WorksPhysics and Astronomy
Probing dark-matter scenarios of Belle II excess in with decays
Abstract
The hypothesis of top-philic dark matter as the origin of the Belle II excess events in decays has direct impact on decays: Both of these decays proceed via loop diagrams~with a top quark emitting some mediator field and the subsequent decay of into a pair of dark-matter fermions . The Belle II results on allowed us to extract the parameters of the mediator with reasonable accuracy, whereas the sensitivity to the fermion mass proved to be rather low. In contrast to decays, the decays exhibit a strong sensitivity to ; therefore, if - upon reducing theoretical and experimental uncertainties - the tension between standard-model predictions for the decays and the most recent NA62 measurements becomes pronounced the latter decays may either enable the extraction of the value of or provide a lower limit on~.
Introduction
In the Standard Model (SM), the missing-energy decay is effected by the decays with a pair of neutrino and antineutrino in the final state. This flavour-changing neutral-current (FCNC) induced process is, in the SM, forbidden at tree level and proceeds via loops (penguins and boxes). For the resulting branching ratios, rather accurate theoretical predictions have been reported [1, 2, 3]:
In 2023, the Belle II collaboration reported [4] a surprising observation significantly exceeding the SM expectation, viz.,
as well as, in addition, the measurement of the differential distribution for the decay . If confirmed, these findings require the contribution of new physics beyond the SM. One of many new-physics scenarios for the enhancement is provided by the interpretation of the Belle II excess events as decays into invisible dark-matter (DM) particles [5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28]. Various specific scenarios have been considered, involving, for instance, a dark photon or decays into a DM fermion, , and antifermion, , via some mediator field, in the following generically labelled . Taking into account the additional DM contribution, the decay width becomes
In the latter context, in our respective previous analyses [6, 14, 19, 20, 28] we studied decays within two specific top-philic DM scenarios defined by assuming the mediator field coupling both to the top quark and to a DM fermion-antifermion pair , for either a scalar mediator field (labeled S-scenario) or a vector mediator field (denoted V-scenario). For both scenarios, a good description of the Belle II excess data could be demonstrated and all relevant DM parameters (that is, the mediator mass , the mediator decay width, and the couplings of the mediator to the top quark, , and to the DM fermion, ) have been extracted.
Noteworthy, within top-philic scenarios the same mechanism contributes to any FCNC and transition, in particular, to the missing-energy decay , with the evident consequence for the latter’s decay width that
Both SM amplitude and corresponding differential distribution are known quite well (cf. e.g., the classical work [29] and its recent update [30]). For the additional contribution , upon adopting the top-philic DM parameter values extracted from the successful description of the Belle II excess events in the decays, one readily obtains parameter-free predictions. We calculate the differential distribution of the decay within both S- and V-scenarios and study further constraints on the DM sector, which may be obtained from a comparison of our theoretical predictions with the NA62 experimental results for decays.
Standard-Model and Dark-Matter Contributions to the Decay
Decays in the SM
The effective Hamiltonian describing the transition in the SM has been available for thirty years [29]:
The corresponding differential decay rate, with respect to the square of the difference of the involved meson momenta, , reads [31, 32]
where . The analytic expressions for and can be found in Sec. XI of [29]. We use the numerical values of all the involved parameters from the comparatively recent analysis [30], which yield
corresponding to
The uncertainty in this estimate arises primarily from the uncertainty in , since the form factor is known to a very good accuracy from lattice QCD [33]:
The uncertainty in the -dependence of this form factor plays a minor role since the relevant range of is limited, , and the maximal is much less than the mass of the lowest resonance in the -channel, , that is, .
Decays in the DM S- and V-Scenarios
For the decay , we work out implications of the conjecture that the Belle II excess events actually arise from the decay into a pair of DM fermions that couple to the top quark via some mediator field . No interaction of the mediator with other SM particles is implied. In this case, both FCNC decays and proceed via the top–-boson loop depicted in Figure 1. Therefore, the amplitude of the decay is obtained from the amplitude of the decay by obvious replacements of the CKM factors, quark masses, and form factors.
Figure 1. Feynman graphs underlying the FCNC decays (a) and (b) in top-philic DM scenarios. The Wilson coefficients corresponding to Diagram (a) and Diagram (b) are related to each other by the simple replacements of the CKM factors and of the involved quark fields.
