The DAMA/LIBRA experiment at Gran Sasso has reported for over two decades a single-hit (only one detector fires) annual modulation in ultra-radiopure NaI(Tl) scintillators that satisfies all the requirements of the model-independent dark-matter annual-modulation signature: period of one year, phase near June 2, presence only at low energy, and absence in multiple-hit events. The final DAMA/LIBRA-phase2-empowered configuration pushed the software energy threshold down to 0.5 keV and traced the modulation spectrum over the full 0.5-6 keV region, reaching a combined confidence level of 15.3 sigma in the 2-6 keV(ee) window over 3.40 ton-year, after which the setup was dismantled in fall 2024. Yet the identity of what produces the signal remains undetermined, and the dark-matter interpretation is now under severe quantitative strain: the same-target experiments ANAIS-112 (six years, 4 sigma incompatible) and COSINE-100 (6.4 years, more than 3 sigma incompatible) observe no modulation, and global-fit analyses find a tension exceeding 5 sigma under physically motivated hypotheses unless contrived cancellations are invoked. The problem is to determine the origin of the DAMA/LIBRA annual modulation: either a particle and halo scenario quantitatively consistent with the DAMA spectrum and with all null results, or a demonstrated systematic mechanism reproducing the complete signature.

Background

A large body of astrophysical and cosmological evidence implies that most of the matter in galaxies is dark matter, hypothetical non-luminous particles beyond the Standard Model; a leading strategy to identify them is direct detection, in which an underground detector searches for the elastic scattering of Galactic dark-matter particles off atomic nuclei. Because the Earth orbits the Sun while the Sun moves through the Galactic dark-matter halo, the flux of dark matter impinging on a detector is expected to vary over the year, producing an annual modulation of the event rate with a period of exactly one year, a maximum near June 2, and an amplitude at the percent level that should appear only at low recoil energies and only in single-hit events (those in which just one detector fires, since a dark-matter particle scatters only once). This multi-feature signature is called model-independent because none of its features assumes a particular particle candidate, in contrast to model-dependent searches that assume specific interaction cross sections and halo distributions.

The DAMA/LIBRA experiment at the Laboratori Nazionali del Gran Sasso operates about 250 kg of highly radiopure NaI(Tl) scintillators (sodium-iodide crystals read out by photomultipliers). Scintillator yields are conventionally quoted in keV(ee), electron-equivalent energy: nuclear recoils produce less light than electrons of the same kinetic energy, a suppression encoded in the quenching factor. Since the 1990s, through the DAMA/NaI, DAMA/LIBRA-phase1 and DAMA/LIBRA-phase2 configurations, the collaboration has reported a single-hit annual modulation meeting all the signature requirements at high confidence: as reported in [1], 2.86 ton-year over 22 annual cycles with a 13.7 sigma confidence level in the 2-6 keV(ee) region and a modulation amplitude of about 0.01 counts/day/kg/keV. Lowering the energy threshold matters because different candidate particles - weakly interacting massive particles (WIMPs) scattering on sodium or iodine, axion-like particles, and mirror (shadow-world) matter - predict different shapes of the modulation amplitude SmS_m as a function of energy. A software-threshold reanalysis of phase2 data added a preliminary modulation point at 0.75 keV(ee), and a hardware upgrade (new low-background preamplifiers and 14-bit digitizers) created the DAMA/LIBRA-phase2-empowered configuration with threshold below 1 keV(ee), intended to measure how far the modulation extends toward lower energies.

Two experiments were built to test the claim with the same target material. ANAIS-112 at Canfranc uses 112.5 kg of NaI(Tl); COSINE-100 at Yangyang used a comparable NaI(Tl) array. Both apply time-series analyses directly analogous to DAMA's. In addition, COSINUS at Gran Sasso is developing cryogenic NaI scintillating calorimeters in which a sodium-iodide crystal is instrumented with both a phonon (heat) channel read out by superconducting transition-edge sensors and the conventional scintillation-light channel; the dual readout measures the total recoil energy and the light yield event by event, separating electron recoils from nuclear recoils and thereby testing the DAMA claim without relying on the scintillation-only quenching assumptions that complicate comparisons between DAMA and the other NaI(Tl) experiments.

