Inelastic Dark Matter

TL;DR

Smith and Weiner propose inelastic dark matter with a small mass splitting $\delta$ (roughly $50$–$100$ keV) to resolve the DAMA/CDMS tension by altering scattering kinematics and enhancing modulation signals. They develop the direct-detection formalism including inelastic transitions, show that DAMA can remain sensitive while CDMS is suppressed, and identify viable parameter space with a concrete sneutrino-based supersymmetric model. They discuss cosmological and indirect-detection uncertainties, noting that these constraints do not rule out the scenario, and highlight upcoming experiments (e.g., CDMS Soudan, GENIUS, CRESST tungsten) as critical tests, with distinctive recoil-spectrum signatures providing clear discriminants. The work offers a plausible mechanism and concrete predictions for future searches, potentially transforming the interpretation of direct-detection signals for non-baryonic dark matter.

Abstract

Many observations suggest that much of the matter of the universe is non-baryonic. Recently, the DAMA NaI dark matter direct detection experiment reported an annual modulation in their event rate consistent with a WIMP relic. However, the Cryogenic Dark Matter Search (CDMS) Ge experiment excludes most of the region preferred by DAMA. We demonstrate that if the dark matter can only scatter by making a transition to a slightly heavier state (Delta m ~ 100kev), the experiments are no longer in conflict. Moreover, differences in the energy spectrum of nuclear recoil events could distinguish such a scenario from the standard WIMP scenario. Finally, we discuss the sneutrino as a candidate for inelastic dark matter in supersymmetric theories.

Figures (6)

• Figure 1: Ratio of total events in iWIMP scenario to ordinary WIMP as a function of splitting $\delta$ for DAMA (solid line) and CDMS (dashed), with $m_{\chi}=50\rm GeV$. For DAMA we have integrated the total events in the $2-10 \rm keV$ energy region, while for CDMS we have integrated in the $10 - 100 \rm keV$ region. For large $\delta$ ($>$ 100 keV), the finite value of the the galactic escape velocity can become important, yielding larger suppressions than shown. This effect is stronger for CDMS than for DAMA.
• Figure 2: Annual modulation of event rate with average normalized to one in the inelastic WIMP scenario (solid line) and standard WIMP scenario (dashed), with $\delta=100 \rm keV$ and $m_{\chi}=50\rm GeV$.
• Figure 3: Normalized modulation ($S_{m}$) as a function of energy for ordinary WIMP scenario (solid), inelastic WIMP scenario with $\delta = 100 \rm keV$ (dashed), and inelastic WIMP scenario with $\delta= 150 \rm keV$ (dotted), all with $m_{\chi}=60 \rm GeV$.
• Figure 4: Normalized spectrum of events at CDMS for ordinary WIMP (solid) and inelastic WIMP (dashed) with $\delta = 100 \rm keV$, both with $m_{\chi}=50 \rm GeV$.
• Figure 5: Regions satisfying both DAMA and CDMS constraints in the $\delta-\sigma_{n}$ plane, for (a) $m_{\chi}=50\rm GeV$, (b) $m_{\chi}=100\rm GeV$, (c) $m_{\chi}=300\rm GeV$. In each plot, the shaded region has an integrated signal in the $2-6\rm keV$ energy range consistent with the DAMA $3\sigma$ region. The solid line gives the CDMS constraint and the dashed line gives the limit from an assumption of the absence of signal in the high energy bins at DAMA. The dot-dashed line gives the upper bound arising from Xe pulse shape analysis limits. The dark shaded region satisfies all constraints simultaneously.