S-Scenario
A rather popular model [34, 35], referred to as -scenario, involves the interaction of some DM fermions with the top quark by exchange of a scalar mediator field , governed by the interaction Lagrangian
here is the vacuum expectation value of the SM Higgs field. The emerging effective Lagrangian encoding the FCNC vertex then reads [34, 35]
As depicted in Figure 1(b), the transition proceeds via the mediator field . One of the ingredients of the amplitude of interest, describing the transition and expressed in terms of a form factor , assumes the form
For the differential distribution in the -scenario, upon neglecting compared with , we obtain [19]
with representing the -dependent width of the mediator [14, 19]. This width can be inferred from the imaginary part of the fermion-loop Feynman diagram with scalar vertices (cf. [36, 37]) in the form of
We emphasize that the -quark mass in the denominator of Eq. (14) cancels the factor in , Eq. (12); it thus drops out from the decay rate. The above expressions involve four DM parameters: , , , and . Below, we assume that and, furthermore, that the mediator decays predominantly into the pair.
V-Scenario
A widely discussed top-philic vector-mediator scenario, involving a vector mediator field that couples both to the top quark and to some DM fermions , is defined by the Lagrangian [38, 39, 40]
The coupling of the mediator to the FCNC occurs by means of the penguin loop diagram shown in Figure 1(b). Integrating out the heavy top and fields leads to the effective Lagrangian [41]
The resulting Wilson coefficient , where denotes the mass of the boson in the penguin loop of Figure 1(b), has the form
The associated differential distribution , involving the width of the mediator , reads [19]
Under the assumption that , the -dependent decay width can be read off from the imaginary part of the two-point Feynman diagram with fermions in the loop and vectorial vertices [37]:
These expressions involve the four DM parameters , , , and :
DM CONTRIBUTIONS TO THE OBSERVABLES
For most of the DM parameters of relevance for the present analysis, their values have been extracted in our earlier studies [14, 19, 20] by fitting the Belle II observation [4] of missing-energy excess events within both S- and V-scenarios. The only DM parameter yet to be pinned down is the mediator–quark coupling , or , respectively.
For the S-scenario, the “best” parameter values of -decay origin, corresponding to the global minimum of , are
In addition, the S-scenario entails a simple relation between -decay relevant and -decay relevant parameters:
Recalling the earlier obtained result and inserting it in Eq. (22) yields for its counterpart
For the V-scenario, proceeding in a manner similar to the S-scenario implies, for the -decay relevant parameters,
In the V-scenario, the relationship between -decay governing and -decay governing parameters is still simpler,
and trivially enables us to deduce the value of from the previously extracted parameter :
Figure 2 summarizes the differential distributions for the missing-energy decay as obtained within the SM and the S- and V-scenarios for the DM parameters from, respectively, Eqs. (21) and (23), or (24) and (26). We separately present the theoretically predicted differential distributions in (a) the SM vs the DM S-scenario and (b) the SM vs the DM V-scenario. Moreover, the binned experimentally measured differential distribution from Ref. [44] vs the SM contribution and the additional DM contribution are shown for both the S-scenario (c) and the V-scenario (d).
Figure 2. Glance at missing-energy decay : Confrontation of the differential distributions composed of SM and DM contributions (top row)—with dash-dotted green lines indicating the SM prediction and dash-dotted blue or red lines highlighting the additional contributions caused by the decay—and of the degree of agreement between the binned experimentally measured differential distribution—extracted from Fig. 8(a) of Ref. [44]—with the expectation arising from combining the SM and the additional DM contributions (bottom row), for both the DM S-scenario (left column) and the DM V-scenario (right column).
Various brief comments on our insights gained, and the conclusions inferred, from the outcomes of this investigation, collectively presented in Figure 2, are in order:
The decays proved, as revealed in earlier studies [14, 19, 20], to be not particularly sensitive to the DM fermion mass . For the S-scenario, practically any mass from the range is admissible. For the V-scenario, the mass constraint was more stringent: . In contrast to this, the decays turn out to be very sensitive to the precise value of for the case of light DM fermions, satisfying , for which the decay is kinematically allowed.
For the S-scenario, the maximal DM contribution to the differential distribution corresponds to a vanishing DM fermion mass, . There, it amounts to some 30 – 40% excess, more or less homogeneously smeared over the full allowed -region. This result is well compatible with the recent NA62 measurement of the differential distributions in [42, 43, 44]. Thus, the DM contribution to in the S-scenario is well compatible with the DM parameters deduced earlier from the Belle results for the decay .
