--- title: 'Minimal Dark Matter: Models & Phenomenology' url: https://www.emergentmind.com/topics/minimal-dark-matter type: topic --- # Minimal Dark Matter: Models & Phenomenology Searching arXiv for recent papers on Minimal Dark Matter and related developments. Minimal Dark Matter (MDM) is a class of dark-matter models in which the Standard Model is extended by a single electroweak multiplet, or by the minimally required set of such multiplets in close variants, with interactions fixed predominantly by gauge invariance and with stability arising accidentally or from closely related residual symmetries rather than from an ad hoc dark parity. In its canonical form, MDM adds one \(SU(2)_L\) multiplet with specified hypercharge \(Y\), no new ad hoc interactions, and typically only one continuous parameter, the dark-matter mass \(M\); thermal freeze-out then determines the preferred mass scale from the observed relic abundance [2508.02778]. The framework is most sharply realized by fermionic \(Y=0\) electroweak multiplets, especially the Majorana quintuplet \((1,5)_0\), which is the smallest accidentally stable real representation in the standard construction [2412.19660][2507.17607]. ## 1. Canonical definition and field content Minimal Dark Matter is defined by adding a single \(SU(2)_L\) multiplet to the Standard Model, with no imposed \(\mathbb Z_2\) or analogous stabilizing symmetry and no new interactions beyond those permitted by gauge invariance and renormalizability [2412.19660][1711.08619]. In the fermionic case emphasized in recent precision work, the added field is an \(SU(2)_L\) \(n\)-plet with hypercharge \(Y\), interacting only through the electroweak gauge sector; the representative choices discussed explicitly are a doublet with \(n=2, Y=1/2\), a triplet with \(n=3, Y=0\), and a quintuplet with \(n=5, Y=0\) [2508.02778]. The canonical quintuplet candidate is a Majorana fermion \[ \chi \sim (1,5)_0 \] under \(SU(3)_c\times SU(2)_L\times U(1)_Y\), with components \[ \chi=\bigl(\chi^{++},\chi^+,\chi^0,\chi^-,\chi^{--}\bigr), \] and a renormalizable Lagrangian \[ \mathcal{L}=\frac12\,\overline{\chi}(iD\hspace{-0.28cm}/\hspace{0.05cm}-M_5)\chi . \] Because of its \(SU(2)_L\) and \(U(1)_Y\) quantum numbers, no renormalizable operator couples a single \(\chi\) to Standard-Model fields in a way that induces decay, so an accidental \(\mathbb Z_2\) emerges at the renormalizable level [2412.19660]. This accidental stability is the defining structural feature of the original MDM idea. By contrast, lower-dimensional candidates such as the triplet and doublet generally require an imposed stabilizing \(\mathbb Z_2\) in phenomenological treatments, because otherwise allowed operators spoil cosmological stability [2508.02778]. The neutral state must also evade direct-detection bounds from tree-level \(Z\)-exchange. This is one reason \(Y=0\) real multiplets occupy a privileged role in the literature. In the early MDM classification, the viable automatically stable options were narrowed to a fermionic quintuplet with \(Y=0\) and a scalar septuplet with \(Y=0\), with the fermionic quintuplet singled out as the most predictive case [0808.3867]. Later work revisited the status of scalar multiplets and emphasized that perturbativity and higher-dimensional operators substantially restrict them, reinforcing the centrality of the fermionic quintuplet in the strict accidental-stability sense [1801.05594]. Radiative electroweak corrections split the charged and neutral components. For the quintuplet, the charged–neutral splitting is \[ \Delta m \equiv m_{\chi^\pm}-m_{\chi^0}\simeq 166~\mathrm{MeV}, \] so the neutral component is the lightest state and therefore the dark-matter candidate [2412.19660]. Closely related analyses report the same \(\mathcal O(100\,\mathrm{MeV})\) splitting pattern for triplet-like realizations and for higher charged components, with the \(Q^2\) scaling of the splitting in real multiplets entering both cosmology and collider phenomenology [1507.01584][2507.15934]. ## 2. Thermal relic mechanism and predicted mass scales In canonical MDM, the relic abundance is fixed by thermal freeze-out. The dark multiplet \({\cal X}\) is