GeV-Scale Dark Vector Models

• GeV-scale dark vector models are theoretical frameworks where new vector bosons (e.g., dark photons) mediate interactions between the Standard Model and a hidden dark sector.
• They employ kinetic mixing and symmetry-breaking mechanisms that naturally generate a GeV mass scale, crucial for viable dark matter freeze-out and asymmetric dark matter scenarios.
• Experimental strategies—including direct, indirect, and collider searches—are tailored to probe these models, with future detectors expected to explore subthermal couplings and resonance regions.

GeV-scale dark vector models describe scenarios where new vector bosons with masses at or below the GeV-scale mediate interactions between Standard Model (SM) fields and particles in a hidden or dark sector. These models have become prominent contexts for dark matter (DM) phenomenology, collider searches, astrophysical observations, and cosmological analysis, encompassing a broad range of constructions including “dark photons,” $U(1)_{B-L}$/$U(1)_B$ gauge extensions, photophobic vectors, and Higgs-portal scenarios. The GeV scale is theoretically motivated by UV-to-IR mechanisms, dynamical symmetry breaking, and empirical signals in direct/indirect detection and collider physics.

1. Generation of the GeV Mass Scale and Kinetic Mixing Mechanisms

The canonical realization introduces a hidden sector charged under a new Abelian gauge group $U(1)_d$, with the corresponding gauge boson ($\gamma_d$ or $A^\prime$) acquiring a GeV-scale mass~\cite{(Cohen et al., 2010)}. Kinetic mixing between $U(1)_d$ and the SM hypercharge is induced by loops of heavy states carrying both charges,

$$\epsilon \sim \frac{g_Y g_d}{16\pi^2}\ln\frac{M'}{M}$$

with $g_Y,g_d$ the respective gauge couplings. For loop-suppressed $\epsilon\sim 10^{-3}$ and a typical D-term $\langle D_Y\rangle \sim (72\,{\rm GeV})^2$, spontaneous symmetry breaking in the dark sector yields a dark Higgs VEV

$U(1)_B$0

which is naturally at the GeV scale. The corresponding dark photon mass is $U(1)_B$1.

This dynamical IR scale can be realized in both supersymmetric and non-supersymmetric setups and is a robust prediction of many UV completions involving new gauge bosons. The resulting models cover a range of dark sector mass scales, interaction strengths, and possible anomaly-canceling matter. Notably, gauge kinetic mixing links the weak scale to the dark sector, establishing the natural appearance of GeV masses beneath the electroweak scale.

2. Dark Vector as Mediator: Thermalization, Freeze-Out, and Asymmetric Mechanisms

GeV-scale dark vector mediators can efficiently deplete the symmetric component of the dark matter relic abundance through $U(1)_B$2, with $U(1)_B$3 subsequently decaying to SM states via kinetic mixing~\cite{(Cohen et al., 2010)}. This process is crucial in asymmetric dark matter (ADM) scenarios, where the relic density is set by a primordial asymmetry, as well as in standard thermal (WIMP) production. For GeV-scale DM, freeze-out cross sections are naturally consistent with the observed abundance for vector couplings $U(1)_B$4 in the range $U(1)_B$5--$U(1)_B$6 depending on masses and portal structure.

The symmetric freeze-out abundance is suppressed if

$U(1)_B$7

so the relic density follows the asymmetry. This mechanism is robust over a wide range of vector masses and allows phenomenologically viable models.

In addition, higher-dimensional operators such as

$U(1)_B$8

mediate $U(1)_B$9 asymmetry transfer between the visible and dark sectors, fixing the DM mass to $U(1)_d$0 for $U(1)_d$1 the $U(1)_d$2 charge of the dark matter candidate~\cite{(Ibe et al., 2011)}. These broad mechanisms operate in both Abelian and non-Abelian dark vector models.

