024, Granelli et al., 2022).

3. Interaction models and scattering regimes

The earliest phenomenology often parameterized the source and detector interactions by constant elastic cross sections, such as $\gamma=1$6 or $\gamma=1$7, supplemented by proton or nuclear form factors. In that language, the boosted-dark-matter flux at Earth from a proton jet is written schematically as

$\gamma=1$8

and the event rate in a detector scales with a second power of the relevant detector cross section. This approximation underlies the early XENON1T, MiniBooNE, Borexino, and Super-Kamiokande analyses, as well as the later CDEX-10 reinterpretation in terms of spin-independent DM–nucleon scattering (Wang et al., 2021, Granelli et al., 2022, Xu et al., 2024).

Later work replaced the constant-cross-section approximation by explicit simplified models. The most studied possibilities are scalar, vector, axial-vector, and pseudoscalar mediators coupling a Dirac fermion $\gamma=1$9 to quarks or electrons. Representative Lagrangians are

$g(r)=\left(1-2R_S/r\right)^{3/2}$0

$g(r)=\left(1-2R_S/r\right)^{3/2}$1

$g(r)=\left(1-2R_S/r\right)^{3/2}$2

$g(r)=\left(1-2R_S/r\right)^{3/2}$3

together with electrophilic scalar or vector couplings in electron-boosting scenarios (Marchi et al., 2024, Marchi et al., 6 Jun 2025, Wang, 28 Mar 2025, Bhowmick et al., 2022).

For low-energy comparison with direct detection, these models are often mapped onto a non-relativistic proxy cross section. In vector-portal proton scattering one commonly uses

$g(r)=\left(1-2R_S/r\right)^{3/2}$4

while neutrino-oriented studies use the equivalent $g(r)=\left(1-2R_S/r\right)^{3/2}$5. This identification is only a proxy: the high-energy source process can be deep inelastic scattering with a full $g(r)=\left(1-2R_S/r\right)^{3/2}$6-dependent amplitude, whereas the detector comparison is made to the zero-momentum-transfer limit (Marchi et al., 2024, Wang, 28 Mar 2025, Marchi et al., 6 Jun 2025).

A central development was the recognition that energy dependence matters. For electrophilic scalar and vector mediators, the full $g(r)=\left(1-2R_S/r\right)^{3/2}$7-channel matrix element significantly alters both the source flux and the detector recoil spectrum relative to the constant-$g(r)=\left(1-2R_S/r\right)^{3/2}$8 approximation; in the heavy-mediator regime the revised exclusions can be orders of magnitude stronger, whereas in the light-mediator regime the large-$g(r)=\left(1-2R_S/r\right)^{3/2}$9 propagator suppression weakens them (Bhowmick et al., 2022). A plausible implication is that constant-cross-section analyses should be regarded as baseline parameterizations rather than universal descriptions.

The other major shift was the inclusion of deep inelastic scattering. When relativistic jet protons encounter sub-GeV dark matter, the center-of-mass energy can exceed the hadronic scale, and the interaction becomes

$M_{\rm BH}$0

In this regime the proton is effectively destroyed, pions are produced, and the decay chains $M_{\rm BH}$1 and $M_{\rm BH}$2 yield neutrinos and gamma rays. This DIS component is the basis of the source-neutrino literature and is also what extends boosted-dark-matter spectra to higher energies in recent multi-messenger analyses (Marchi et al., 2024, Wang, 28 Mar 2025, Marchi et al., 6 Jun 2025).

4. Detection channels and terrestrial constraints

The direct-detection branch of blazar-boosted dark matter has produced constraints from water Cherenkov detectors, scintillator experiments, germanium detectors, and xenon TPCs. In the proton-coupling case, the first systematic direct-detection study used XENON1T, MiniBooNE, and Borexino data and found that, depending on source modeling, the resulting bounds on spin-independent and spin-dependent DM–proton cross sections improve on other available limits for light dark matter candidates by 1 up to 5 orders of magnitude (Wang et al., 2021).