initially in equilibrium with the Standard Model; when the temperature drops below \(M\), annihilations freeze out and the surviving abundance is set by the thermally averaged effective annihilation cross section [2508.02778][2412.19660]. The standard parametric relation is \[ \Omega_{\rm DM}h^2 \propto \frac{1}{\langle \sigma v\rangle}, \] so once the gauge representation is chosen and the couplings are fixed by the Standard Model, the observed relic density selects a specific mass. For a fermionic \(SU(2)_L\) \(n\)-plet with hypercharge \(Y\), the short-distance \(s\)-wave annihilation cross section quoted in recent precision work is \[ \sigma v_{\rm rel} = \frac{g_2^4 (n^2-1)(2n^2+19+18\alpha_3/\pi) + 4 Y^2 g_Y^4 (41+8Y^2 + 22 \alpha_3/\pi) +16g_2^2 g_Y^2 Y^2 (n^2-1)}{256\pi\, M^2\, g_{\cal X}}, \] with \(g_{\cal X}=2n\) for Majorana \(Y=0\) multiplets and \(g_{\cal X}=4n\) for Dirac \(Y\neq0\) multiplets [2508.02778]. This expression already displays the central scaling \(\sigma v\propto 1/M^2\), which underlies the predictive character of the framework. The thermal mass values depend sensitively on non-perturbative electroweak effects. In the modern literature summarized in the supplied material, representative thermal masses are as follows. | Candidate | Thermal mass | Context | |---|---:|---| | Doublet \((n=2,Y=1/2)\) | \(1.08\pm0.02~\mathrm{TeV}\) | cosmology (previous) [2508.02778] | | Triplet \((1,3)_0\) | \(2.86\pm0.04~\mathrm{TeV}\) | cosmology (previous) [2508.02778] | | Quintuplet \((1,5)_0\) | \(13.7^{+0.6}_{-0.3}~\mathrm{TeV}\) | cosmology (previous) [2508.02778] | | Quintuplet \((1,5)_0\) | \(14~\mathrm{TeV}\) | \(SU(5)\) embedding input [2412.19660] | The canonical fermionic quintuplet was historically associated with a mass near \(9.6\pm0.2\) TeV when Sommerfeld effects were included in earlier analyses [0808.3867], whereas later treatments that incorporate bound-state formation and more complete electroweak dynamics shifted the preferred mass upward to about \(14\) TeV [2412.19660]. A plausible implication is that the numerical value of the “canonical” MDM mass is theory-systematics dependent at the several-TeV level once higher-order electroweak effects are included; the supplied recent indirect-detection analyses therefore use the updated thermal window around \(13.7\) TeV for the real quintuplet [2507.17607]. For the triplet and quintuplet, Sommerfeld enhancement and bound-state formation are not perturbative corrections but leading dynamical ingredients. The \(SU(2)_L\) long-range potentials are of Coulombic form \[ V=-\lambda \alpha_2/r, \] with \(\lambda=(1+I^2-2n^2)/8\) in the group-theory basis used in the precision relic-density analysis [2508.02778]. These effects can be \(\mathcal O(10)\) in the annihilation rate and dominate over subleading loop corrections. ## 3. Stability mechanisms and model variations The original MDM concept is accidental stability: the representation is chosen so that no renormalizable decay operator exists. The quintuplet \((1,5)_0\) is the prototypical realization [2412.19660]. However, a substantial literature generalizes the stability mechanism while keeping the electroweak-multiplet logic. In the minimal local \(U(1)_{B-L}\) extension, an unbroken residual matter parity \[ P_M=(-1)^{3(B-L)} \] survives after symmetry breaking by a field with even \(3(B-L)\), and combined with spin one may define \[ V_p = (-1)^{3(B-L)+2s}. \] This gauged remnant stabilizes a much broader range of electroweak multiplets, including candidates that would be unstable in the Standard Model alone [1801.05594]. The construction is especially important for hypercharged scalar multiplets, where a small CP-even/odd mass splitting can realize inelastic dark matter and evade tree-level \(Z\)-mediated direct-detection bounds. Two mechanisms are described: dimension-5 operators effective for very high \(B-L\) breaking scale, and mixing with hypercharge-zero fields effective for lower \(B-L\) breaking scale [1801.05594]. Left–right symmetric theories offer a related but distinct mechanism. In \(SU(2)_L\times SU(2)_R\times U(1)_{B-L}\) models, the breaking of \(U(1)_{B-L}\) can leave a residual \(\mathbb