3. Direct and Indirect Detection Phenomenology

Direct Detection

GeV-scale dark vectors mediate DM-nucleon interactions via both tree and loop-level processes. For dark photons, loop-induced couplings lead to predicted direct detection cross sections~\cite{(Cohen et al., 2010)}

$U(1)_d$3

for suitable scalar couplings $U(1)_d$4, covering the $U(1)_d$5--$U(1)_d$6 GeV mass range. For Higgs-portal models, $U(1)_d$7 exchange cross sections scale with singlet-doublet scalar mixing and the gauge coupling, remaining within the reach of LZ, XENON1T, and future detectors~\cite{(Das et al., 22 May 2025)}. Nuclear recoil energies are enhanced for low-mass nuclear targets (H, He), offering advantages in detectors such as NEWS~\cite{(Profumo, 2015)}.

Indirect Detection and Astrophysical Constraints

Annihilation via a light dark vector produces distinct indirect signals in γ-rays, X-rays, and cosmic-ray $U(1)_d$8. For vector-portal models with $U(1)_d$9 few GeV, annihilation to mesonic states and multiphoton final states is accurately computed using chiral perturbation theory~\cite{(Coogan et al., 2021)}, capturing complex hadronic branching and diffuse γ-ray signals.

Only models that avoid overproduction of secondary emissions or satisfy CMB constraints for s-wave annihilation cross sections $\gamma_d$0 remain viable in the GeV mass window~\cite{(Cirelli et al., 5 Aug 2025)}. Future MeV-range telescopes (COSI, AMEGO, GECCO) will probe orders of magnitude deeper, making sub-thermal cross sections accessible~\cite{(Coogan et al., 2021, Cirelli et al., 5 Aug 2025)}.

Astrophysical constraints from stellar cooling (plasmon decay, Compton scattering, and bremsstrahlung) as well as from SN1987A restrict sub-GeV multipole vector DM models, with higher-dimensional electromagnetic form factor couplings ($\gamma_d$1, $\gamma_d$2, etc.) tightly constrained by anomalous stellar energy loss~\cite{(Chu et al., 2023)}.

4. Collider Searches and Form Factor-Driven Production Rates

GeV-scale dark vectors are actively searched for at both $\gamma_d$3 flavor factories (KLOE, Belle II), proton fixed-target, and forward collider experiments~\cite{(Barze' et al., 2010, Kling et al., 11 Sep 2025)}. At $\gamma_d$4 machines, dark photons are produced via $\gamma_d$5 with $\gamma_d$6 decaying to $\gamma_d$7, producing a narrow resonance in $\gamma_d$8. Sensitivities to kinetic mixing as small as $\gamma_d$9--$A^\prime$0 are achieved for integrated luminosities $A^\prime$1--$A^\prime$2.

In high-energy $A^\prime$3 collisions, dark vectors are produced via bremsstrahlung off protons and neutrons, with amplitudes governed by nucleon timelike vector form factors~\cite{(Kling et al., 11 Sep 2025)}. The physically motivated resonance-based form factor model constructed with Breit–Wigner sums over ω, φ, ρ states enforces normalization at $A^\prime$4 and imposed QCD-motivated fall-off at large $A^\prime$5, enabling its application to generic charge assignments for both dark photon and non-photophilic vectors (e.g., $A^\prime$6, $A^\prime$7, or protophobic models).

Experimental yields are sensitive to both the form factor normalization and its resonance structure, impacting the predicted acceptance for forward detectors such as FASER. The inclusion of neutron-induced production and a careful account of nuclear damping and off-shell effects extend reach and reduce uncertainties relative to older approximations. This framework enables the reinterpretation of bounds and projections for a wide range of dark vector models and associated charges.

