Direct detection of solar chameleons with electron recoil data from XENONnT

TL;DR

This work tests solar chameleon dark-energy models using XENONnT electron recoil data, updating solar flux modeling to include Primakoff production and full-solar magnetic conversion. The detection signal in XENONnT is governed by the effective coupling $\beta_{\text{eff}} = \beta_\gamma M_e^{-4}$, which combines production in the Sun and absorption on Earth; under the DE-scale choice $\Lambda = 2.4~\text{meV}$, the analysis yields a 95% CL upper limit $\log_{10} \beta_{\text{eff}} < -6.9$, largely independent of $\beta_e$ and $n$. The results show that Primakoff production dominates the electron-recoil signal at keV energies, and that the constraints extend across the broader class of inverse-power-law chameleons, not just the $n=1$ case. This demonstrates that existing multi-tonne detectors can probe screened dark-energy models, motivating future multi-target and lower-threshold analyses to distinguish solar chameleons from axions and other light scalars.

Abstract

We reassess prospects for direct detection of solar chameleons, in light of recent progress in modeling their production, and the availability of new XENONnT data. We show that the contribution from Primakoff production in the electric fields of electrons and ions dominates the electron recoil event rate, which is enhanced compared to earlier estimates based on magnetic conversion in the tachocline alone. We argue that the signal is governed by the effective coupling $β_{\text{eff}} \equiv β_γM_e^{-4}$, which encodes the combined effects of production and detection, where $β_γ$ and $M_e$ are the chameleon-photon (conformal) coupling and chameleon-electron disformal coupling scale, respectively. Setting the height of the chameleon potential to the dark energy (DE) scale $Λ\simeq 2.4\,{\text{meV}}$, we show that XENONnT electron recoil data set the upper limit $\log_{10}β_{\text{eff}}<-6.9$. This limit is independent of the conformal matter coupling $β_m$ and index $n$, and applies to the whole class of inverse power-law chameleons, well beyond the $n=1$ case usually studied. We comment on how future multi-target experiments and lower-threshold analyses could distinguish solar chameleons from other light (pseudo)scalar particles such as axions. Our work demonstrates that existing dark matter direct detection experiments can probe regions of parameter space relevant to screened DE models, providing complementary tests to astrophysical and fifth-force searches at no additional experimental cost.

Figures (8)

• Figure 1: Comparison of the two components of the solar chameleon spectrum: Primakoff production from transverse photons in the electric fields of electrons and ions (black solid curve), and magnetic conversion from the solar bulk magnetic field (blue dash-dotted curve and blue band). The width of the blue band reflects the uncertainty on the strength of the solar magnetic field (discussed in Appendix \ref{['app:solarmodel']}). The yellow dashed curve and associated band correspond instead to the magnetic conversion component calculated in Paper II, updated here (to the blue dash-dotted curve and associated band) to account for a few issues in the earlier code, as discussed in Sec. \ref{['subsubsec:completespectrum']}. The spectrum is computed assuming the AGSS09 solar model, while fixing the chameleon parameters to $\beta_m=10^2$, $\Lambda=2.4\,{\text{meV}}$, and $n=1$. Note that the spectrum has been normalized by $\beta_{\gamma}^2$ to factor out the dependence on the chameleon-photon coupling strength, see Eqs. (\ref{['eq:productionprimakoff']},\ref{['eq:productionmagnetic']}).
• Figure 2: Regions of $\beta_m$-$\Lambda$ parameter space where solar chameleons can escape the solar core, or conversely remain trapped. For $n=1$ chameleons the two regions are those above the solid black line and the light gray shaded region below the solid black line respectively, whereas for $n=4$ chameleons these are instead the region above the dash-dotted black line and the dark gray shaded region below the dash-dotted black line respectively. The solid black and dash-dotted black lines correspond to the boundaries defined by the "critical values" $\Lambda_{\text{crit}}^{(n=1)}(\beta_m)$ and $\Lambda_{\text{crit}}^{(n=4)}(\beta_m)$ given by Eqs. (\ref{['eq:lcritn1']},\ref{['eq:lcritn4']}), where the effective mass of the ($n=1$ or $n=4$) chameleon in the solar core is equal to the temperature of the latter, $m_{\text{eff}}(\rho_{\text{core}})=T_{\text{core}}$. The blue, orange, and green shaded regions are those excluded by atom interferometry, levitated force sensors, and torsion balance experiments respectively, for $n=1$ chameleons. For $n=4$ chameleons the atom interferometry and torsion balance limits (not shown here to avoid overcrowding the figure) are weaker, and the levitated force sensor constraints do not apply. The horizontal dashed line corresponds to the DE scale $\Lambda \simeq 2.4\,{\text{meV}}$.
• Figure 3: Benchmark example of solar chameleon fit to the XENONnT electron recoil event rate, with chameleon parameters fixed to $\beta_e=10$, $\beta_{\gamma}=10^{9.4}$, $M_e=10^4\,{\text{eV}}$, $\Lambda=2.4\,{\text{meV}}$, and $n=1$. This choice of parameters corresponds to an effective coupling strength $\beta_{\text{eff}}=10^{-6.6}$ [see Eq. (\ref{['eq:betaeff']})], which is allowed by the data (see Fig. \ref{['fig:betaeffupperlimit']}). The XENONnT data is given by the black datapoints, the solid red curve is the XENONnT energy-dependent background $B_0$, and the dashed blue curve is the total (background plus signal) solar chameleon event rate. Detector efficiency and energy resolution effects are already accounted for when plotting the total signal.
• Figure 4: As in Fig. \ref{['fig:benchmarksignal']}, but isolating the solar chameleon contribution to the total signal, for different values of the effective coupling strength $\beta_{\text{eff}}$ between $10^{-6.8}$ and $10^{-6.2}$, as indicated by the color coding. Note that the solar chameleon contribution needs to be summed to the energy-dependent background $B_0$ in order to be meaningfully compared to the XENONnT electron recoil data. It is visually clear that, for most of the plotted values of $\beta_{\text{eff}}$, the total (chameleon+background) signal would exceed the XENONnT measurements, and the frequentist limits we will later obtain on $\beta_{\text{eff}}$ should therefore fall broadly within the plotted range.
• Figure 5: As in Fig. \ref{['fig:benchmarksignal']}, but for an effective coupling strength $\beta_{\text{eff}}=10^{-6.2}$, and separating the contributions to the chameleon signal arising from different components of the solar chameleon spectrum: Primakoff production from transverse photons in the electric fields of electrons and ions (dashed blue curve), and magnetic conversion from the solar bulk magnetic field within the radiative zone (dashed violet curve), tachocline (dashed green curve), and convective zone (dotted yellow curve). Of all the components, the Primakoff one clearly dominates by a large margin, given the recoil energy range of interest to XENONnT (compare against the spectrum of Fig. \ref{['fig:spectrum']}). The inset zooms into the low-energy region where the relative contribution from the bulk magnetic field components is least insignificant. Note that the chosen value of $\beta_{\text{eff}}$ is excluded by the data (see Fig. \ref{['fig:betaeffupperlimit']}; note that the four components need to be summed to the energy-dependent background $B_0$ in order to be meaningfully compared to the XENONnT electron recoil measurements), and has been chosen to grossly exaggerate the difference between the components, since the bulk magnetic field ones would otherwise be impossible to see by eye.