Z_2^{B-L}\), while high-dimensional \(SU(2)_{L,R}\) multiplets may also be accidentally stable in the original MDM sense [1507.01584]. The specific fermion multiplets \[ \phi_L\sim(2n+1,1,0),\qquad \phi_R\sim(1,2n+1,0) \] generate a two-component dark sector whose left-handed part behaves like standard MDM and whose right-handed part is governed by the additional heavy gauge bosons \(W_R\) and \(Z_R\) [1507.01584]. This is not “minimal” in the original single-multiplet sense, but it preserves the central idea that stability need not come from an imposed dark parity. A different extension allows Higgs Yukawa couplings between electroweak multiplets. In “Minimal Dark Matter coupled to the Higgs,” a Majorana multiplet with \(Y=0\) and a Dirac multiplet with \(Y=\pm 1/2\) mix through the Standard-Model Higgs doublet [1711.08619]. This class includes \(1_M2_D\), \(3_M2_D\), \(3_M4_D\), and \(5_M4_D\) models. Here an explicit dark \(Z_2\) must be imposed, because the original accidental stability of the pure quintuplet is lost once extra multiplets are added [1711.08619]. The resulting models are no longer MDM in the strictest sense, but they form a controlled generalization in which coannihilation structure, direct detection, and mass splittings become tunable. At the opposite conceptual extreme, “Minimal Proton-Mass Dark Matter” uses a single complex scalar singlet carrying baryon and lepton number, with no new exact stabilizing symmetry; stability is kinematic and tied to the narrow window \[ m_p-m_e \le m_\phi \le m_p+m_e. \] This framework is explicitly contrasted with canonical electroweak-multiplet MDM and should be regarded as a distinct use of the phrase “minimal dark matter” rather than a continuation of the Cirelli-type program [2606.20792]. This suggests that the terminology has broadened, while the electroweak-multiplet construction remains the standard meaning in the MDM literature. ## 4. Precision annihilation dynamics The annihilation phenomenology of fermionic MDM is dominated by electroweak gauge channels, but annihilation into quarks and Higgs doublets contributes non-negligibly near freeze-out [2508.02778]. Recent precision work has focused on QCD corrections to the quark final states. For fermionic MDM, the one-loop QCD corrections include virtual gluon vertex corrections and real gluon emission in \[ {\rm DM}(p_1)\,\overline{\rm DM}(p_2)\to q(q_1)\,\bar q(q_2)\,g(q_3), \] computed in Feynman gauge with dimensional regularization in \(d=4-2\epsilon\), massless quarks, and \(\overline{\rm MS}\) renormalization at \(\bar\mu\sim M\) [2508.02778]. The real-emission contribution is \[ \sigma v_{\rm rel} |_{\rm real}= \sigma v_{\rm rel}|_{\rm tree} \left[ 1 + \frac{\alpha_3}{\pi} \left( \frac{4}{3\epsilon^2} +\frac{2-4\ell/3}{\epsilon}+\frac{19}{3}-\frac{7\pi^2}{9}-2\ell+\frac23\ell^2\right)\right], \] and the virtual correction is \[ \sigma v_{\rm rel} |_{\rm virtual}= \sigma v_{\rm rel}|_{\rm tree} \left[ 1 - \frac{\alpha_3}{\pi} \left( \frac{4}{3\epsilon^2} +\frac{2-4\ell/3}{\epsilon}+\frac{16}{3}-\frac{7\pi^2}{9}-2\ell+\frac23\ell^2\right)\right], \] with \(\ell=\ln(4M^2/\bar\mu^2)\). Infrared divergences cancel in the sum, giving the universal finite correction \[ \sigma v_{\rm rel} |_{\rm QCD} = \sigma v_{\rm rel}|_{\rm tree} \Big[1 + \frac{\alpha_3}{\pi} \Big]. \] This correction is directly analogous to classic results for \(e^+e^-\to q\bar q\) and \(Z\to q\bar q\) [2508.02778]. Its impact on the total MDM annihilation rate depends on representation, because the weight of quark final states decreases with increasing \(n\). The net enhancement of the full annihilation rate is about \(+2\%\) for the doublet, \(+1.3\%\) for the triplet, and \(+0.5\%\) for the quintuplet [2508.02778]. Since \(M^2\propto \sigma\) at fixed relic density, the corresponding mass shifts are about \(+1\%\), \(+0.6\%\), and \(+0.25\%\), respectively [2508.02778]. The paper explicitly emphasizes that electroweak loop corrections beyond Sommerfeld and bound-state physics can be of comparable or larger size, so a fully consistent NLO electroweak treatment remains outstanding. Indirect detection calculations for the real quintuplet have recently been