*mps*

Atmospheric neutrino up-scattering explanation of LZ 2026 excess

Abstract

The recent observation of an isolated nuclear recoil at $248\pm 23\pm 23$ keV energy by LUX-ZEPLIN (LZ) experiment has motivated the community to look for a new physics explanation, as the Standard model background estimation fails to accomodate that. Most of the existing literature hitherto considers a galactic halo dark matter with a heavier partner. In this work, we traverse the alternate route of atmospheric neutrino ($\nu$) up-scattering, thus producing a massive beyond standard model (BSM) particle $\chi$. The kinematic requirement of such a scattering poses a cut off in the lower recoil energies providing an explanation of the unique isolated event at such a higher recoil energy. Such up-scattering with the nucleons ($\mathcal{N}$), $\nu\mathcal{N}\to\chi\mathcal{N}$ can be naturally realized in sterile neutrino models, though we keep our analysis generic without specifying $\chi$. We identify the region of parameter space that can produce such an isolated event assuming a scalar mediator with mass $m_{\phi}$ and coupling $y_{\chi,q}$. For example, with $m_{\chi}\sim 1$ GeV, and $\sqrt{y_{\chi}y_{q}}/m_{\phi}=2\times 10^{-2}$ GeV^-1 can satisfy such an excess of events while remaining allowed by other existing constraints as well.

Despite the remarkable success of the Standard model (SM) of particle physics, many natural phenomena are yet to be resolved [1]. Such shortcomings of SM include dark matter [2], neutrino mass [3], the strong CP problem etc. For these reasons, the quest for beyond standard model (BSM) physics persists for almost half a century. The ton-scale dark matter (DM) detectors [4, 5, 6, 7], though originally proposed to identify DM, have served as a powerful strategy to exclude or constrain many such low energy BSM scenarios and strengthened our understanding of the SM neutrino sector. The recent observation by LUX-ZEPLIN (LZ) experiment has suggested an isolated nuclear recoil event around $248\pm 23\pm 23$ keV recoil energy ($T_{\mathcal{N}}$) [8]. Interestingly, the origin of such an excess event remains unresolved from our known background arising from the SM interactions. The most dominant background for direct detection experiments is the solar neutrinos, which are accumulated below $T_{\mathcal{N}}\lesssim 10$ keV [8]. And above $T_{\mathcal{N}}\gtrsim 100$ keV, events due to the atmospheric neutrinos, mostly arising from pion decay, are too suppressed. Since the standard weak interactions can not explain such an event at high recoil energy, this strongly points toward BSM interactions.

The characteristic feature of this unknown event also make its quite distinct from the events arising from BSM scenarios with a continuous falling rate with recoil energy. To explain the absence of any such BSM excess events in low recoil energy inelastic dark matter (DM) models are proposed [9, 10, 11, 12]. In such scenarios, the lighter DM component scatters with a nucleon to produce a heavier dark state. This poses a lower cutoff on the recoil energy and thus can well explain such a feature. On the contrary, in this paper we propose an alternative explanation of such an excess by BSM interactions of neutrinos. The existence of neutrino masses is indeed a strong indication of BSM physics, and many well-motivated BSM scenarios propose dark particles to explain the neutrino mass issue. Neutrinos undergoing elastic scattering (e.g. $\nu\mathcal{N}\to\nu\mathcal{N}$) through any BSM interactions can in principle induce excess events [13]. However, such scatterings exhibit higher recoil events in the lower $T_{\mathcal{N}}$ and the required interaction rate to produce one event in high $T_{\mathcal{N}}\sim 250$ keV is ruled out from low energy ($T_{\mathcal{N}}\lesssim 30$ keV) observations [13, 14].

Figure 1: Nuclear recoil events spectrum at LZ for the benchmark $m_{\chi}=2$ GeV and $\sqrt{y_{\chi}y_{q}}/m_{\phi}=0.25$ GeV^-1. The blue curve shows the total SM neutrino background (solar, DSNB and atmospheric). The black curve shows the BSM up-scattering contribution $\nu\mathcal{N}\to\chi\mathcal{N}$. The magenta star marks the LZ event observed at $T_{\mathcal{N}}=248$ keV [8]. Grey shaded bands indicate recoil energies outside the signal region of interest.