For the V-scenario, the emerging overall picture is quite different: There, a pronounced sensitivity of the excess events induced by some DM contribution in is observed. The region GeV is practically excluded by the NA62 measurements, in excellent agreement with the constraints inferred from the Belle II data on . A highly interesting case is the region GeV: There, one will note a strong enhancement of the differential distribution in the region GeV. Optimists may see—or even interpret such behavior as—some indications of an enhancement of the experimental results compared to the SM predictions in this region. If this enhancement is established for greater statistics, an explanation of these findings within the DM V-scenario will be a viable possibility.
Summary
In conclusion, we analyzed the consequences, for the missing-energy decay , of the conjecture of some top-philic DM origin of the Belle II excess events observed for the decay . Within the framework of this conjecture, the branching ratio of the decay under study receives two contributions:
Herein, the DM contribution involves precisely the same DM parameters as those determined previously [19] from the analysis of the decay . Therefore, the decay provides an unambiguous probe of the two top-philic scenarios employed.
The present experimental results reported by NA62 [44] are compatible within uncertainties with the SM predictions. Nevertheless, an optimistic eye may tend to see some excess events in the data. If (ultimately) established at a higher statistics, such excess events might be a useful additional probe of DM (as well as of any other new physics) scenarios.
Our conclusions may be summarized as follows:
The mediator mass was determined from the Belle II data with reasonable accuracy, whereas the DM fermion mass extracted from these data had a larger uncertainty. The decay , however, provides a very promising possibility to probe the DM fermion mass in the region , where the decay is kinematically allowed: The differential distribution of is sensitive to the precise value of , thus potentially providing constraints on .
For the S-scenario, the allowed range of extracted from the Belle II data on essentially tolerates for the mass any value in accordance with the bounds GeV. The maximal DM contribution, of branching ratio , is expected for . Figure 2(a) provides the theoretical differential distributions for both the SM and the DM S-scenario, whereas Figure 2(c) shows the binned differential distribution measured by the NA62 experiment vs the SM and DM S-scenario contributions. The access events are distributed smoothly over the kinematically allowed range . So, any smoothly distributed excess events would speak in favour of the S-scenario with a small value of the DM fermion mass parameter .
For the V-scenario, the allowed range of demanded by the Belle II data on is GeV. However, the region of smaller is likewise not totally excluded. Figure 2(b) provides the theoretical differential distributions for both the SM and the DM V-scenario, and Figure 2(d) presents the binned differential distribution measured by the NA62 experiment vs the SM and DM V-scenario contributions. In this case, a rather particular DM contribution to the decay emerges for GeV: the contribution is peaked in the region GeV and may be sufficiently large. So, if higher-statistics data establish excess events peaked at large , this behavior will give a strong support to the V-scenario with DM fermion mass GeV.
Acknowledgments
The research of A.B. and D.M. was carried out within the scientific program of the National Center for Physics and Mathematics, Section 5 Particle Physics and Cosmology. Stage 2026–2027.
References
- [1]D. Bečirević, G. Piazza, and O. Sumensari, Revisiting \(B \to K^{(*)}\nu\bar{\nu}\) decays in the Standard Model and beyond, Eur. Phys. J. C 83, 252 (2023).
- [2]W. G. Parrott, C. Bouchard, and C. T. H. Davies [HPQCD collaboration], Standard Model predictions for \(B \to K\ell^+\ell^-\), \(B \to K\ell_1^+\ell_2^-\), and \(B \to K\nu\bar{\nu}\) using form factors from \(N_f = 2 + 1 + 1\) lattice QCD, Phys. Rev. D 107, 014511 (2023); 107, 119903(E) (2023).
- [3]L. Allwicher, D. Bečirević, G. Piazza, S. Rosauro-Alcaraz, and O. Sumensari, Understanding the first measurement of \(B(B \to K\nu\nu)\), Phys. Lett. B 848, 138411 (2024).DOI
- [4]I. Adachi et al. (Belle-II Collaboration), Evidence for \(B^+ \to K^+\nu\nu\) decays, Phys. Rev. D 109, 112006 (2024).