updated to include NLO electroweak corrections to the non-relativistic potential and NLL resummation for the endpoint spectrum [2507.17607]. In the modern coupled-channel treatment, the neutral \(Q=0\), \(S=0\) two-body basis is \[ \{\chi^{++}\chi^{--},\ \chi^+\chi^-,\ \chi^0\chi^0\}, \] with the potential matrix \[ {\cal V}(r)= \begin{pmatrix} -4A+8\delta_0 & 2B & 0 \\ 2B & -A+2\delta_0 & 3\sqrt{2}B \\ 0 & 3\sqrt{2}B & 0 \end{pmatrix}, \] where \(A(r)\) and \(B(r)\) include Coulomb, Yukawa, and NLO potential terms [2507.17607]. The resulting annihilation dynamics generate both the familiar high-energy line/endpoint features near \(E_\gamma\simeq M_\chi\) and a low-energy structure from bound-state formation at energies of order the binding energy. A key result of that analysis is that, in the Milky Way halo, bound-state formation dominates the gamma-ray flux near \(100\) GeV for the thermal quintuplet. For \(M_\chi=13.7\) TeV, the dominant \(^{1}(1s)_3\) bound state has binding energy \[ E_b=72.94~\text{GeV}, \] and over the thermal window \[ \frac{E_b}{72.94~\text{GeV}} \approx -0.1~\text{GeV} + 1.1~\text{GeV}\left(\frac{M_\chi}{13.7~\text{TeV}}\right) \] is only weakly mass dependent [2507.17607]. This low-energy feature is central to current Fermi-LAT exclusions of the lower thermal mass window. ## 5. Phenomenology: indirect detection, direct detection, and colliders Indirect detection is the most constraining probe of canonical fermionic MDM. For the real quintuplet, updated gamma-ray calculations show that Fermi-LAT diffuse emission data strongly disfavor the lower edge of the thermal mass window, even under conservative assumptions about the inner Milky Way profile [2507.17607]. In that analysis, the thermal mass window is \[ M_\chi = 13.7^{+0.6}_{-0.3}\ \text{TeV}, \] and several hundred hours of CTAO observations of northern dwarf spheroidals are projected to test the central value [2507.17607]. A separate Fermi-based analysis argues more strongly that minimal fermionic \(Y=0\) dark matter making up \(100\%\) of DM is excluded for all \(n<9\) under the standard cosmological history [2507.15934]. Using 14 years of Fermi inner-Galaxy data between 30 GeV and 2 TeV and accounting for continuum photons from annihilation and bound-state formation, it finds 95% upper limits on the signal normalization parameter \(\mu\) below unity for the thermal wino, quintuplet, and \(7\)-plet, even for the conservative FIRE-2 halo “Thelma” profile: \[ \mu^{95}_{\rm Thelma}(3_0)=0.12,\qquad \mu^{95}_{\rm Thelma}(5_0)=0.41,\qquad \mu^{95}_{\rm Thelma}(7_0)=0.80, \] where \(\mu=1\) corresponds to the nominal thermal MDM prediction [2507.15934]. The \(9\)-plet remains allowed in that study, with \(\mu^{95}_{\rm Thelma}=2.1\) [2507.15934]. The thermal wino triplet receives especially detailed treatment. Its thermal relic mass is \[ m_\chi = 2.86\pm 0.01~\text{TeV}, \] and the same Fermi analysis concludes that it is excluded even allowing for cored Milky Way profiles up to about \(6.7\) kpc at \(\rho_\odot=0.38~\text{GeV/cm}^3\), or about \(3.7\) kpc even if \(\rho_\odot\) is reduced to \(0.2~\text{GeV/cm}^3\) [2507.15934]. This conclusion is stronger than older HESS line-based exclusions precisely because the continuum signal persists at large angles where a core suppresses the line signal less efficiently. A controversy remains, however: the alternate updated indirect-detection study of the quintuplet emphasizes that its lower thermal mass range is strongly disfavored but that the central mass value should still be testable by forthcoming CTAO observations rather than already decisively excluded [2507.17607]. This suggests a genuine methodological tension between analyses, driven by spectral treatment, region-of-interest choices, and assumptions about diffuse backgrounds and halo profiles. Direct detection in canonical \(Y=0\) MDM is loop-induced and therefore typically weak relative to indirect constraints. In the \(SU(5)\) quintuplet embedding, the quoted spin-independent prediction is \[ \sigma_\text{SI}^\text{theory}\sim 1.0\times 10^{-46}\ \text{cm}^2 \] for \(M_5=14\) TeV, compared with an LZ bound \[ \sigma_\text{SI}^\text{LZ}\lesssim 4.2\times 