In this work we consider, the upscattering of $\nu$ to a heavier dark particle $\chi$ with mass $m_{\chi}$ (e.g. $\nu\mathcal{N}\to\chi\mathcal{N}$). Such scenario is well explored for neutrino scattering events [15, 16, 17] and also in light DM absorption searches [18, 19, 20] when $\chi$ can dposit its mass to recoil energy. However, for our scenario $\chi$ can be any generic BSM particle, not necessarily DM. The striking feature of such $\nu$ up-scattering to massive state is that, only neutrinos with energy $\gtrsim m_{\chi}$ can undergo such scattering and produce the recoil. Naturally the recoil spectra will be suppressed to higher $T_{\mathcal{N}}$ as the incoming $\nu$ flux decreases with energy. Hence it can well explain the peaked event at $T_{\mathcal{N}}\sim 250$ keV due to the atmospheric $\nu$ scattering.

To realize such a scenario we consider the following BSM lagrangian,

$$\mathcal{L}_{\rm BSM}\supset y_{\chi}\overline{\chi}\,P_{L}\,\nu_{\ell}\phi+y_{q}\overline{q}q\phi+\mathrm{H.c.},$$ (1)

where $\phi$ is the scalar mediator with mass $m_{\phi}$. $y_{\chi,q}$ describe respective Yukawa couplings with $\chi$ and quarks $q~(u,d)$. One can also consider other type of mediators with different Lorentz structure giving rise to different event rates. However, for simplicity, in this work we restrict our analysis only to scalar mediators. The qualitative feature for other mediators remain same and one can easily extend our analysis to other mediator types. The differential cross-section of $\nu-\mathcal{N}$ scattering is given by [17],

$$\dfrac{\mathrm{d}\sigma_{\nu\mathcal{N}}}{\mathrm{d}T_{\mathcal{N}}}(E_{\nu},T_{\mathcal{N}}) = \dfrac{m_{\mathcal{N}}C_{S}^{4}}{4\pi(m_{\phi}^{2}+2m_{\mathcal{N}}T_{\mathcal{N}})^{2}}F^{2}(\left|\mathbf{q}\right|^{2})$$ (2)

$$\left(1+\dfrac{T_{\mathcal{N}}}{2m_{\mathcal{N}}}\right)\left(\dfrac{m_{\mathcal{N}}T_{\mathcal{N}}}{E_{\nu}^{2}}+\dfrac{m_{\chi}^{2}}{2E_{\nu}^{2}}\right),~~$$

where for nuclear form factor we adopt Klein-Nystrand (KN) parametrization, which is read as [21],

$$F_{W}(\left|\mathbf{q}\right|^{2})=\dfrac{3\,j_{1}(\left|\mathbf{q}\right|R_{A})}{\left|\mathbf{q}\right|R_{A}(1+a^{2}\left|\mathbf{q}\right|^{2})}.$$ (3)

Here, $j_{1}(x)=\sin(x)/x^{2}-\cos(x)/x$ stands as the spherical Bessel function of order one while $R_{A}=1.23\,A^{1/3}~\mathrm{fm}$ signifies nuclear radius. $A=131$ is the mass number of the Xe atom and $a=0.7~\mathrm{fm}$ is the screening depth. On the other hand, the factor $C_{s}$ encodes the mapping of elementary quark level interactions to nuclear level scattering given by [17],

$$C_{S}^{2}=y_{\chi}y_{q}\left(Z\sum_{q=u,d}\dfrac{m_{p}}{m_{q}}f_{T_{q}}^{(p)}+N\sum_{q=u,d}\dfrac{m_{n}}{m_{q}}f_{T_{q}}^{(n)}\right)$$ (4)

with $Z$ ($m_{p}$) and $N=A-Z$ ($m_{n}$) being the number (mass) of protons and neutrons, respectively, while $m_{q}$ are light quark masses. $f_{T_{q}}^{(p)}$ are hadronic structure functions, and the numerical values of these quantities adopted in this work are [22]

$$f_{T_{u}}^{p}=0.026,\quad f_{T_{d}}^{p}=0.038,\quad f_{T_{u}}^{n}=0.018,\quad f_{T_{d}}^{n}=0.056.$$ (5)

