Inelastic Dark Matter, Non-Standard Halos and the DAMA/LIBRA Results

TL;DR

This paper examines whether inelastic dark matter can reconcile the DAMA/LIBRA modulation signal with null results from other direct detection experiments by exploring non-standard Galactic velocity distributions. Using Via Lactea and Dark Disc simulations, it shows that a Via Lactea–like halo with a suitably chosen velocity profile can significantly expand the compatible iDM parameter space, particularly at high WIMP masses, while Dark Disc provides little improvement. The analysis also highlights that astrophysical uncertainties, notably the iodine quenching factor and the local escape velocity, strongly influence conclusions, and that heavy-element detectors are essential for a definitive test of iDM in light of these velocity distributions. The work underscores the sensitivity of DM interpretation to halo modeling and motivates future data from heavy-nucleus experiments to decisively test DAMA's claim.

Abstract

The DAMA collaboration have claimed to detect particle dark matter (DM) via an annual modulation in their observed recoil event rate. This appears to be in strong disagreement with the null results of other experiments if interpreted in terms of elastic DM scattering, while agreement for a small region of parameter space is possible for inelastic DM (iDM) due to the altered kinematics of the collision. To date most analyses assume a simple galactic halo DM velocity distribution, the Standard Halo Model, but direct experimental support for the SHM is severely lacking and theoretical studies indicate possible significant differences. We investigate the dependence of DAMA and the other direct detection experiments on the local DM velocity distribution, utilizing the results of the Via Lactea and Dark Disc numerical simulations. We also investigate effects of varying the solar circular velocity, the DM escape velocity, and the DAMA quenching factor within experimental limits. Our data set includes the latest ZEPLIN-III results, as well as full publicly available data sets. Due to the more sensitive dependence of the inelastic cross section on the velocity distribution, we find that with Via Lactea the DAMA results can be consistent with all other experiments over an enlarged region of iDM parameter space, with higher mass particles being preferred, while Dark Disc does not lead to an improvement. A definitive test of DAMA for iDM requires heavy element detectors.

Figures (12)

• Figure 1: Recoil energy spectrum for scattering off of tungsten. On the left the inelastic scattering rates are shown for the two choices of form factor: the Fermi Two-Parameter (dashed) and Helm (solid) form factor. Both choices show that iDM leads to a suppression of low-energy events. The right panel shows the recoil energy spectrum for elastic scattering; in this case the difference between the two form factors is negligible. For elastic scattering the recoil spectrum peaks at low $E_R$. All calculations assume $M_\chi = 200 \:$GeV, $\sigma_n = 10^{-40}$ cm$^{2}$ and the SHM velocity distribution truncated at $v_{esc} = 500$ km/s.
• Figure 2: Here we show the variation in the exclusion limits set by the experiments as $\delta$ is varied and $M_{\chi}$ is held constant. These limits are calculated using the SHM. The preferred region of parameter space for the DAMA results is shown at 90% and 99.5%, and the DAMA best fit point is plotted with a dot. As one can see there is a small region of agreement between all experiments for low masses and $\delta \sim 130$ keV. At higher masses both CRESST-II and CDMS II exclude the DAMA results and the region of agreement with the other experiments is greatly reduced.
• Figure 3: Change in limits when the SHM is replaced by the $\text{VL}_{220}$ halo, cf, Figure 2. At low masses there is a smaller region of agreement between CRESST-II and DAMA, with CRESST-II almost excluding DAMA at the 90% level over all masses. At high masses, however, CRESST-II and CDMS II are significantly less constraining on the DAMA region than for the SHM, this can be seen by comparing the bottom right panels of both figures. One can also see that the typical cross sections are an order of magnitude higher for the $\text{VL}_{220}$ halo at low masses than for the SHM (note also that the lower panels have a different scale for the cross section).
• Figure 4: The allowed parameter space for fixed $\delta=100$keV and varying $M_{\chi}$. For the SHM (left panel) and this value of $\delta$, CDMS II excludes the DAMA region at 90%. For the $\text{VL}_{220}$ halo (right panel) the tightest constraints are set by CRESST-II, and there is agreement between DAMA and CDMS II up to high WIMP masses. Again one can see that the typical allowed cross sections are an order of magnitude higher for the $\text{VL}_{220}$ halo.
• Figure 5: Detectable particle distributions in a germanium detector as a function of tangential velocity. The detectable region is shaded in grey. The solid, dashed and dot-dashed lines show the SHM, $\text{VL}_{220}$ and $\text{VL}_{270}$ distributions respectively. The left edge of the grey region corresponds to $M_{\chi}=500 \text{ GeV}$.