Vector-Like Lepton Dark Matter

• Vector-like lepton dark matter models are beyond-Standard-Model constructs featuring new leptons with identical left- and right-handed gauge quantum numbers, stabilized by discrete symmetries.
• These models analyze relic-density mechanisms through annihilation, co-annihilation, resonance effects, and direct detection constraints, with parameters often tuned via singlet-doublet mixing.
• Collider signatures such as displaced vertices and multi-lepton final states from compressed spectra offer tangible experimental tests of these dark-sector theories.

Searching arXiv for recent and foundational papers on vector-like lepton dark matter models. Vector-like lepton dark matter models are a class of beyond-the-Standard-Model constructions in which new leptons with identical gauge quantum numbers for their left- and right-handed components participate directly in the dark sector. In the literature, these models appear in several structurally distinct forms: the dark matter can be the neutral component of a vector-like lepton multiplet, a scalar coupled through a lepton portal built from vector-like leptons, or a vector state whose dominant interactions arise through vector-like lepton mixing. Across these realizations, three ingredients recur: an exact or remnant discrete symmetry that stabilizes the lightest dark-sector state, electroweak-scale or TeV-scale vector-like leptons, and a relic-density mechanism controlled by annihilation, co-annihilation, resonance effects, or non-thermal production (Bhattacharya et al., 2015, Schwaller et al., 2013, Bouzeraib et al., 10 Mar 2026).

1. Taxonomy and historical development

Early formulations emphasized the neutral component of an electroweak multiplet. In the singlet-doublet realization, the Standard Model is augmented by a vector-like doublet and a singlet, both odd under an unbroken $Z_2$, so that the lighter neutral mass eigenstate is stable and behaves as a weakly interacting massive particle (Bhattacharya et al., 2015). Closely related constructions later incorporated a scalar triplet, generating a pseudo-Dirac splitting that suppresses $Z$-mediated elastic scattering while connecting the dark sector to neutrino mass generation (Bhattacharya et al., 2017, Bhattacharya et al., 2018).

A second line of development treats vector-like leptons as a portal rather than as the dark matter itself. Kawamura, Okawa, and Omura introduced a lepton-portal model in which a real scalar dark matter field $X$ couples exclusively to extra vector-like leptons and the muon, allowing simultaneous discussion of the $W$-boson mass shift, the muon anomalous magnetic moment, and the relic density (Kawamura et al., 2022). Related scalar-portal structures appear in the inert Zee model, where new vector-like fermions open a lepton portal for inert scalar dark matter, and in $U(1)_{L_\mu-L_\tau}$ models where scalar dark matter co-annihilates with vector-like leptons (Gaviria et al., 2020, Zhou et al., 2022).

A third branch embeds vector-like leptons in enlarged gauge sectors. Examples include gauged lepton number, left-right symmetric models, an alternative left-right model with an extra non-abelian $SU(2)_V$, and dark $SU(2)_D$ portal models. In these cases, the stable dark matter candidate may be a mostly singlet vector-like neutrino, the neutral component of a new vector-like multiplet, or a dark gauge boson whose phenomenology is controlled by vector-like lepton interactions (Schwaller et al., 2013, Bahrami et al., 2016, Bouzeraib et al., 10 Mar 2026, Kim et al., 2022).

This diversity suggests that “vector-like lepton dark matter model” is best understood not as a single Lagrangian but as a model family unified by symmetry protection, vector-like matter content, and electroweak or TeV-scale dark-sector phenomenology.

2. Field content, gauge structure, and stabilizing symmetries

The minimal electroweak realization contains a vector-like $SU(2)_L$ doublet and a singlet,

$$N = \begin{pmatrix}N^0\ N^- \end{pmatrix},\qquad \chi^0,$$

with a $Z_2$ parity under which the new fermions are odd and all Standard Model fields are even. After electroweak symmetry breaking, the neutral states mix and the lighter mass eigenstate becomes the dark matter candidate (Bhattacharya et al., 2015). Reviews of this framework emphasize that the same $Z$0 symmetry is the central mechanism forbidding mixing with Standard Model leptons and ensuring stability (Bhattacharya et al., 2018).

Other realizations enlarge the field content substantially. In the gauged-lepton-number model of Schwaller, Tait, and Vega-Morales, anomaly cancellation requires exotic leptons and right-handed neutrinos, while a singlet scalar $Z$1 with lepton-number charge breaks $Z$2 and leaves an accidental global $Z$3 under which the exotic sector is odd; the lightest neutral exotic is then a stable Dirac dark matter state (Schwaller et al., 2013). In left-right models, one full vector-like lepton family plus mirrors is introduced, again with a parity that forbids mixing with ordinary leptons and stabilizes the lightest new neutral state (Bahrami et al., 2016). In the alternative left-right model with $Z$4, the vector-like leptons transform under the new gauge symmetry and a parity $Z$5 makes the neutral component $Z$6 stable on cosmological time scales (Bouzeraib et al., 10 Mar 2026).