- [5]M. Abdughani and Y. Reyimuaji, Constraining light dark matter and mediator with \(B \to K\nu\nu\) data, Phys. Rev. D 110, 055013 (2024).
- [6]A. Berezhnoy and D. Melikhov, \(B \to K^*M_X\) vs \(B \to KM_X\) as a probe of a scalar-mediator dark matter scenario, EPL 145, 14001 (2024).
- [7]A. Datta, D. Marfatia, and L. Mukherjee, \(B \to K\nu\bar{\nu}\), MiniBooNE and muon \(g-2\) anomalies from a dark sector, Phys. Rev. D 109, L031701 (2024).
- [8]L. Calibbi, T. Li, L. Mukherjee, and M. A. Schmidt, Is Dark Matter the origin of the \(B \to K\nu\nu\) excess at Belle II?, Phys. Rev. D 112, 075020 (2025). arXiv:2502.04900 [hep-ph].arxiv.org/abs/2502.04900
- [9]S.-Y. Ho, J. Kim, and P. Ko, Recent \(B \to K\nu\nu\) Excess and Muon \(g-2\) Illuminating Light Dark Sector with Higgs Portal, Phys. Rev. D 111, 055029 (2025).arxiv.org/abs/2401.10112
- [10]X.-G. He, X.-D. Ma, M. A. Schmidt, G. Valencia, and R. R. Volkas, Scalar dark matter explanation of the excess in the Belle II \(B \to K+\) invisible measurement, JHEP 07, 168 (2024).
- [11]B.-F. Hou, X.-Q. Li, M. Shen, Y.-D. Yang, and X.-B. Yuan, Deciphering the Belle II data on \(B \to K\nu\nu\) decay in the (dark) SMEFT with minimal flavour violation, JHEP 06, 172 (2024).
- [12]E. Gabrielli, L. Marzola, K. Muursepp, and M. Raidal, Explaining the B → Kνν excess via a massless dark photon, Eur. Phys. J. C 84, 460 (2024).
- [13]X. G. He, X. D. Ma, J. Tandean and G. Valencia, B → K+invisible, dark matter, and CP violation in hyperon decays, JHEP 07, 078 (2025).
- [14]A. Berezhnoy, W. Lucha, and D. Melikhov, Analysis of q2-distribution for B → KMX and B → K*MX in a scalar-mediator dark matter scenario, Phys. Rev. D 111, 075035 (2025).
- [15]L. Kloy and S. Nandi, Flavour and Electroweak Precision Constraints on a Simplified Dark Matter Model with a Light Spin-0 Mediator, arXiv:2503.15609 [hep-ph].
- [16]T. M. Aliev, A. Elpe, I. Turan and L. Selbuz, Explaining Belle Data on B → K(*)νν̄ Decays via Dark Photon Resonances, Phys. Rev. D 112, 015025 (2025).
- [17]K. Ding, Y. Li, X. Liu, Y. Liu, C. T. Lu and B. Zhu, Resonant ALP-Portal Dark Matter Annihilation as a Solution to the B± → K±νν̄ Excess, Phys. Rev. D 112, 115034 (2025).DOI
- [18]X. G. He, X. D. Ma, J. Tandean and G. Valencia, Light dark-matter window in K+ → π+ℓ̸, Phys. Rev. D 112, 055025 (2025).
- [19]A. Berezhnoy, W. Lucha and D. Melikhov, Probing vector- versus scalar-mediator dark-matter scenarios in B → (K, K*)MX decays, Eur. Phys. J. Plus 141, 495 (2026).DOI
- [20]A. Berezhnoy, W. Lucha, and D. Melikhov, Scrutinizing Dark-Matter Scenarios with B → (K, K*)νν̄ Decays, Universe 11 (2025) 12, 385.DOI
- [21]Z. K. Liu, Y. Li, B. F. Hou and Q. Chang, Search for light dark matter in rare meson decays, Phys. Rev. D 113, 035030 (2026).
- [22]J. Kim and P. Ko, B+ → K+νν̄ excess and DM semi-annihilation, Phys. Rev. D 113, 075023 (2026).DOI
- [23]F. Mescia, S. Okawa, J. Swallow and C. Toni, Dark matter emission at Belle II and NA62 in the minimal flavor violation framework, Phys. Rev. D 113, 095014 (2026).