For DM–electron scattering, Super-Kamiokande has been the canonical probe. A dedicated analysis of TXS 0506+056 and BL Lacertae derived limits on $M_{\rm BH}$3 from directional electron recoils and found that for dark-matter masses below $M_{\rm BH}$4 the exclusion can reach $M_{\rm BH}$5, with the strongest limits arising for BL Lacertae under optimistic spike assumptions (Granelli et al., 2022). A later study of Centaurus A showed that a misaligned but nearby AGN can compensate geometric suppression with proximity and push the Super-Kamiokande reach on the DM–electron cross section down to $M_{\rm BH}$6 for $M_{\rm BH}$7 dark matter (Xia et al., 2024).

Solid-state and xenon detectors extend the same logic to nuclear recoils. Using 205.4 kg·day of CDEX-10 data at CJPL, one study derived $M_{\rm BH}$8 C.L. exclusions on spin-independent DM–nucleon scattering from $M_{\rm BH}$9 to $4R_S$0 for TXS 0506+56 and from $4R_S$1 to $4R_S$2 for BL Lacertae over $4R_S$3. In the sub-MeV region these were identified as the best sensitivities among solid-state detector experiments (Xu et al., 2024). Xenon-based reinterpretations later incorporated explicit detector response: with TXS 0506+056 as the source, XENON1T excludes $4R_S$4 near $4R_S$5, while LZ EFT searches constrain $4R_S$6 (Barillier et al., 8 Sep 2025).

The detector rate always depends on a convolution of the BBDM flux with elastic scattering in the detector medium. For nuclear recoils in CDEX-10, for example,

$4R_S$7

while electron-recoil searches use the analogous $4R_S$8 kernel (Xu et al., 2024, Granelli et al., 2022). Because the source flux itself is proportional to the source scattering cross section, and the detector signal requires another scattering, event rates often scale quadratically with the underlying DM–SM coupling in the purely elastic literature.

A characteristic feature of these exclusions is that they are frequently bounded both from below and from above in cross section. The lower edge reflects the minimum interaction strength needed to produce a detectable rate, whereas the upper edge arises because the Earth, mountain overburden, or detector surroundings attenuate or isotropize the flux for highly interacting dark matter. This behavior is explicit in CDEX-10 and xenon-TPC analyses and also appears in Super-Kamiokande studies of electron-coupled BBDM (Xu et al., 2024, Xia et al., 2024, Barillier et al., 8 Sep 2025).

5. Neutrinos, gamma rays, and multi-messenger realizations

The inelastic branch of the subject recasts the blazar spike as a dense hadronic target. In this picture, relativistic jet protons undergo deep inelastic scattering on sub-GeV dark matter, producing hadronic showers and therefore neutrinos and gamma rays. One analysis showed that, within realistic lepto-hadronic jet models and a conservative spike normalization, the additional proton–dark-matter DIS channel can explain the TXS 0506+056 neutrino observations for dark-matter parameters allowed by laboratory, direct, and indirect searches, and argued that this was the first systematic connection between DM–proton DIS in blazar jets and the observed IceCube neutrino from TXS 0506+056 (Marchi et al., 2024).

The neutrino flux from a single blazar is commonly written in the form

$4R_S$9

with the factor $10^5R_S$0 accounting for flavor equipartition after oscillations. In this formulation the same spike column density $10^5R_S$1 that governs BBDM production also controls the source neutrino luminosity (Marchi et al., 2024, Marchi et al., 6 Jun 2025).

A major extension was the demonstration that DM-induced neutrinos from a stacked sample of blazars can saturate the diffuse astrophysical neutrino flux observed by IceCube at high energies. In a sample of 324 blazars, the cumulative DM-induced flux for a spin-1 mediator with $10^5R_S$2, $10^5R_S$3, couplings $10^5R_S$4, and BMCII can saturate IceCube’s observed diffuse muon-neutrino flux above $10^5R_S$5. In that framework BL Lacs dominate over FSRQs, in contrast to many standard $10^5R_S$6 models, because the DM-induced neutrino luminosity scales roughly with proton luminosity and spike column density rather than with target photon density (Marchi et al., 6 Jun 2025).