Figure 2: 90% CL region of the $(m_{\chi},\,\sqrt{y_{\chi}y_{q}}/m_{\phi})$ plane (red) in which atmospheric neutrino up-scattering accounts for the LZ excess, defined by $R_{I}=1$ event in $202\text{--}296$ keV while respecting the low energy requirement $R_{II}<10^{-3}R_{\rm BKG}$ over $5.4\text{--}202$ keV. The region closes below $m_{\chi}\sim 0.7$ GeV, where the kinematic threshold falls into the low recoil window and the excess events are excluded by the $R_{II}$ criterion, and above $m_{\chi}\sim 2$ GeV, where the atmospheric neutrino flux above $E_{\nu}^{\rm min}$ is too suppressed to yield one event.

Compared to the elastic scattering of neutrino, the expected spectra of such up-scattering has a strong dependence on the produced BSM particle $\chi$. The minimum (maximum) recoil energy is given by,

$$T_{\mathcal{N}}^{\rm min/max}=\dfrac{2m_{\mathcal{N}}{(E_{\nu}^{\mathrm{max}})}^{2}-m_{\chi}^{2}\left(E_{\nu}^{\mathrm{max}}+m_{\mathcal{N}}\right)\pm E_{\nu}^{\mathrm{max}}\sqrt{4m_{\mathcal{N}}^{2}{(E_{\nu}^{\mathrm{max}})}^{2}-4m_{\mathcal{N}}m_{\chi}^{2}\left(E_{\nu}^{\mathrm{max}}+m_{\mathcal{N}}\right)+m_{\chi}^{4}}}{2m_{\mathcal{N}}\left(2E_{\nu}^{\mathrm{max}}+m_{\mathcal{N}}\right)}.$$ (6)

The dependence of the kinematic cut off (of $T_{\mathcal{N}}$) on $m_{\chi}$ has an interesting consequence. To produce a particle with $m_{\chi}=2$ GeV and $E_{\nu}^{\max}\sim$ few GeV (above which the flux is suppressed), the minimum recoil energy is $\sim 200$ keV. This indeed explains the absence of any tail-like spectrum at lower recoil energies. On the other hand, such a scattering is also forbidden for solar (DSNB) neutrinos with energy $\ll m_{\chi}$. The recoil spectrum is given by,

$$\frac{\mathrm{d}R}{\mathrm{d}T_{\mathcal{N}}}=t_{\rm exp}\mathcal{E}N_{T}~\int_{E_{\nu}^{\mathrm{min}}}^{E_{\nu}^{\mathrm{max}}}\mathrm{d}E_{\nu}~\sum_{\ell}~\dfrac{\mathrm{d}\Phi_{\nu_{\ell}}}{\mathrm{d}E_{\nu}}~\dfrac{\mathrm{d}\sigma_{\nu_{\ell}\mathcal{A}}}{\mathrm{d}T_{\mathcal{N}}}~\Theta(T_{\mathcal{N}}-T_{\mathcal{N}}^{\mathrm{min}})~\Theta(T_{\mathcal{N}}^{\mathrm{max}}-T_{\mathcal{N}}),$$ (7)

where we consider all different components on $\nu$ flux with all three flavors. $t_{\rm exp}$ and $N_{A}$ signify time of exposure and number of target nucleus. $E_{\nu}^{\mathrm{max}}$ denotes the maximum energy of each component and $E_{\nu}^{\mathrm{min}}$ describes the minimum energy required to up-scatter,

$$E_{\nu}^{\mathrm{min}}(T_{\mathcal{N}})=\dfrac{1}{2}\left(T_{\mathcal{N}}+\sqrt{2m_{\mathcal{N}}T_{\mathcal{N}}+(T_{\mathcal{N}})^{2}}\right)\left(1+\dfrac{m_{\chi}^{2}}{2m_{\mathcal{N}}T_{\mathcal{N}}}\right).$$ (8)