Portal models use similar symmetry logic even when the dark matter is not itself a vector-like lepton. In the lepton-portal scalar model, the new odd fields are vector-like lepton doublets $Z$7, vector-like singlets $Z$8, and a real scalar dark matter field $Z$9, all odd under a $X$0 parity (Kawamura et al., 2022). In the inert Zee model, the exact $X$1 instead renders the inert scalar sector and the new vector-like fermions odd, with the relic density carried by the scalar $X$2 rather than the fermions (Gaviria et al., 2020).

Realization	Stable dark-sector state	Stabilizing symmetry
Singlet-doublet electroweak model	Lighter neutral VLL eigenstate $X$3	exact $X$4 (Bhattacharya et al., 2015)
Gauged lepton number	Dirac singlet-like neutral exotic $X$5	accidental global $X$6 (Schwaller et al., 2013)
Lepton-portal scalar DM	real scalar $X$7	$X$8 on $X$9 (Kawamura et al., 2022)
Alternative left-right with $W$0	neutral VLL component $W$1	parity $W$2 (Bouzeraib et al., 10 Mar 2026)
$W$3 lepton portal	dark gauge boson $W$4	remnant $W$5 (Kim et al., 2022)

A common misconception is that the phrase necessarily refers to a neutral vector-like lepton as dark matter. The literature does not support that restriction: scalar dark matter and vector dark matter realizations are both explicit and phenomenologically important (Kawamura et al., 2022, Kim et al., 2022).

3. Mass matrices, mixing patterns, and stability mechanisms

In the minimal singlet-doublet construction, the neutral mass matrix after electroweak symmetry breaking is

$W$6

with

$W$7

The lighter eigenstate $W$8 is stable and, for small $W$9, predominantly singlet-like (Bhattacharya et al., 2015). This mixing angle simultaneously governs relic annihilation, direct detection, and invisible $U(1)_{L_\mu-L_\tau}$0 decays, making it the central parameter of the minimal model.

Triplet extensions alter this structure qualitatively. When a hypercharge-2 scalar triplet develops an induced vev $U(1)_{L_\mu-L_\tau}$1, the vector-like doublet receives a Majorana mass $U(1)_{L_\mu-L_\tau}$2, splitting the Dirac dark matter into a pseudo-Dirac pair with $U(1)_{L_\mu-L_\tau}$3. This suppresses elastic $U(1)_{L_\mu-L_\tau}$4 exchange once $U(1)_{L_\mu-L_\tau}$5, thereby relaxing direct-detection limits (Bhattacharya et al., 2017). The same induced vev also generates active neutrino masses, so the dark-matter mass splitting and neutrino sector become linked.

Portal models exhibit different but equally consequential mixing structures. In the scalar muon-portal model, the charged vector-like leptons mix through

$U(1)_{L_\mu-L_\tau}$6

while the neutral component of the doublet remains unmixed with mass $U(1)_{L_\mu-L_\tau}$7 (Kawamura et al., 2022). In the $U(1)_{L_\mu-L_\tau}$8 construction, the charged states $U(1)_{L_\mu-L_\tau}$9 and $SU(2)_V$0 mix through

$SU(2)_V$1

and nonzero left- and right-handed mixing angles are required for the simultaneous description of the $SU(2)_V$2-mass shift and $SU(2)_V$3 (Zhou et al., 2022).

The stability mechanism is not always an imposed elementary $SU(2)_V$4. In gauged lepton number, stability follows from an accidental global symmetry of the renormalizable Lagrangian even after $SU(2)_V$5 breaking (Schwaller et al., 2013). In $SU(2)_V$6, the term $SU(2)_V$7 breaks the gauge symmetry to a remnant $SU(2)_V$8 under which $SU(2)_V$9, stabilizing the lighter real scalar component $SU(2)_D$0 (Zhou et al., 2022). In the alternative left-right model, the parity $SU(2)_D$1 forbids all Yukawa mixings of the new vector-like leptons with ordinary leptons, and the neutral component $SU(2)_D$2 cannot decay into Standard Model states (Bouzeraib et al., 10 Mar 2026).