Multi-messenger analyses strengthened this point by adding explicit gamma-ray and neutrino secondaries from elastic and inelastic DM–proton scattering in vector-portal models. These studies found that the distinctive high-energy gamma-ray and neutrino fluxes from DIS provide constraints on $10^5R_S$7 that significantly surpass those from traditional direct detection experiments and can improve on earlier constant-cross-section, pure-elastic BBDM limits by several orders of magnitude (Wang, 28 Mar 2025). A later comparison of BBDM searches against DM-induced neutrino interpretations concluded that proton-recoil searches at Super-Kamiokande, KamLAND, Borexino, JUNO, Hyper-Kamiokande, and DUNE leave room for a variety of DM models to explain the neutrino data, and that neutrinos are often the more promising first handle on these interactions (Marchi et al., 16 Jul 2025).

These results clarify that blazar-boosted dark matter is not a single observable but a linked program: at the source, $10^5R_S$8 DIS produces neutrinos and gamma rays; in transit, boosted $10^5R_S$9 free-streams; at Earth, the same $R_{\rm sp}\simeq R_\star\sim 10^6R_S$0 may scatter in detectors. This suggests that the most discriminating tests will be joint ones that confront source-side neutrino and gamma-ray spectra with Earth-side boosted-dark-matter searches.

6. Systematics, controversies, and prospective directions

The dominant uncertainty throughout the subject is the dark-matter distribution near the supermassive black hole. Nearly every analysis emphasizes that the signal normalization is controlled first by the spike profile and its inner cutoff, not by detector systematics. Assumptions about adiabatic growth, annihilation-induced cores, stellar heating, mergers, and the minimum jet radius $R_{\rm sp}\simeq R_\star\sim 10^6R_S$1 or $R_{\rm sp}\simeq R_\star\sim 10^6R_S$2 can change $R_{\rm sp}\simeq R_\star\sim 10^6R_S$3 by orders of magnitude, and therefore shift inferred limits on cross sections by comparable amounts (Xu et al., 2024, Marchi et al., 2024, Barillier et al., 8 Sep 2025). A related controversy concerns whether long-lived dense spikes survive around realistic AGN; the literature therefore uses benchmark profiles such as BMP1/BMP2/BMP3 or BMCI/BMCII rather than claiming a unique astrophysical prediction.

Jet modeling is the second major systematic. Proton luminosity $R_{\rm sp}\simeq R_\star\sim 10^6R_S$4, spectral index $R_{\rm sp}\simeq R_\star\sim 10^6R_S$5, maximum particle energy, Doppler factor, and line-of-sight angle all directly enter the source flux. The papers repeatedly note that blazars are highly variable and that SED fits correspond to specific epochs. Time variability also complicates interpretation: source neutrinos arrive promptly, whereas BBDM can be delayed by its subluminal propagation, so transient source activity is smeared over long travel times (Marchi et al., 2024, Marchi et al., 16 Jul 2025).

There are also particle-physics ambiguities. Constant elastic cross sections are useful baselines but are not generic; mediator mass and Lorentz structure change both the energy dependence and the relevant terrestrial constraints. Vector and scalar mediators tend to map cleanly onto spin-independent limits, axial vectors to spin-dependent limits, and pseudoscalars to momentum-suppressed scattering with important loop-induced effects in direct detection (Bhowmick et al., 2022, Marchi et al., 6 Jun 2025, Marchi et al., 16 Jul 2025). This suggests that a complete treatment of blazar-boosted dark matter is inseparable from the mediator model.

A further issue is self-consistency of the spike itself. Recent work has checked whether DM–proton scatterings in the jet or DM self-interactions can deplete the spike enough to invalidate the signal. For stacked blazar populations, the depletion induced by DM–proton and DM–DM interactions was found not to compromise the DM interpretation for high-energy neutrinos, though it can challenge other blazar–DM signals under more aggressive assumptions about long-lived jet activity (Marchi et al., 16 Jul 2025).

Prospective directions are already well defined in the literature. They include additional blazar targets, stacked analyses of hundreds of sources, larger exposures and lower thresholds in germanium or silicon detectors, end-to-end detector-response modeling in xenon experiments, and proton-recoil searches in Hyper-Kamiokande, DUNE, and JUNO (Xu et al., 2024, Barillier et al., 8 Sep 2025, Marchi et al., 16 Jul 2025). This suggests that the next phase of the subject will be driven less by new kinematic ideas than by improved astrophysical modeling of SMBH environments and by genuinely multi-messenger fits that connect direct detection, neutrino astronomy, and gamma-ray data within a common blazar–dark-matter framework.