Global interpretation of direct Dark Matter searches after CDMS-II results

TL;DR

This work reexamines the DAMA annual modulation in light of CDMS-II and XENON10 using a global fit across four DM-nucleus interaction classes: elastic and inelastic with spin-independent ($e$SI, $i$SI) and spin-dependent ($e$SD, $i$SD) scattering. It shows that elastic scenarios (both SI and SD) struggle to reconcile DAMA with other limits, while inelastic spin-dependent scattering off protons (iSD) provides a consistent DAMA explanation compatible with all other experiments, aided by a simple tensor-current, Majorana-mass-splitting model. For inelastic SI (iSI), conventional iDM is disfavored by CRESST-II, and a low-mass, channeling-driven region requires fine-tuning. The analysis highlights how DAMA’s channeling hypothesis and the Galactic escape-velocity tail crucially shape viable regions, and it notes that the two CDMS events are not a robust DM signal when combined with other constraints. Overall, the paper narrows the viable DM parameter space and proposes a concrete iSD mechanism that can accommodate DAMA while remaining consistent with cross-experiment limits.

Abstract

We perform a global fit to data from Dark Matter (DM) direct detection experiments, including the recent CDMS-II results. We discuss possible interpretations of the DAMA annual modulation signal in terms of spin-independent and spin-dependent DM-nucleus interactions, both for elastic and inelastic scattering. We find that in the spin-dependent inelastic scattering off protons a good fit to all data is obtained. We present a simple toy model realizing such a scenario. In all the remaining cases the DAMA allowed regions are disfavored by other experiments or suffer from severe fine tuning of DM parameters with respect to the galactic escape velocity. Finally, we also entertain the possibility that the two events observed in CDMS-II are an actual signal of elastic DM scattering, and we compare the resulting CDMS-II allowed regions to the exclusion limits from other experiments. In this arXiv version of the manuscript we also provide in appendix A the updated fits including recent CoGeNT results.

Figures (9)

• Figure 1: DAMA allowed regions (90% and 3$\sigma$ CL) and constraints from other experiments (90% CL) for SI scattering (left) and SD scattering off protons (right). Shaded DAMA regions have been obtained assuming the channeling effect according to Bernabei:2007hw, while the black contour curves correspond to no channeling.
• Figure 2: Comparison of different data sets for CDMS and XENON for SI elastic scattering. XENON 07 refers to the analysis from Angle:2007uj, while XENON 09 corresponds to the re-analysis of the same data from XENON:2009xb. The CDMS 2008 and 2009 data sets are from Ahmed:2008eu and :2009zw, respectively. For XENON, different assumptions on the effective light yield $L_{\rm eff}$ are used. Thick curves are based on the measurement Sorensen:2008ec, while thin curves are based on the alternative data set Manzur:2009hp.
• Figure 3: Allowed regions for CDMS 2009 data (1$\sigma$, 90% and 3$\sigma$ CL), DAMA (90% and 3$\sigma$ CL), and constraints from other experiments (90% CL) for elastic SI scattering (left), SD scattering off protons (middle), and SD scattering off neutrons (right).
• Figure 4: Global fit of inelastic, spin-independent DM. We show projections of the 3-dimensional regions at 90% and 3$\sigma$ CL onto the three 2-dimensional planes, by minimizing the global $\chi^2$ in each case with respect to the third (un-displayed) parameter. Confidence regions are defined for 2 dof. The fit includes CDMS (2008 + 2009 data), XENON (2009 analysis), DAMA, CRESST-II, ZEPLIN-III, and KIMS. The dark shaded regions are the allowed regions defined with respect to the global minimum, while the light shaded regions are defined relative to a local minimum, which by itself is disfavored relative to the global minimum with $\Delta\chi^2 = 16.1$. The open contours correspond to DAMA data only. In the upper panel we show also contours of $v_\mathrm{min} = v_\mathrm{esc} + v_\mathrm{earth}$ for a recoil energy of $E = 3$ keV, with and without quenching of iodine scatters, where $v_\mathrm{earth}$ is the velocity of the earth relative to the halo, depending on the time in the year.
• Figure 5: DAMA allowed regions (90% and 3$\sigma$ CL) and constraints from other experiments (90% CL) for inelastic DM scattering with SI interactions (left) and SD interactions with protons (right). We show regions in the $(m_\chi, \sigma_p)$ plane (2 dof) for fixed DM mass splitting $\delta$. The upper panels ($\delta = 130$ keV) correspond to a signal in DAMA from quenched events on iodine, whereas the lower panels ($\delta = 40$ keV) correspond to channeled events on iodine according to Bernabei:2007hw.