10^{-46}\ \text{cm}^2 \] at similar mass, placing the model just below current sensitivity [2412.19660]. In Higgs-coupled MDM generalizations, by contrast, tree-level Higgs exchange can raise the direct-detection cross section into the \(10^{-46}\)–\(10^{-45}\ \mathrm{cm}^2\) regime, with blind-spot cancellations when \(y_1\simeq -y_2\) [1711.08619]. Collider signatures depend strongly on the realization. Pure MDM multiplets produce disappearing tracks or long-lived charged states when the charged–neutral splitting is \(\mathcal O(100\,\mathrm{MeV})\). In the left–right symmetric case, the right-handed charged states can become lighter than the neutral component for \(M\gtrsim 0.9\,M_{W_R}\), excluding that region because it would give charged dark matter [1507.01584]. In asymmetry-based supersymmetric triplet models, the charged fermion lifetime from \[ \Sigma^\pm\to \Sigma^0\pi^\pm \] is characterized by \[ \Gamma^{-1}_{\pi^\pm}\approx 106~\text{cm} \] for \(T=1\), while scalar partners can be effectively detector-stable because their decays are controlled by \(\xi\sim 10^{-12}\) [1102.3455]. In the non-supersymmetric \(SU(5)\) quintuplet embedding, the key collider signatures arise instead from the required colored sextet fermions, with a conservative reinterpretation of ATLAS long-lived \(R\)-hadron searches giving \[ M_6\gtrsim 1.58~\mathrm{TeV}, \] still compatible with the unification-preferred range [2412.19660]. ## 6. Embeddings, ultraviolet structure, and naturalness The canonical quintuplet can be embedded into non-supersymmetric \(SU(5)\) by placing it in a fermionic \(\mathbf{200}_F\), whose decomposition contains \((\mathbf1,\mathbf5)_0\) [2412.19660]. In that construction, Standard Model plus the quintuplet alone does not unify the gauge couplings; two pairs of colored sextet fermions \[ (\mathbf6,\mathbf1)_{-2/3}\oplus(\overline{\mathbf6},\mathbf1)_{2/3} \] at \(\mathcal O(1\!-\!10)\) TeV are required [2412.19660]. Optimal unification occurs near \[ \sqrt{M_6\tilde M_6}\approx 4.3~\mathrm{TeV}, \] with a representative point \[ M_6=2~\mathrm{TeV},\qquad \tilde M_6=9.7~\mathrm{TeV}, \] giving \[ M_U\approx 2.1\times 10^{18}~\mathrm{GeV},\qquad \alpha_U^{-1}\approx 9.0 \] and a rough proton lifetime estimate \[ \tau_p\approx 3.9\times 10^{39}~\text{years} \] [2412.19660]. The unification scale is thus near the reduced Planck scale, and the sextets become metastable because their decays are suppressed by the unification scale. A different ultraviolet concern is the little hierarchy problem induced by a heavy electroweak quintuplet. In “Natural minimal dark matter,” the dark multiplet, Higgs sector, and weak gauge sector are supersymmetrized while the rest of the Standard Model remains non-supersymmetric [1510.03861]. In non-supersymmetric MDM, the quintuplet mass contributes to the Higgs mass parameter at two loops; in the partially supersymmetric construction these \(M^2\)-enhanced two-loop corrections cancel, pushing the first non-vanishing contribution to three loops: \[ \delta m_h^2 \sim \frac{g^6}{(4\pi)^6} M^2. \] This yields a naturalness bound of about \(15\) TeV for the fermionic quintuplet, larger than the \(7\) TeV mass required in that model’s relic-density analysis, where both scalar and fermion quintuplets contribute and Sommerfeld enhancement is included [1510.03861]. This suggests that partial supersymmetrization can reconcile MDM-scale dark matter with Higgs naturalness without fully supersymmetrizing the Standard Model. The ultraviolet behavior of the weak coupling itself has motivated another line of work: minimal asymptotically safe dark matter. There one introduces \(N_F\) copies of a fermionic triplet or quintuplet with common mass \(M_{\rm DM}\) and an \(\mathrm O(N_F)\) flavor symmetry [1905.04227]. The large-\(N_F\) resummed beta function for \(\alpha_2\) develops a non-trivial fixed point, \[ \alpha_2^* \simeq \frac{6\pi}{T(R_f)N_F}, \] avoiding a Landau pole [1905.04227]. The relic-density condition scales as \(\Omega_{\rm DM}h^2\propto N_F M_{\rm DM}^2\), so the required mass drops like \(1/\sqrt{N_F}\). For triplets, viable models remain in the range \[ 3\le N_F\le 20,\qquad 620~\text{GeV}