Using the Energy Spectrum at DAMA/LIBRA to Probe Light Dark Matter

TL;DR

This paper tests whether a light WIMP can explain the DAMA/LIBRA modulation by incorporating the observed recoil-energy spectrum and constraints from unmodulated rates. Under a standard Maxwellian halo and spin-independent scattering, the spectral data disfavor the light-mass solution and the high-mass solution is excluded by other direct-detection experiments. Modifications to the astrophysical velocity distribution or to particle physics (e.g., DM streams or inelastic scattering) can reopen narrow windows, but such scenarios face significant model-building and experimental challenges. The work highlights the critical role of spectral information and cross-experiment constraints in assessing DM interpretations of DAMA/LIBRA.

Abstract

A weakly interacting massive particle (WIMP) weighing only a few GeV has been invoked as an explanation for the signal from the DAMA/LIBRA experiment. We show that the data from DAMA/LIBRA are now powerful enough to strongly constrain the properties of any putative WIMP. Accounting for the detailed recoil spectrum, a light WIMP with a Maxwellian velocity distribution and a spin-independent (SI) interaction cannot account for the data. Even neglecting the spectrum, much of the parameter space is excluded by limits from the DAMA unmodulated signal at low energies. Significant modifications to the astrophysics or particle physics can open light mass windows.

Figures (4)

• Figure 1: We show the region consistent with the DAMA/LIBRA modulation signal (at 68%, 90% and 99% CL), accounting for the detailed modulation energy spectrum between 2-6 keVee and the single bin modulation between 6-14 keVee (colored contours). Around this, we show the envelope (gray line) consistent with the DAMA/LIBRA modulation signal at 99% CL, fitting the total modulation rate from the two bins, 2-6 keVee and 6-14 keVee (i.e., no spectral information). Also shown are bounds calculated by considering data from the CDMS-II (Si) and XENON10 experiments. The limit curve labeled "DAMA-Total" arises from the self-consistency with the total (unmodulated) rates in the DAMA-LIBRA data, as described in Section \ref{['sec:DAMAvDAMA']}. At right, we zoom in on the light mass region. When zooming in, we study variations about the minimum $\chi^2$ in the low mass region, as described in the text.
• Figure 2: We show the modulation spectra for the best fit point where scattering off iodine dominates, $m_{\chi}$ = 77 GeV (dot-dashed orange), and three points where scattering off of sodium dominates. The best fit point off sodium is $m_{\chi}$ = 12 GeV (solid red). We also show $m_{\chi} = 2$ GeV (dashed green) and $m_{\chi}=7$ GeV (dotted blue). The points with error bars are the published DAMA/LIBRA data.
• Figure 3: Similar to Fig. \ref{['fig:NewContours']}, but incorporating a stream of DM, with $v_{str} = 900 \; {\rm km/s},$ and $\sigma_{str}=20 \; {\rm km/s}$. Insets: magnification of parameter space near 2 GeV and 4 GeV. The 2 GeV relies on both channeling and the stream, the 4 GeV region arises from unchanneled events from the stream.
• Figure 4: Similar to Fig. \ref{['fig:NewContours']}, but including only inelastic scatterings, with $\delta = 35 \;{\rm keV}$.