5. Model Variants: $A^\prime$8, Higgs-Portal, Trinification, and Multipole Scenarios

GeV-scale dark vector models are not monolithic, encompassing a diversity of UV embeddings:

• $A^\prime$9-based models ensure dark matter stability by exploiting the residual $U(1)_d$0 symmetry after gauge symmetry breaking, linking the DM mass to its $U(1)_d$1 charge and generating associated GeV-scale vectors that mediate both relic abundance and collider signatures~\cite{(Ibe et al., 2011)}.
• Higgs-portal constructions with dark vector dark matter and extra singlet scalars (2HDM+S, renormalizable U(1) extensions) utilize scalar mixing to communicate between the dark and visible sectors, accommodating parameter regimes with $U(1)_d$2--$U(1)_d$3~GeV and suppressed direct detection cross sections~\cite{(Das et al., 22 May 2025, Ko et al., 2014)}.
• Models with electromagnetic multipole interactions explore higher-dimensional couplings to the photon (magnetic/electric dipole, charge radius, toroidal/anapole), presenting unique stellar and cosmological constraints and requiring UV completion to regulate unitarity and the $U(1)_d$4 limit~\cite{(Chu et al., 2023)}.
• Trinification scenarios embed the vector boson DM in the gauge structure $U(1)_d$5, naturally generating $U(1)_d$6-odd, stable vector bosons and off-diagonal interactions with SM and vectorlike fermions, with mass limits $U(1)_d$7~GeV and LHC bounds on companion vectorlike fermions up to several TeV~\cite{(Babu et al., 2021)}.

6. Constraints from Cosmology, Cosmic-Rays, and Direct Detection Experimental Design

Strongly-interacting GeV-scale DM, particularly that communicated by light vector mediators, faces severe constraints from cosmic-ray upscattering. Even regions where conventional underground experiments lose sensitivity due to atmospheric and rock overburden suppression are tightly excluded when accounting for upscattered relativistic DM fluxes generated by cosmic-ray collisions~\cite{(Alvey et al., 2022)}. Cross sections above a few $U(1)_d$8 for GeV masses are essentially ruled out, regardless of momentum-dependent nuclear structure corrections. This result is robust against changes in the DM-nucleus interaction model, including both light mediator and puffy dark matter realizations.

Direct detection strategy evolves accordingly: light-target, low-threshold detectors (NEWS, SENSEI) extend the reach to GeV and sub-GeV DM masses~\cite{(Profumo, 2015)}. These designs are particularly effective for vector-mediated DM, filling gaps in the parameter space not covered by standard dual-phase noble liquid detectors.

7. Future Experimental Reach and Theoretical Implications

The remaining viable parameter space for GeV-scale thermal vector-mediated DM is narrow—often limited to resonance regions where $U(1)_d$9 and subthermal dark couplings $$\epsilon \sim \frac{g_Y g_d}{16\pi^2}\ln\frac{M'}{M}$$0 to $$\epsilon \sim \frac{g_Y g_d}{16\pi^2}\ln\frac{M'}{M}$$1~\cite{(Alonso-González et al., 15 Jul 2025)}. Upcoming direct detection (DARWIN/XLZD), advanced fixed-target, flavor factory, and forward detector experiments (FASER, FORMOSA, Belle II, COSI) will probe or close these allowed windows, while high-precision MeV–GeV telescopes (AMEGO, GECCO) target indirect signatures via combined prompt and secondary emission.

Continued theoretical advances, especially in nucleon/nucleus form factor modeling for production and detection, remain critical for accurate interpretation of all leading and next-generation experiments~\cite{(Kling et al., 11 Sep 2025)}. The explicit, resonance-based description of nucleon form factors now enables broader class analyses for arbitrary vector gauge boson charges and couplings.

---

References:

• (Cohen et al., 2010, Ibe et al., 2011, Barze' et al., 2010, Abe et al., 2012, Farzan et al., 2012, Choi et al., 2013, Ko et al., 2014, Izaguirre et al., 2014, Profumo, 2015, Kawamura et al., 2016, Davoudiasl et al., 2018, Chakraborti et al., 2018, Coogan et al., 2021, Babu et al., 2021, Alvey et al., 2022, Chu et al., 2023, Das et al., 22 May 2025, Alonso-González et al., 15 Jul 2025, Cirelli et al., 5 Aug 2025, Kling et al., 11 Sep 2025)