4. Relic-density mechanisms

The relic-density calculation is typically organized around the Boltzmann equation

$SU(2)_D$3

or its yield-based form. This structure appears in singlet-doublet models, portal models, and gauge-extended models alike (Bhattacharya et al., 2015, Kawamura et al., 2022, Bouzeraib et al., 10 Mar 2026).

In the minimal singlet-doublet fermion case, the dominant channels are $SU(2)_D$4-channel $SU(2)_D$5- or $SU(2)_D$6-exchange into $SU(2)_D$7, $SU(2)_D$8, and $SU(2)_D$9, together with co-annihilation involving the heavier neutral state and the charged partner when the mass splitting is sufficiently small (Bhattacharya et al., 2015). Reviews of the framework summarize the usual pattern: without a triplet, viable relic density generally prefers small singlet-doublet mixing, $SU(2)_L$0, and a small mass difference with the charged next-to-lightest state, $SU(2)_L$1 GeV, so that co-annihilation is effective (Bhattacharya et al., 2018).

Several models replace this by resonance control. In the alternative left-right construction with $SU(2)_L$2, the dominant channels are $SU(2)_L$3 via $SU(2)_L$4 and co-annihilations $SU(2)_L$5 via $SU(2)_L$6, with narrow viable strips near $SU(2)_L$7 or $SU(2)_L$8 (Bouzeraib et al., 10 Mar 2026). In gauged lepton number, the leading annihilation proceeds through $SU(2)_L$9-channel $$N = \begin{pmatrix}N^0\ N^- \end{pmatrix},\qquad \chi^0,$$0 exchange into Standard Model leptons, and the relic density is compatible with $$N = \begin{pmatrix}N^0\ N^- \end{pmatrix},\qquad \chi^0,$$1–$$N = \begin{pmatrix}N^0\ N^- \end{pmatrix},\qquad \chi^0,$$2 GeV for $$N = \begin{pmatrix}N^0\ N^- \end{pmatrix},\qquad \chi^0,$$3 TeV and $$N = \begin{pmatrix}N^0\ N^- \end{pmatrix},\qquad \chi^0,$$4–$$N = \begin{pmatrix}N^0\ N^- \end{pmatrix},\qquad \chi^0,$$5 (Schwaller et al., 2013).

Portal models often rely on chirality structure and co-annihilation simultaneously. In the muon-specific scalar portal of Kawamura, Okawa, and Omura, the relic density receives contributions from $$N = \begin{pmatrix}N^0\ N^- \end{pmatrix},\qquad \chi^0,$$6 and from co-annihilation with the lighter charged vector-like lepton $$N = \begin{pmatrix}N^0\ N^- \end{pmatrix},\qquad \chi^0,$$7; using micrOMEGAs, $$N = \begin{pmatrix}N^0\ N^- \end{pmatrix},\qquad \chi^0,$$8 is found for $$N = \begin{pmatrix}N^0\ N^- \end{pmatrix},\qquad \chi^0,$$9–$Z_2$0 with Yukawas $Z_2$1–$Z_2$2 (Kawamura et al., 2022). In the inert Zee model, the lepton portal generates $Z_2$3 through $Z_2$4-channel vector-like charged fermions; the low-mass region $Z_2$5 GeV is recovered when $Z_2$6 and annihilation proceeds through the leptonic portal (Gaviria et al., 2020).

Non-thermal production is a notable outlier rather than the rule. A pure electroweak doublet Dirac fermion with full-strength $Z_2$7 couplings is viable only as a tiny dark-matter subcomponent if its abundance is diluted or regenerated by late modulus decay, with direct-detection bounds requiring $Z_2$8 for $Z_2$9 GeV (Halverson et al., 2014). This case is therefore best viewed as an exceptional corner of the broader vector-like lepton landscape.

5. Direct detection, indirect detection, and precision observables

Direct-detection phenomenology depends sharply on whether the dark matter retains an unsuppressed vector coupling to the $Z$00. In the minimal singlet-doublet model, the $Z$01–$Z$02 coupling is suppressed by $Z$03, and the per-nucleon cross section is approximately

$Z$04

which drives the well-known constraint $Z$05 for $Z$06–$Z$07 GeV from LUX-era bounds (Bhattacharya et al., 2015). Triplet-induced pseudo-Dirac splitting avoids this by kinematically forbidding elastic $Z$08 exchange (Bhattacharya et al., 2017).

