Light Dark Matter Search with 7.8 Tonne-Year of Ionization-Only Data in XENONnT

Abstract

We report on a blinded search for dark matter (DM) using ionization-only (S2-only) signals in XENONnT with a total exposure of $7.83\mathrm{tonne}\times\mathrm{year}$ over 579 days in three science runs. Dedicated background suppression techniques and the first complete S2-only background model in XENONnT provide sensitivity to nuclear recoils of [0.5, 5.0] $\mathrm{keV_\mathrm{nr}}$ and electronic recoils of [0.04, 0.7] $\mathrm{keV_\mathrm{ee}}$. No significant excess over the expected background is observed, and we set 90\% confidence level upper limits on spin-independent DM--nucleon and spin-dependent DM--neutron scattering for DM masses between 3 and 8 $\mathrm{GeV}/c^2$, as well as on DM--electron scattering, axion-like particles, and dark photons, improving on previous constraints. For spin-independent DM--nucleon scattering, we exclude cross sections above $6.0\times10^{-45} $cm$^2$ at a DM mass of 5 $\mathrm{GeV}/c^2$, pushing the XENONnT sensitivity closer to the region where coherent elastic neutrino-nucleus scattering ($\text{CE}ν\text{NS}$) becomes an irreducible background.

Figures (4)

• Figure 1: Top: The SI DM-nucleon spectra for 3, 6, and 8 $\mathrm{GeV}/c^2$ masses with a cross section of $4.4\times10^{-45}$ cm$^2$, shown before (dotted) and after (solid) applying the total signal detection efficiency. The solar $^8$B CE$\nu$NS spectrum (gray) is shown for comparison. Bottom: The S2-only signal efficiency (black) is the product of the individual efficiencies for the S1 region of interest (ROI) (blue), S2 ROI (green), and S2 quality selection (red), compared to the S1-S2 analysis efficiency XENON:2024hup (gray dashed). All efficiencies shown are exposure-weighted averages over SR0, SR1, and SR2.
• Figure 2: Comparison of observed events (black dots) with the expected background components in the enlarged ROI of all science runs. Colored contours show the modeled distributions of cathode (red), delayed electrons (blue), accidental electrons (orange), and ${}^8\mathrm{B}$ CE$\nu$NS (green), with dark and light shades representing the $1\sigma$ and $2\sigma$ contours. A signal distribution (cyan) $1\sigma$ contour is shown, assuming a flat spectrum of signal recoil energy. The validation is performed in four dimensions: cS2 (energy scale), S2 CNF score (ambience), cathode BDT score and S2 width (waveform). Projections along each axis are shown as top and side histograms; shaded bands indicate background model statistical and systematic uncertainties.
• Figure 3: Comparison of observed events with expected background components in the science ROI, combining all science runs. The description is the same as in Fig. \ref{['fig:background_origin']}. After applying all data selections, the remaining background has a similar shape to the signal. The side projections show the best-fit model and data (black), showing no significant excess. The inference is performed in the $\mathrm{cS2}$ dimension.
• Figure 4: The 90% confidence level upper limits on the DM--particle scattering with 1$\sigma$ (green) and 2$\sigma$ (yellow) sensitivity bands. The black dashed (solid) lines show limits before (after) $-1\sigma$ power-constrained limit (PCL). Previous published results from XENON10 XENON10:2011prx, XENON1T XENON:2019gfn, XENONnT XENON:2022ltvXENON:2024hupXENON:2024znc, LUX-ZEPLIN LZ:2023poo, PandaX PandaX:2022xqxZhang:2025ajc, DarkSide DarkSide-50:2022qzhDarkSide:2022knj and SuperCDMS SuperCDMS:2019jxx are shown for comparison. The neutrino fog region OHare:2021utq in panel (a) is indicated in gray bands for SI DM--nucleon scattering. The ${}^8\mathrm{B}$ CE$\nu$NS equivalent DM of 5.5 $\mathrm{GeV}/c^2$ with cross-section 4.4e-45cm^2 is shown.