Problem Statement

Determine the origin of the DAMA/LIBRA annual modulation signal. The full experimental situation to be explained is: a single-hit annual modulation with period of one year, phase near June 2, modulation amplitude of order 0.01 counts/day/kg/keV, present over the 0.5-6 keV(ee) energy interval (traced down to 0.5 keV(ee) by the final empowered configuration, 3.40 ton-year total exposure at 15.3 sigma confidence level in 2-6 keV(ee)), absent in multiple-hit events and absent above 6 keV(ee); and, in tension with it, the null annual-modulation results of ANAIS-112 (six years) and COSINE-100 (6.4 years) in the same NaI(Tl) target material, incompatible with the DAMA amplitude at 4 sigma and more than 3 sigma respectively. A complete answer must take one of two forms and account for the entire data pattern, not merely a subset of it: (i) a particle-physics and halo scenario (for example a WIMP, axion-like particle, or mirror-matter interaction, including the needed treatment of sodium and iodine quenching and detector response) that quantitatively reproduces the measured Sm(E)S_m(E) spectrum - matching the published modulation amplitudes of the cited DAMA/LIBRA analyses bin by bin within their quoted uncertainties - as well as the period, the phase, the single-hit-only character and the absence of modulation above 6 keV(ee), while remaining compatible with the ANAIS-112 and COSINE-100 null results and with limits from other target materials, without requiring finely tuned cancellations such as opposite-sign sodium and iodine modulations; or (ii) a demonstrated instrumental, background or environmental mechanism that quantitatively reproduces that same complete set of signature features in the DAMA setup.

References

  1. R. Bernabei, P. Belli, F. Cappella, V. Caracciolo, R. Cerulli, C. J. Dai, A. d'Angelo, A. Incicchitti, A. Leoncini, X. H. Ma, V. Merlo, F. Montecchia, X. D. Sheng, Z. P. Ye, "Dark Matter: DAMA/LIBRA and its perspectives", Proceedings of IDM 2022, arXiv:2209.00882, published as SciPost Phys. Proc. 12, 025 (2023), doi:10.21468/SciPostPhysProc.12.025, https://arxiv.org/abs/2209.00882
  2. R. Bernabei, P. Belli, A. Bussolotti, V. Caracciolo, F. Cappella, R. Cerulli, C. J. Dai, A. d'Angelo, N. Ferrari, A. Incicchitti, A. Leoncini, X. H. Ma, A. Mattei, V. Merlo, F. Montecchia, X. D. Sheng, Z. P. Ye, "Further results from DAMA/LIBRA-phase2 and perspectives", Nucl. Phys. At. Energy 22, 329 (2021), doi:10.15407/jnpae2021.04.329, https://doi.org/10.15407/jnpae2021.04.329
  3. R. Bernabei, P. Belli, A. Bussolotti, F. Cappella, V. Caracciolo, R. Cerulli, C. J. Dai, A. d'Angelo, A. Incicchitti, A. Leoncini, X. H. Ma, A. Mattei, V. Merlo, F. Montecchia, X. D. Sheng, Z. P. Ye, "Model-Independent Results from DAMA/LIBRA-phase2-empowered", Nucl. Phys. At. Energy 27, 105 (2026), doi:10.15407/jnpae2026.02.105, https://doi.org/10.15407/jnpae2026.02.105
  4. J. Amaré et al. (ANAIS collaboration), "Towards a Robust Model-Independent Test of the DAMA/LIBRA Dark Matter Signal: ANAIS-112 Results with Six Years of Data", Phys. Rev. Lett. 135, 051001 (2025), arXiv:2502.01542, doi:10.1103/ntnl-zrn9, https://arxiv.org/abs/2502.01542
  5. N. Carlin et al. (COSINE-100 collaboration), "COSINE-100 full dataset challenges the annual modulation signal of DAMA/LIBRA", Sci. Adv. 11, eadv6503 (2025), arXiv:2409.13226, doi:10.1126/sciadv.adv6503, https://arxiv.org/abs/2409.13226
  6. N. Carlin et al. (COSINE-100 and ANAIS-112 collaborations), "Combined Annual Modulation Dark Matter Search with COSINE-100 and ANAIS-112", Phys. Rev. Lett. 135, 121002 (2025), arXiv:2503.19559, doi:10.1103/9j7w-qp1c, https://arxiv.org/abs/2503.19559
  7. G. Busoni, J. M. Cornell, W. Handley, F. Kahlhoefer, A. Kvellestad, M. Pitts, L. Street, A. C. Vincent, M. White, "DAMA/LIBRA and dark matter: decisive tension or contrived cancellation", JCAP 08 (2026) 047, arXiv:2510.05216, doi:10.1088/1475-7516/2026/08/047, https://arxiv.org/abs/2510.05216
  8. G. Angloher et al. (COSINUS collaboration), "COSINUS model-independent sensitivity to the DAMA/LIBRA dark matter signal", JCAP 09 (2025) 072, arXiv:2504.18263, doi:10.1088/1475-7516/2025/09/072, https://arxiv.org/abs/2504.18263
  9. G. Angloher et al. (COSINUS collaboration), "COSINUS: a model-independent challenge to the DAMA/LIBRA dark matter claim using cryogenic NaI detectors", Commun. Phys. 9, 213 (2026), doi:10.1038/s42005-026-02620-9, https://doi.org/10.1038/s42005-026-02620-9
  10. R. S. James, K. Rule, E. Barberio, V. U. Bashu, L. J. Bignell, I. Bolognino, G. Brooks, S. S. Chhun, et al., "The DAMA/LIBRA signal: an induced modulation effect?", arXiv:2408.08697 (2024), https://arxiv.org/abs/2408.08697