- [24]P. D. Bolton, J. F. Kamenik and M. Novoa-Brunet, Dark light shining on B → K(*)Emiss, Phys. Rev. D 113, 115001 (2026).DOI
- [25]S. Jahedi, J. H. Liang, Y. Liao, X. D. Ma and Y. Uchida, Systematic study of lepton-flavor-violating dark matter interactions via indirect detection in effective field theories, Phys. Rev. D 114, 035007 (2026).
- [26]L. Gärtner, N. Krug, T. Kuhr, M. A. Schmidt, S. Stefkova and B. Yabsley, Constraints on invisible B+ → K+X decays from the Belle II B+ → K+νν̄ measurement, Phys. Rev. D 114, no.3, 032003 (2026).
- [27]S. Y. Ho, J. Kim and P. Ko, A Unified Dark Matter Explanation for B+ → K+νν̄ and the Super-Kamiokande Antineutrino Excess, [arXiv:2607.03206 [hep-ph]].
- [28]A. Berezhnoy, W. Lucha and D. Melikhov, B(s) → V MX decays as probes of dark-matter scenarios of Belle II enhancement in B → KMX decays, [arXiv:2609.13069 [hep-ph]].arxiv.org/abs/2609.13069
- [29]G. Buchalla, A. J. Buras and M. E. Lautenbacher, Weak Decays beyond Leading Logarithms, Rev. Mod. Phys. 68, 1125 (1996).DOI
- [30]J. Brod, M. Gorbahn and E. Stamou, Updated Standard Model Prediction for K → πνν̄ and εK, PoS BEAUTY2020, 056 (2021).
- [31]P. Colangelo, F. De Fazio, P. Santorelli and E. Scrimieri, Rare B → K(*) neutrino anti-neutrino decays at B factories, Phys. Lett. B 395 339 (1997).
- [32]D. Melikhov, N. Nikitin and S. Simula, Rare exclusive semileptonic b → s transitions in the standard model, Phys. Rev. D 57, 6814 (1998).
- [33]Y. Aoki et al. [Flavour Lattice Averaging Group (FLAG)], FLAG review 2024, Phys. Rev. D 113, 014508 (2026).
- [34]B. Batell, M. Pospelov, and A. Ritz, Multi-lepton Signatures of a Hidden Sector in Rare B Decays, Phys. Rev. D 83, 054005 (2011).DOI
- [35]K. Schmidt-Hoberg, F. Staub, and M. W. Winkler, Constraints on light mediators: confronting dark matter searches with B physics, Phys. Lett. B 727, 506 (2013).
- [36]G. J. Gounaris and J. J. Sakurai, Finite width corrections to the vector meson dominance prediction for ρ → e+e−, Phys. Rev. Lett. 21, 244 (1968).DOI
- [37]D. Melikhov, O. Nachtmann, V. Nikonov, and T. Paulos, Masses and couplings of vector mesons from the pion electromagnetic weak, and pi gamma transition form-factors, Eur. Phys. J. C 34, 345 (2004).
- [38]P. Langacker, The Physics of Heavy Z′ Gauge Bosons, Rev. Mod. Phys. 81, 1199 (2009).DOI
- [39]P. Cox, A. D. Medina, T. S. Ray and A. Spray, Novel collider and dark matter phenomenology of a top-philic Z′, JHEP 06, 10 (2016).DOI
- [40]Y. Hu, Y. Liu and Y. Liu, A study on vector mediator top-philic dark matter, Commun. Theor. Phys. 76, 085201 (2024).DOI
- [41]T. Inami and C. S. Lim, Effects of Superheavy Quarks and Leptons in Low-Energy Weak Processes KL → μμ̄, K+ → π+νν̄ and K0 ↔ K̄0, Prog. Theor. Phys. 65, 297 (1981) [erratum: Prog. Theor. Phys. 65, 1772 (1981)].
- [42]E. Cortina Gil et al. [NA62], Observation of the K+ → π+νν̄ decay and measurement of its branching ratio, JHEP 02, 191 (2025).
- [43]X. Chang [NA62], New measurement of K+ → π+νν̄ branching ratio at the NA62 experiment, [arXiv:2604.12649 [hep-ex]].arxiv.org/abs/2604.12649
- [44]S. Alibocus et al. [NA62], Measurement of the branching ratio of the K+ → π+νν̄ decay, [arXiv:2607.16413 [hep-ex]].