The event rate for our region of interest is given by

$$R_{I}=~\int_{202{\rm~keV}}^{296{\rm~keV}}\mathrm{d}T_{\mathcal{N}}~\frac{\mathrm{d}R}{\mathrm{d}T_{\mathcal{N}}}.$$ (9)

To explain the observed event, we require $R_{I}=1$. To prevent any excess events in other energy ranges, we also impose the criterion,

$$R_{II}=~\int_{5.4{\rm~keV}}^{202{\rm~keV}}\mathrm{d}T_{\mathcal{N}}~\frac{\mathrm{d}R}{\mathrm{d}T_{\mathcal{N}}}<10^{-4}R_{\rm BKG}.$$ (10)

$R_{\rm BKG}=1713$ stands as the SM background in the low recoil energy range [8]. This crude cut on the excess events in lower recoil energy ensures that we BSM events in that range remain negligible.

The predicted recoil spectrum for a representative benchmark parameter of $m_{\chi}=2$ GeV and $\sqrt{y_{\chi}y_{q}}/m_{\phi}=0.25$ GeV^-1 is shown in Fig. 1. While the SM neutrino background falls steeply and is further suppressed by the Helm form factor zero near $T_{\mathcal{N}}\simeq 90$ keV, the up-scattering contribution switches on only above $T_{\mathcal{N}}^{\rm min}$ and peaks almost where the LZ event is observed. Below $T_{\mathcal{N}}^{\rm min}$ i.e. at the low energy tail, up-scattering contribution generates no events.

The region of parameter space that accounts for the observed event is presented in Fig. 2. The band is closed from all sides: for $m_{\chi}\lesssim 0.7$ GeV the kinematic threshold moves into the low recoil window and the associated events violate our $R_{II}$ criterion, whereas for $m_{\chi}\gtrsim 2$ GeV the atmospheric flux above $E_{\nu}^{\rm min}$ becomes too small to yield a single event. Within this window, the required coupling strength $\sqrt{y_{\chi}y_{q}}/m_{\phi}$ grows with $m_{\chi}$ to compensate for the shrinking flux, and remains consistent with all existing constraints discussed below.

High energy colliders like LHC so far set the strongest constraint on such BSM operator ($y_{\chi}y_{q}/m_{\phi}^{2}$) through mono-jet signals (i.e. $pp\to\chi+\bar{\nu}$+jet) [23]. However, the center of mass energy (COM) considered in the LHC analysis was taken $\sqrt{s}=$13 TeV and thus models featuring mediator $m_{\phi}\lesssim 13$ TeV can not be realized in the EFT approach. On the other hand, light mediators with mass $\ll\sqrt{s}$ will be produced on-shell and thus the LHC constraints do not apply for our low energy scenario. A similar argument holds for other searches with higher COM energies. Astrophysical constraints like supernova cooling [24] and meson decay constraints [25] also do not hold for $m_{\chi}\gtrsim 800$ MeV.

In this work, we demonstrate that atmospheric neutrino conversion to a heavy dark fermion through a BSM scalar mediator can in principle explain the single isolated nuclear recoil event observed by LZ. The heavy final state $\chi$ with mass $\sim$GeV, imposes a kinematic cutoff in the lower recoil energies, mimicking an isolated BSM event. Indeed, a single event is insufficient to pin down uniquely the microscopic BSM physics. In the near future, other direct search facilities with extremely low backgrounds can offer an ideal test of this explanation. Larger detectors like KamLAND [26] and Borexino [27] can also probe the best-fit parameter space identified here, which we keep for our future work.

Acknowledgement. SJ thanks Xiaoyong Chu and ICTP-AP, Beijing, for arranging an academic visit where this work was accomplished. SJ acknowledges financial support from the National Natural Science Foundation of China (12425506 and 12375101) and State Key Laboratory of Dark Matter Physics.