Higgs-mediated scattering is the dominant residual channel in many fermionic realizations. In the gauged-lepton-number model, the Higgs-portal cross section can lie just below then-current XENON-100 limits and within reach of LZ/XENONnT for $Z$09 and $Z$10 (Schwaller et al., 2013). In the inert Zee model, the tree-level contribution is controlled by $Z$11, while one-loop vector-like lepton contributions are typically $Z$12 of the Higgs-mediated rate (Gaviria et al., 2020). In the alternative left-right model with $Z$13, the spin-independent cross section arises from $Z$14-channel $Z$15 exchange and current LZ limits exclude much of the parameter space unless the model sits very close to a resonance (Bouzeraib et al., 10 Mar 2026).

Several recent models were constructed explicitly to exploit vector-like lepton loops or mixing in precision observables. In the muon-portal scalar model, one-loop oblique corrections from the new charged and neutral vector-like leptons generate typical values $Z$16–$Z$17, $Z$18–$Z$19, and $Z$20, corresponding to $Z$21–$Z$22 consistent with the CDF anomaly, while the same setup gives chirality-enhanced contributions to $Z$23 if both $Z$24 and $Z$25 are nonzero and the charged-state mixing angles are sizable (Kawamura et al., 2022). The $Z$26 lepton-portal model instead generates the muon $Z$27 through dark gauge boson loops and shifts the $Z$28 mass through tree-level $Z$29–$Z$30 mixing, with a viable region at $Z$31–$Z$32, $Z$33–$Z$34 GeV, and $Z$35–$Z$36 (Kim et al., 2022).

These anomaly-oriented constructions should not be mistaken for generic predictions of the class. The ability to fit the CDF $Z$37 mass or the muon anomaly is model-specific rather than intrinsic to vector-like lepton dark matter.

6. Collider signatures, misconceptions, and model-building directions

Collider phenomenology is dominated by electroweak pair production of charged vector-like leptons and by the compressed nature of many viable spectra. In the minimal singlet-doublet model, the charged partner decays as $Z$38; if $Z$39, the three-body width is suppressed and for $Z$40 the proper decay length can reach the centimeter scale, yielding a displaced-vertex signature (Bhattacharya et al., 2015). The triplet extension preserves this logic and likewise predicts displaced vertices for small mixing and sub-$Z$41 mass splittings (Bhattacharya et al., 2017).

Compressed spectra also weaken conventional searches. In the scalar muon-portal model, light extra leptons with $Z$42 can evade collider bounds when nearly degenerate with the scalar dark matter, because the muons from $Z$43 are soft; the paper summarizes that for $Z$44 GeV the LHC slepton limits collapse to $Z$45–$Z$46 GeV depending on $Z$47 (Kawamura et al., 2022). An analogous statement appears in the $Z$48 model, where $Z$49 searches exclude $Z$50 GeV only for large mass splitting, while $Z$51 GeV remains allowed if $Z$52 GeV (Zhou et al., 2022).

Gauge-extended realizations can instead produce multi-lepton plus missing-energy signatures or heavy-gauge-boson resonance phenomenology. The left-right symmetric model with vector-like leptons predicts $Z$53 and a clean $Z$54 final state, with the International Linear Collider considerably outperforming the LHC for light vector-like leptons (Bahrami et al., 2016). In the $Z$55-symmetric 2HDM with two generations of $Z$56-odd vector-like leptons, high-multiplicity leptonic final states are especially promising: at $Z$57 TeV and $Z$58, the $Z$59 and $Z$60 channels provide the best reach for benchmark points with dark-matter masses up to approximately $Z$61 GeV (Chakraborty et al., 2021).

Three recurrent misconceptions can be stated precisely. First, vector-like lepton dark matter is not synonymous with a pure electroweak doublet WIMP; such a state is generically overconstrained by $Z$62-mediated scattering unless it is pseudo-Dirac, highly subdominant, or embedded in an extended sector (Halverson et al., 2014, Dey et al., 2 Sep 2025). Second, the viable parameter space is not determined by relic density alone: direct detection, electroweak precision data, invisible decays, and collider compression effects are structurally inseparable from the dark-matter calculation (Bhattacharya et al., 2015, Bhattacharya et al., 2018). Third, the class is not phenomenologically uniform. Some realizations are designed around neutrino mass generation, some around lepton-flavor structure, and some around simultaneous explanations of $Z$63 and the 2022 CDF $Z$64-mass result (Bhattacharya et al., 2017, Carone et al., 2018, Kawamura et al., 2022).

Taken together, the literature defines a broad but coherent research program: vector-like leptons supply a renormalizable dark-sector interface whose phenomenology is controlled by symmetry protection, mixing structure, and electroweak-scale spectroscopy. The central model-building problem is not merely to obtain a stable neutral state, but to do so while reconciling relic abundance, suppressed direct detection, and experimentally accessible charged-partner signatures.